From owner-ietf-usefor@mail.imc.org Wed Apr 6 15:37:05 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id PAA21074 for ; Wed, 6 Apr 2005 15:37:04 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j36JZWPx053318 for ; Wed, 6 Apr 2005 12:35:32 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j36JZWbw053317 for ietf-usefor-skb; Wed, 6 Apr 2005 12:35:32 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from ietf.org (odin.ietf.org [132.151.1.176]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j36JZUaM053311 for ; Wed, 6 Apr 2005 12:35:31 -0700 (PDT) (envelope-from dinaras@cnri.reston.va.us) Received: from CNRI.Reston.VA.US (localhost [127.0.0.1]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id PAA20760; Wed, 6 Apr 2005 15:35:27 -0400 (EDT) Message-Id: <200504061935.PAA20760@ietf.org> Mime-Version: 1.0 Content-Type: Multipart/Mixed; Boundary="NextPart" To: i-d-announce@ietf.org Cc: ietf-usefor@imc.org From: Internet-Drafts@ietf.org Subject: I-D ACTION:draft-ietf-usefor-usefor-03.txt Date: Wed, 06 Apr 2005 15:35:27 -0400 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: --NextPart A New Internet-Draft is available from the on-line Internet-Drafts directories. This draft is a work item of the Usenet Article Standard Update Working Group of the IETF. Title : News Article Format Author(s) : C. Lindsey, et al. Filename : draft-ietf-usefor-usefor-03.txt Pages : 31 Date : 2005-4-6 This document specifies the syntax of network news (Netnews) articles in the context of the 'Internet Message Format' (RFC 2822) and 'Multipurpose Internet Mail Extensions (MIME)' (RFC 2045). This document supersedes RFC 1036, updating it to reflect current practice and incorporating incremental changes specified in other documents. A URL for this Internet-Draft is: http://www.ietf.org/internet-drafts/draft-ietf-usefor-usefor-03.txt To remove yourself from the I-D Announcement list, send a message to i-d-announce-request@ietf.org with the word unsubscribe in the body of the message. You can also visit https://www1.ietf.org/mailman/listinfo/I-D-announce to change your subscription settings. Internet-Drafts are also available by anonymous FTP. Login with the username "anonymous" and a password of your e-mail address. After logging in, type "cd internet-drafts" and then "get draft-ietf-usefor-usefor-03.txt". A list of Internet-Drafts directories can be found in http://www.ietf.org/shadow.html or ftp://ftp.ietf.org/ietf/1shadow-sites.txt Internet-Drafts can also be obtained by e-mail. Send a message to: mailserv@ietf.org. In the body type: "FILE /internet-drafts/draft-ietf-usefor-usefor-03.txt". NOTE: The mail server at ietf.org can return the document in MIME-encoded form by using the "mpack" utility. To use this feature, insert the command "ENCODING mime" before the "FILE" command. To decode the response(s), you will need "munpack" or a MIME-compliant mail reader. Different MIME-compliant mail readers exhibit different behavior, especially when dealing with "multipart" MIME messages (i.e. documents which have been split up into multiple messages), so check your local documentation on how to manipulate these messages. Below is the data which will enable a MIME compliant mail reader implementation to automatically retrieve the ASCII version of the Internet-Draft. --NextPart Content-Type: Multipart/Alternative; Boundary="OtherAccess" --OtherAccess Content-Type: Message/External-body; access-type="mail-server"; server="mailserv@ietf.org" Content-Type: text/plain Content-ID: <2005-4-6160717.I-D@ietf.org> ENCODING mime FILE /internet-drafts/draft-ietf-usefor-usefor-03.txt --OtherAccess Content-Type: Message/External-body; name="draft-ietf-usefor-usefor-03.txt"; site="ftp.ietf.org"; access-type="anon-ftp"; directory="internet-drafts" Content-Type: text/plain Content-ID: <2005-4-6160717.I-D@ietf.org> --OtherAccess-- --NextPart-- From owner-ietf-usefor@mail.imc.org Thu Apr 7 07:13:23 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id HAA08883 for ; Thu, 7 Apr 2005 07:13:21 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j37BBewT032549 for ; Thu, 7 Apr 2005 04:11:40 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j37BBeGO032548 for ietf-usefor-skb; Thu, 7 Apr 2005 04:11:40 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from rufus.isode.com (rufus.isode.com [62.3.217.251]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j37BBcKY032526 for ; Thu, 7 Apr 2005 04:11:39 -0700 (PDT) (envelope-from alexey.melnikov-usefor@isode.com) Received: from [192.168.0.7] ([62.3.217.253]) by rufus.isode.com via TCP (internal) with ESMTPA; Thu, 7 Apr 2005 12:11:33 +0100 Message-ID: <4255153E.3080500@isode.com> Date: Thu, 07 Apr 2005 12:10:54 +0100 From: Alexey Melnikov User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.7.2) Gecko/20040804 Netscape/7.2 (ax) X-Accept-Language: en-us, en MIME-Version: 1.0 To: ietf-usefor@imc.org Subject: Re: Issues outstanding References: In-Reply-To: Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Content-Transfer-Encoding: 7bit Ken has just published a new revision of USEFOR (03). Some comments on the changes and how they relate to the list of issues below: Charles Lindsey wrote: >Some while back I posted a list of issues (and our Chair added some more). >We have now reached the point where we cannot continue working on our >drafts until these are resolved. > >So here is the list again, with my comments on where we are at on each >one. SO PLEASE CAN WE HAVE INPUT ON THESE, especially on the ones which >still appear to be OPEN? > > >>We are coming to the point where there is little more that can be done on >>the documents we are supposed to be producing without deciding how various >>outstanding issues are to be resolved. >> >> > >1. Complaints-To > > > >>I published a list of 4 options (and invited other options) in >>http://www.imc.org/ietf-usefor/mail-archive/msg00151.html. Only three >>people have expressed any preference amongst them. I think #4 is dead, and >>#2 is the one most people could live with (but on such a small sample, >>that is hardly meaningful). >> >> >>#2 was to do it in Injection-Info rather than in a Complaints-To header, >>and to provide only a mail-complaints-to parameter (which would leave open >>the option to provide a separate url-complaints parameter as a future >>extension). >> > >I think the conclusion we reached on this was to have a >'mail-complaints-to' parameter in the Injection-Info header with an > for its parameter. And we decided not to have any provision >for URLs at this time, though a url-complaints-to parameter could be added >as a future extension if there was a demand for it. > >If that is agreed, then this issue is CLOSED, > This change is done as well as description of different parameters. >except for deciding whether >multiple
s meant you were supposed to reply to ALL of them, or to >ANY ONE of them. Input on this is still needed. > This minor issue is still open. My personal opinion is that if there are multiple email addresses they all should be treated as equal and mailing to ANY of them should suffice. > >2. Path header delimiters > > This is still open. [...] > > >3. Mail-Copies-To and Posted-And Mailed > > > >>Available options appear to be: >> >>1. Include them as in draft-13 >>2a. Defer them to a future document (standards-track) >>2b. Defer them to a future document (experimental) >>3. Drop them entirely >> >> >>Earlier discussions were inconclusive. I gather our Chair prefers #2 (a or >>b), but he has made no definitive pronouncement. >> >> > >I think all that the discussion established was that it was as much effort to >remove them (from USEPRO) as to add them (to USEFOR). It is still not clear >(to me) what the objection to keeping them is, and I see no merit at all in >#2b (since these headers are in moderately common use, and the "experiment" >has, in effect, been done). > >So this issue is still OPEN. > This issue is closed now: the headers will not appear in the USEFOR document. The choice between 2a/2b/3 is up to the WG. > > >4. Terminology for followups. > > > >>1. A followup is a response, and MUST have a References header. A part of >> a multi-part FAQ (or anything similar) is not a followup, but it MAY >> nevertheless have a References header. >> >>2. A followup is a response, or a part of a multi-part FAQ (or anything >> similar). A followup MUST have a References header, and anything else >> MUST NOT have one. >> >>It has been established that there is no technical difference between >>these formulations. It is just a matter of wording, so a simple majority >>for one of the other should settle the matter. >> >>There are alternative definitions in USEPRO, but no corresponding wordings >>for the References header in USEFOR yet, so maybe we should wait until >>there are. >> >> > >This one is still OPEN. There are two alternative texts in USEPRO, but the >matching alternative texts for USEFOR are not in place yet (I hope Ken is >working on them). So I am happy to let this one be for now. There is no >technical issue involved - just a question of how to define things. > I don't believe that anything in the USEFOR should be changed, so this issue concerns the USEPRO document at best. > > >5. Review Injection-Info syntax (this might be related to Complaints-To) > >I invited proposals from anyone who wanted to pursue this. I received none, so >I think this one is CLOSED. > The updated USEFOR draft now includes description of different parameters. If people want an alternative syntax, please speak up now! > >We all agree that RFC 2231 is ugly, but most of it is quite unnecessary in >Netnews. I would be happy for this to be pointed out in USEFOR with suitably >discouraging wording. > >6. Remove filename parameter from the Archive header. > >I think we concluded that the filename-parameter (and perhaps other >parameters) might well be useful in the future, but there was no need to >define them now. Therefore, we should just keep provision for MIME-style >parameters in this header (so software would be required to ignore such >parameters for now), but leave the definition of any actual parameters for >future extensions. > In the latest USEFOR draft the filename parameter was replaced by a generic parameters. > > >7. FWS issue in headers. > >Frank was very keen to introduce *FWS rather than *CFWS or *FWS in various >headers to cope with the rule that folding should not result in empty lines, >or even in lines with empty content. It was established, however, that the >present verbiage covering this issue would still be needed because it was not >possible to solve all such cases syntactically. I argued that there was no >point in changing only those cases where it would work, thereby introducing >differences from RFC 2822. Note that this issue involves no technical change - >just the method of description. > >Frank received no other support, and I propose to do nothing. If Ken wants to >make these changes to USEFOR, then so be it. I regard this one as CLOSED. > I tend to agree. > > >8. Define a Message-ID compatible with NNTP, get rid of NO-WS-CTL. > >We agreed to get rid of NO-WS-CTL (it would have been incompatible with >the new NNTP draft), but our Chair rules that further departures from RFC >2822 were not to be allowed. So I think this is CLOSED. > > > > NO-WS-CTL have been removed. Can people check that the new syntax is Ok? I suspect that some minor issues raised by Frank are yet to be addressed. > So could people who disagree with the ones I have marked CLOSED please >speak up, and otherwise will our Chair please confirm that they are >CLOSED. > >And please may we have discussion of the ones still OPEN, especially the >Path header one. > > > Alexey From owner-ietf-usefor@mail.imc.org Fri Apr 8 02:16:27 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id CAA02250 for ; Fri, 8 Apr 2005 02:16:27 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j386FFAu004709 for ; Thu, 7 Apr 2005 23:15:15 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j386FF1P004708 for ietf-usefor-skb; Thu, 7 Apr 2005 23:15:15 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from ciao.gmane.org (main.gmane.org [80.91.229.2]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j386FDj7004686 for ; Thu, 7 Apr 2005 23:15:14 -0700 (PDT) (envelope-from usenet-format@gmane.org) Received: from list by ciao.gmane.org with local (Exim 4.43) id 1DJmjT-0000VH-4G for ietf-usefor@imc.org; Fri, 08 Apr 2005 08:12:43 +0200 Received: from 212.82.251.127 ([212.82.251.127]) by main.gmane.org with esmtp (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Fri, 08 Apr 2005 08:12:43 +0200 Received: from nobody by 212.82.251.127 with local (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Fri, 08 Apr 2005 08:12:43 +0200 X-Injected-Via-Gmane: http://gmane.org/ To: ietf-usefor@imc.org From: Frank Ellermann Subject: =?UTF-8*de-DE-1996?B?RG9uYXVkYW1wZnNjaGlmZmZhaHJ0c2thcGl0w6Ruc23DvHR6ZQ==?= Date: Fri, 08 Apr 2005 08:10:53 +0200 Organization: Lines: 23 Message-ID: <4256206D.6512@xyzzy.claranet.de> References: <4255153E.3080500@isode.com> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: quoted-printable X-Complaints-To: usenet@sea.gmane.org X-Gmane-NNTP-Posting-Host: 212.82.251.127 X-Mailer: Mozilla 3.0 (OS/2; U) Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Content-Transfer-Encoding: quoted-printable Alexey Melnikov wrote: >> Frank received no other support, and I propose to do >> nothing. If Ken wants to make these changes to USEFOR, then >> so be it. I regard this one as CLOSED. > I tend to agree. One last attempt, the subject of this article is a German fun word: Donaudampfschifffahrtskapit=E4nsm=FCtze. The 3 "fff" are a case of de-DE-1996, I use RfC 2231 and one B64 UTF-8 word. It is too long for RfC2047, therefore it's folded with a FWS. This violates a MUST in Usefor-03, but it's a valid RfC 2822 mail header field. So now what, shoot my UA (in theory, in practice I edited the subject manually) ? Let some "injection agent" fix it ? Or reject it ? What about mail2news gateways ? = Bye, Frank From owner-ietf-usefor@mail.imc.org Fri Apr 8 22:13:39 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id WAA05467 for ; Fri, 8 Apr 2005 22:13:38 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j392CW9M017594 for ; Fri, 8 Apr 2005 19:12:32 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j392CWUt017592 for ietf-usefor-skb; Fri, 8 Apr 2005 19:12:32 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtp809.mail.ukl.yahoo.com (smtp809.mail.ukl.yahoo.com [217.12.12.199]) by above.proper.com (8.12.11/8.12.9) with SMTP id j392CVxE017583 for ; Fri, 8 Apr 2005 19:12:32 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from unknown (HELO host81-144-66-5.midband.mdip.bt.net) (ietf-usefor@imc.org@81.144.66.5 with poptime) by smtp809.mail.ukl.yahoo.com with SMTP; 9 Apr 2005 02:12:25 -0000 Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j392CCE25526 for ietf-usefor@imc.org; Sat, 9 Apr 2005 03:12:12 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20637 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: =?UTF-8*de-DE-?B?RG9uYXVkYW1wZnNjaGlmZmZhaHJ0c2thcGl0w6Ruc23DvHR6ZQ==?= Content-Type: text/plain; charset=iso-8859-1 Message-ID: Content-Transfer-Encoding: 8bit X-Newsreader: NN version 6.5.2 (NOV) References: <4255153E.3080500@isode.com> <4256206D.6512@xyzzy.claranet.de> Mime-Version: 1.0 Date: Fri, 8 Apr 2005 20:11:05 GMT Lines: 55 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Content-Transfer-Encoding: 8bit In <4256206D.6512@xyzzy.claranet.de> Frank Ellermann writes: >Alexey Melnikov wrote: >>> Frank received no other support, and I propose to do >>> nothing. If Ken wants to make these changes to USEFOR, then >>> so be it. I regard this one as CLOSED. >> I tend to agree. >One last attempt, the subject of this article is a German fun >word: Donaudampfschifffahrtskapitänsmütze. The 3 "fff" are a >case of de-DE-1996, I use RfC 2231 and one B64 UTF-8 word. >It is too long for RfC2047, therefore it's folded with a FWS. >This violates a MUST in Usefor-03, but it's a valid RfC 2822 >mail header field. It is indeed, but since it is in breach of RFC 2047, a conforming reading agent SHOULD render it exactly as received ('=?...?=' and all). However, some kind agents _might_ try to decode it for you. The proper way to create that header is as: Subject: =?UTF-8*de-DE-?B?RG9uYXVkYW1wZnNjaGlmZmZh?= =?UTF-8*de-DE-?B?aHJ0c2thcGl0w6Ruc23DvHR6ZQ==?= That works because, if you have two encoded words with nothing but FWS between them, then the decoder is required to ignore the FWS. And you have to be careful where you make the split, because each encoded-word must encode an integral number of octets (so after a multiple of 4 for base64, and watch out for '=xy' in Q-P). And on top of that you must not split any multi-octet character between the encoded words (which could be tricky in UTF-8, where the number of octets per character is variable). But any decent MUA ought to be able to do it. I tried it in Opera, and it produced: Subject: =?utf-8?Q?Donaudampfschifffahrtskapit?= =?utf-8?Q?=C3=A4nsm=C3=BCtze?= OK, it used Q-P rather than Base 64, but it was quite happy to display the original word without any spaces in the middle when read back. BTW, your message achieved a Spammassassin score of 3 :-( . -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 From owner-ietf-usefor@mail.imc.org Fri Apr 8 22:14:01 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id WAA05736 for ; Fri, 8 Apr 2005 22:14:00 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j392CYMg017602 for ; Fri, 8 Apr 2005 19:12:34 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j392CY3O017601 for ietf-usefor-skb; Fri, 8 Apr 2005 19:12:34 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtp809.mail.ukl.yahoo.com (smtp809.mail.ukl.yahoo.com [217.12.12.199]) by above.proper.com (8.12.11/8.12.9) with SMTP id j392CXwS017584 for ; Fri, 8 Apr 2005 19:12:33 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from unknown (HELO host81-144-66-5.midband.mdip.bt.net) (ietf-usefor@imc.org@81.144.66.5 with poptime) by smtp809.mail.ukl.yahoo.com with SMTP; 9 Apr 2005 02:12:26 -0000 Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j392CDr25532 for ietf-usefor@imc.org; Sat, 9 Apr 2005 03:12:13 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20638 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: Issues outstanding Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: <4255153E.3080500@isode.com> Date: Fri, 8 Apr 2005 21:32:34 GMT Lines: 188 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In <4255153E.3080500@isode.com> Alexey Melnikov writes: >Ken has just published a new revision of USEFOR (03). Some comments on >the changes and how they relate to the list of issues below: There is still an awful lot of stuff that is meant to be covered in USEFOR, but still isn't. But that is for another thread. >Charles Lindsey wrote: >>>We are coming to the point where there is little more that can be done on >>>the documents we are supposed to be producing without deciding how various >>>outstanding issues are to be resolved. >> >>1. Complaints-To >> >>I think the conclusion we reached on this was to have a >>'mail-complaints-to' parameter in the Injection-Info header with an >> for its parameter. And we decided not to have any provision >>for URLs at this time, though a url-complaints-to parameter could be added >>as a future extension if there was a demand for it. >> >>If that is agreed, then this issue is CLOSED, >> >This change is done as well as description of different parameters. Yes, but not quite as agreed, because the token he has given for the parameter is 'complaints-to', rather than 'mail-complaints-to' as we agreed. That will not satisfy the people who wanted to leave room for a 'url-complaints-to' parameter as a future extension. >>except for deciding whether >>multiple
s meant you were supposed to reply to ALL of them, or to >>ANY ONE of them. Input on this is still needed. >> >This minor issue is still open. As currently written, the of that parameter is an , which by default means "send to them all" (cf. the Reply-To header). >My personal opinion is that if there are multiple email addresses they >all should be treated as equal and mailing to ANY of them should suffice. That could be achieved by adding some wording to say so. I am easy either way, so we need to hear more opinions. >>2. Path header delimiters >> >> >This is still open. >[...] Shame! I though we had more-or less agreed on that one. OK, I shall start another thread to discuss that. >>3. Mail-Copies-To and Posted-And Mailed >> >>>1. Include them as in draft-13 >>>2a. Defer them to a future document (standards-track) >>>2b. Defer them to a future document (experimental) >>>3. Drop them entirely >> >This issue is closed now: the headers will not appear in the USEFOR >document. The choice between 2a/2b/3 is up to the WG. Actually, I have now come to the conclusion that this problem is far worse with mailing lists than it is on Usenet, so it might be better to fix it there, maybe based on Mail-Followups-To with some extra features to incorporate News. Indeed, there was some discussion on the ietf-822 list about Mail-Followups-To as a possible solution, with many in favour but two diehards implacably opposed (for completely opposite reasons). Ours is not the only list that suffers from long discussions with no decision at the end :-( . (Oddly, I noticed that John Stanley was in favour of Posted-And-Mailed - he is quite right, but if Mail- Copies-To has to wait, then Posted-And-Mailed must wait too.) I shall now remove all mention of both these headers from USEPRO. >>4. Terminology for followups. >> >>This one is still OPEN. There are two alternative texts in USEPRO, but the >>matching alternative texts for USEFOR are not in place yet (I hope Ken is >>working on them). So I am happy to let this one be for now. There is no >>technical issue involved - just a question of how to define things. >> >I don't believe that anything in the USEFOR should be changed, so this >issue concerns the USEPRO document at best. Could you please explain your reasoning here? ISTM that any header described in USEFOR needs the following information, as appropriate: 1. A brief statement of what the header is supposed to achieve. 2. Its syntax. 3. Its semantics (i.e. what information is conveyed by its various syntactic parts). 4. Any restrictions or requirements on when it is to be used (e.g. it is "mandatory", or it MUST/SHOULD [NOT] be present if such and such other circumstances pertain). Following #4, there has always been a statement like the following associated with this header: A followup MUST have a References-header, and an article that is not a followup MUST NOT have a References-header. It is particularly important to say that here, because it is a change from RFC 2822, where the word used is only SHOULD. Now the precise wording of that varies according to exactly how the term "followup" is defined (and that is what this issue is all about). But both sides to this argument are agreed that that "MUST" needs to be said, and this is (AFAICS) the only place where it could be said. Yes, USEPRO describes in detail how to construct this header in the particular case of responses/replies to earlier articles, but there are other applications for it also, for example multipart FAQs and message/partial (and again it is common ground that these are legitimate applications). >>5. Review Injection-Info syntax (this might be related to Complaints-To) >> >The updated USEFOR draft now includes description of different parameters. >If people want an alternative syntax, please speak up now! Well nobody has spoken on this for a long time. I think the important thing is that Russ declared that he could live with this syntax, and I think his opinion is important. But I think we also need some deprecatory remarks to discourage the wilder (and unnecessary) extremes of RFC 2231 (and please can we use that lovely word "gibbous" in them :-) ). It is nice to see all the other Injection-Info parameters fully described. I have a few niggles about the precise details, and there are issues regarding the way they have been introduced syntactically, but that too is for another thread. >>6. Remove filename parameter from the Archive header. >> >In the latest USEFOR draft the filename parameter was replaced by a >generic parameters. Fine! >>7. FWS issue in headers. >> >>Frank was very keen to introduce *FWS rather than *CFWS or *FWS in various >>headers to cope with the rule that folding should not result in empty lines, >>or even in lines with empty content. It was established, however, that the >>present verbiage covering this issue would still be needed because it was not >>possible to solve all such cases syntactically. I argued that there was no >>point in changing only those cases where it would work, thereby introducing >>differences from RFC 2822. Note that this issue involves no technical change - >>just the method of description. >> >>Frank received no other support, and I propose to do nothing. If Ken wants to >>make these changes to USEFOR, then so be it. I regard this one as CLOSED. >> >I tend to agree. >>8. Define a Message-ID compatible with NNTP, get rid of NO-WS-CTL. >> >>We agreed to get rid of NO-WS-CTL (it would have been incompatible with >>the new NNTP draft), but our Chair rules that further departures from RFC >>2822 were not to be allowed. So I think this is CLOSED. >NO-WS-CTL have been removed. Can people check that the new syntax is Ok? >I suspect that some minor issues raised by Frank are yet to be addressed. Yes, there is a placeholder "[[Adjacent dots should not be allowed]]" for Frank's problem, but we still need some syntax to plug in there. We also need some wording drawing attention to this extra departure from RFC 2822, and referring to [NNTP] which necessitated it. -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 From owner-ietf-usefor@mail.imc.org Sat Apr 9 00:26:01 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id AAA09317 for ; Sat, 9 Apr 2005 00:26:01 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j394Oqd8030346 for ; Fri, 8 Apr 2005 21:24:52 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j394Oq04030345 for ietf-usefor-skb; Fri, 8 Apr 2005 21:24:52 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from ciao.gmane.org (main.gmane.org [80.91.229.2]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j394OnCL030339 for ; Fri, 8 Apr 2005 21:24:50 -0700 (PDT) (envelope-from usenet-format@gmane.org) Received: from list by ciao.gmane.org with local (Exim 4.43) id 1DK7Tz-0002KG-W0 for ietf-usefor@imc.org; Sat, 09 Apr 2005 06:22:08 +0200 Received: from du-001-228.access.de.clara.net ([212.82.227.228]) by main.gmane.org with esmtp (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Sat, 09 Apr 2005 06:22:07 +0200 Received: from nobody by du-001-228.access.de.clara.net with local (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Sat, 09 Apr 2005 06:22:07 +0200 X-Injected-Via-Gmane: http://gmane.org/ To: ietf-usefor@imc.org From: Frank Ellermann Subject: =?UTF-8*de-DE-1996?B?RG9uYXVkYW1wZnNjaGlmZmZhaHJ0c2thcGl0w6Ruc23DvHR6ZQ==?= Date: Sat, 09 Apr 2005 06:14:38 +0200 Organization: Lines: 25 Message-ID: <425756AE.2232@xyzzy.claranet.de> References: <4255153E.3080500@isode.com> <4256206D.6512@xyzzy.claranet.de> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Complaints-To: usenet@sea.gmane.org X-Gmane-NNTP-Posting-Host: du-001-228.access.de.clara.net X-Mailer: Mozilla 3.0 (OS/2; U) Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Content-Transfer-Encoding: 7bit Charles Lindsey wrote: >> it's a valid RfC 2822 mail header field. > It is indeed, but since it is in breach of RFC 2047 It shouldn't, I've sent it in two lines, 2nd line 76 characters: Subject: =?UTF-8*de-DE-1996?B?RG9uYXVkYW1wZnNjaGlmZmZhaHJ0c2thcGl0w6Ruc23DvHR6ZQ==?= ....5...10....5...20....5...30....5...40....5...50....5...60....5...70....5. http://article.gmane.org/gmane.ietf.usenet.format:28154:raw says that "something" between me and GmaNe "fixed" this, but generally my UA is innocent if it comes to attempts of being smart. > Subject: =?UTF-8*de-DE-?B?RG9uYXVkYW1wZnNjaGlmZmZh?= > =?UTF-8*de-DE-?B?aHJ0c2thcGl0w6Ruc23DvHR6ZQ==?= Sure, you and me know this, but I'm talking about existing MUAs, mail2news gateways, injection agents, servers, and newsreaders. > BTW, your message achieved a Spammassassin score of 3 :-( . 1st UTF-8, 2nd B64, what was the 3rd ? Bye, Frank From owner-ietf-usefor@mail.imc.org Sat Apr 9 00:57:39 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id AAA15190 for ; Sat, 9 Apr 2005 00:57:39 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j394uqGc031776 for ; Fri, 8 Apr 2005 21:56:52 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j394uqKx031775 for ietf-usefor-skb; Fri, 8 Apr 2005 21:56:52 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from ciao.gmane.org (main.gmane.org [80.91.229.2]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j394upSL031769 for ; Fri, 8 Apr 2005 21:56:51 -0700 (PDT) (envelope-from usenet-format@gmane.org) Received: from list by ciao.gmane.org with local (Exim 4.43) id 1DK7yz-0003nh-QC for ietf-usefor@imc.org; Sat, 09 Apr 2005 06:54:09 +0200 Received: from du-001-228.access.de.clara.net ([212.82.227.228]) by main.gmane.org with esmtp (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Sat, 09 Apr 2005 06:54:09 +0200 Received: from nobody by du-001-228.access.de.clara.net with local (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Sat, 09 Apr 2005 06:54:09 +0200 X-Injected-Via-Gmane: http://gmane.org/ To: ietf-usefor@imc.org From: Frank Ellermann Subject: Re: =?UTF-8*de-DE-1996?B?RG9uYXVkYW1wZnNjaGlmZmZhaHJ0c2thcGl0w6Ruc23DvHR6ZQ==?= Date: Sat, 09 Apr 2005 06:55:19 +0200 Organization: Lines: 9 Message-ID: <42576037.190F@xyzzy.claranet.de> References: <4255153E.3080500@isode.com> <4256206D.6512@xyzzy.claranet.de> <425756AE.2232@xyzzy.claranet.de> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Complaints-To: usenet@sea.gmane.org X-Gmane-NNTP-Posting-Host: du-001-228.access.de.clara.net X-Mailer: Mozilla 3.0 (OS/2; U) Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Content-Transfer-Encoding: 7bit I've just tested it with two news servers, and both rejected a "Subject:" SP CRLF SP "stuff" CRLF So "fixing" the FWS issue by removing the separate MUST about non-empty header body lines Usefor-03 would be a very bad idea. Maybe I should a post an I-D "Subject: Re: considered effective at least" mentioning this case. From owner-ietf-usefor@mail.imc.org Sat Apr 9 16:11:56 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id QAA04424 for ; Sat, 9 Apr 2005 16:11:55 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j39KAE6o022796 for ; Sat, 9 Apr 2005 13:10:14 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j39KAEbO022795 for ietf-usefor-skb; Sat, 9 Apr 2005 13:10:14 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtp801.mail.ukl.yahoo.com (smtp801.mail.ukl.yahoo.com [217.12.12.138]) by above.proper.com (8.12.11/8.12.9) with SMTP id j39KADWn022779 for ; Sat, 9 Apr 2005 13:10:13 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from unknown (HELO host81-144-73-19.midband.mdip.bt.net) (ietf-usefor@imc.org@81.144.73.19 with poptime) by smtp801.mail.ukl.yahoo.com with SMTP; 9 Apr 2005 20:10:06 -0000 Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j39K9o229793 for ietf-usefor@imc.org; Sat, 9 Apr 2005 21:09:50 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20642 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Differences between RFC 2822 and Usefor Message-ID: X-Newsreader: NN version 6.5.2 (NOV) Date: Sat, 9 Apr 2005 20:00:42 GMT Lines: 318 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: I published a list of differences from RFC 2822 as they stood in the old draft-13 last July (and it has been on our web site ever since). Most of those differences were agreed years ago when this WG was first formed and have been in the drafts ever since. I don't recall any disagreement with the list when I published it. There have been a few things changed since then (like the MIME-style parameters have gone), and also things are described somewhat differently in USEFOR, so I have updated it. The new texts follows (and it will shortly be put on the website), and after that there are the diffs from the previous one. Note that many of these agreed changes have not made it into the new USEFOR yet, and so I have marked the missing ones with "**". Differences between RFC 2822 and Usefor. ---------------------------------------- 08 April 2005 NOTE: Some items have been removed from the previous version, hence some discontinuities in the listing below. A '**' before an item indicates that the restriction in question has not yet been incorporated into the USEFOR document. 1. Differences enforced syntactically. -------------------------------------- 1.1 There is a REQUIRED SP (not even WSP) after the ':' in each header (even if the header has no content). 1.2 An MUST have at least one character. Thus the Subject (for example) of an article cannot be empty. 1.3**In the References-header, there is an obligatory CFWS (currently restricted to FWS by 2.1.1 below) in between each msg-id. NOTE: all the above were in RFC 1036, and the first is also required by NNTP. 1.5**Body lines are restricted to 998 characters plus CRLF (as in RFC 2822). However, all agents SHOULD, and relaying agents MUST, process lines of arbitrary length. 1.8 All the "obsolete" syntax in RFC 2822 is NOT REQUIRED (though it MAY be recognized). There are two small exceptions, listed below. 1.8.1 In the Date-header, the obsolete "UT" and "GMT" forms of zone MUST be recognized (because of their current widespread use), but MUST NOT be generated. 1.8.2**The (mis-named) "obsolete" syntax for phrase from RFC 2822, which allows for 'John D. Smith' to remain unquoted, is retained (but renamed as extended-phrase). NOTE: NNTP imposes similar restrictions. 1.11 Message-ID 1.11.1 No CFWS allowed (only FWS, which in practice means only non-folded WSP). 1.11.2 No quoted-pairs in msg-id, except for '\\' and '\"' in id-left '\[', '\]' and '\\' in id-right 1.11.3 No quoted-strings in id-left unless they contain one of the "specials". 1.11.4 No '>' anywhere within a msg-id, even within a quoted-string (to conform with RFC 1036). 1.11.5 All these restrictions on msg-id apply also to the References- and Supersedes-headers. 1.11.6 No control characters in msg-id. 2. Differences enforced by verbiage. ------------------------------------ 2.1 The following MUST be accepted, but SHOULD NOT be generated (yet): 2.1.1**Comments, except after a mailbox (where there is a now-deprecated convention for indicating the mailbox owner), or at the end of a date-time (which conventionally indicates the timezone). They are, however, freely allowed in headers that are newly defined in Usefor. 2.1.2**Extended-phrases (see 1.8.2 above). 2.2 The content of the first line of a header MUST NOT consist of WSP only (though such SHOULD be accepted). Observe that continuation lines of headers also MUST NOT consist of WSP only, as in RFC 2822. 2.3 Headers with empty content are deprecated (but if present that SP after the ':" is still required). 2.4 All agents MUST support header lines up to 998 octets, but there is no RECOMENDED limit of 78 characters as in RFC 2822. There is mention of a purely advisory limit of 79 (with a reference to USEAGE). 2.5 Relaying agents MUST NOT refold headers in transit. 2.6 There must not be more than one header with a given header-name, except where explicitly sanctioned by the appropriate standard. In particular, there MUST NOT be more than one Keywords-header. 2.7 The length of a msg-id MUST NOT exceed 250 octets. 2.8**The body of an article SHOULD NOT be empty. 2.9 (was 1.6) RFC 2047 and RFC 2231 are fully integrated into the Netnews. 2.10 (was 1.7) All the Content-* MIME headers are considered to be incorporated into Netnews and MUST be accepted in articles at, least to the extent required by RFC 2049. 2.11 A References header MUST be provided for followups (as opposed to SHOULD be provided for replies in RFC 2822). 3. Rules specific to Netnews headers. ------------------------------------- 3.2 Comments (but not FWS) are forbidden in the Newsgroups-, Distribution- Path- and Followup-To-headers (also see 1.11.1 above for Message-ID). 3.3**WSP and folding in Newsgroup- and Followup-To-headers MUST be accepted, but SHOULD NOT be generated (yet). NOTE: The effect of all these differences still preserves the property that the articles that Usefor permits to be generated form a proper subset of the articles that are required to be acceptable to RFC 2822. *** rfc2822-diffs.old Thu Jul 1 23:24:57 2004 --- rfc2822-diffs.txt Fri Apr 8 22:05:05 2005 *************** *** 3,6 **** ! 01 July 2004 1. Differences enforced syntactically. --- 3,12 ---- ! 08 April 2005 + NOTE: Some items have been removed from the previous version, hence some + discontinuities in the listing below. + + A '**' before an item indicates that the restriction in question has not yet + been incorporated into the USEFOR document. + 1. Differences enforced syntactically. *************** *** 11,13 **** ! 1.2 An 'unstructured' MUST have at least one character. Thus the Subject (for example) of an article cannot be empty. --- 17,19 ---- ! 1.2 An MUST have at least one character. Thus the Subject (for example) of an article cannot be empty. *************** *** 14,16 **** ! 1.3 In the References-header, there is an obligatory CFWS (currently restricted to FWS by 2.1.1 below) in between each msg-id. --- 20,22 ---- ! 1.3**In the References-header, there is an obligatory CFWS (currently restricted to FWS by 2.1.1 below) in between each msg-id. *************** *** 20,26 **** ! 1.4 The allowed characters in a header-name are restricted to ALPHA, ! DIGIT and embedded '-'. However, agents SHOULD accept all printables ! except SP and ':'. ! ! 1.5 Body lines are restricted to 998 characters plus CRLF (as in RFC 2822). However, all agents SHOULD, and relaying agents MUST, process --- 26,28 ---- ! 1.5**Body lines are restricted to 998 characters plus CRLF (as in RFC 2822). However, all agents SHOULD, and relaying agents MUST, process *************** *** 28,38 **** - 1.6 RFC 2047 and RFC 2231 are fully integrated into the syntax. Thus - encoded-words are explicitly included within 'unstructured', - 'ccontent' and 'phrase'. Moreover, it is the RFC 2231 version of - encoded-word that is used. - - 1.7 All the Content-* MIME headers are considered to be incorporated into - the syntax (i.e. they are to be accepted in articles, though not all - of them are required to have their semantic intentions implemented). - 1.8 All the "obsolete" syntax in RFC 2822 is NOT REQUIRED (though it MAY --- 30,31 ---- *************** *** 44,46 **** ! 1.8.2 The (mis-named) "obsolete" syntax for phrase from RFC 2822, which allows for 'John D. Smith' to remain unquoted, is retained (but --- 37,39 ---- ! 1.8.2**The (mis-named) "obsolete" syntax for phrase from RFC 2822, which allows for 'John D. Smith' to remain unquoted, is retained (but *************** *** 68,70 **** --- 61,65 ---- + 1.11.6 No control characters in msg-id. + 2. Differences enforced by verbiage. *************** *** 74,83 **** ! 2.1.1 Comments, except after a mailbox (where there is a now-deprecated convention for indicating the mailbox owner), or at the end of a ! date-time (which conventionally indicates the timezone). ! BUG: they ought to be freely allowed in headers that are newly ! defined in Usefor (and indeed the text positively encourages them ! in some of those places). ! 2.1.2 Extended-phrases (see 1.8.2 above). --- 69,76 ---- ! 2.1.1**Comments, except after a mailbox (where there is a now-deprecated convention for indicating the mailbox owner), or at the end of a ! date-time (which conventionally indicates the timezone). They are, ! however, freely allowed in headers that are newly defined in Usefor. ! 2.1.2**Extended-phrases (see 1.8.2 above). *************** *** 88,91 **** 2.3 Headers with empty content are deprecated (but if present that SP ! after the ':" is still required). Injecting agents SHOULD delete such ! headers, but other agents MUST propagate them. --- 81,83 ---- 2.3 Headers with empty content are deprecated (but if present that SP ! after the ':" is still required). *************** *** 103,107 **** ! 2.8 The body of an article SHOULD NOT be empty. 3. Rules specific to Netnews headers. --- 95,107 ---- ! 2.8**The body of an article SHOULD NOT be empty. + 2.9 (was 1.6) RFC 2047 and RFC 2231 are fully integrated into the Netnews. + 2.10 (was 1.7) All the Content-* MIME headers are considered to be + incorporated into Netnews and MUST be accepted in articles at, least + to the extent required by RFC 2049. + + 2.11 A References header MUST be provided for followups (as opposed to SHOULD + be provided for replies in RFC 2822). + 3. Rules specific to Netnews headers. *************** *** 109,118 **** - 3.1 All structured headers have MIME-style extension-parameters, with - x-attributes or to be defined in future standards. Some have explicit - parameters defined in this standard. However, this does not apply to - headers which are taken from RFC 2822 or other mail standards, nor to - the Mail-Copies-To, Complaints-to and Supersedes-header defined in - this standard. Nevertheless, such parameters SHOULD be recognized - (and ignored) in all headers. - 3.2 Comments (but not FWS) are forbidden in the Newsgroups-, --- 109,110 ---- *************** *** 121,128 **** ! 3.3 The following MUST be accepted, but SHOULD NOT be generated (yet): - 3.3.1 MIME-style parameters in headers defined prior to this standard. - - 3.3.2 WSP and folding in Newsgroup- and Followup-To-headers. - NOTE: The effect of all these differences still preserves the property --- 113,117 ---- ! 3.3**WSP and folding in Newsgroup- and Followup-To-headers MUST be ! accepted, but SHOULD NOT be generated (yet). NOTE: The effect of all these differences still preserves the property -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 From owner-ietf-usefor@mail.imc.org Sat Apr 9 16:11:57 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id QAA04441 for ; Sat, 9 Apr 2005 16:11:57 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j39KACH9022787 for ; Sat, 9 Apr 2005 13:10:12 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j39KACFG022786 for ietf-usefor-skb; Sat, 9 Apr 2005 13:10:12 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtp801.mail.ukl.yahoo.com (smtp801.mail.ukl.yahoo.com [217.12.12.138]) by above.proper.com (8.12.11/8.12.9) with SMTP id j39KABRT022778 for ; Sat, 9 Apr 2005 13:10:12 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from unknown (HELO host81-144-73-19.midband.mdip.bt.net) (ietf-usefor@imc.org@81.144.73.19 with poptime) by smtp801.mail.ukl.yahoo.com with SMTP; 9 Apr 2005 20:10:05 -0000 Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j39K9na29787 for ietf-usefor@imc.org; Sat, 9 Apr 2005 21:09:49 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20641 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: Babble from Frank Ellermann Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: <4255153E.3080500@isode.com> <4256206D.6512@xyzzy.claranet.de> <425756AE.2232@xyzzy.claranet.de> Date: Sat, 9 Apr 2005 19:33:52 GMT Lines: 48 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In <425756AE.2232@xyzzy.claranet.de> Frank Ellermann writes: >Charles Lindsey wrote: >>> it's a valid RfC 2822 mail header field. >> It is indeed, but since it is in breach of RFC 2047 >It shouldn't, I've sent it in two lines, 2nd line 76 characters: >Subject: > =?UTF-8*de-DE-1996?B?RG9uYXVkYW1wZnNjaGlmZmZhaHJ0c2thcGl0w6Ruc23DvHR6ZQ==?= >....5...10....5...20....5...30....5...40....5...50....5...60....5...70....5. Ah! It was all on one line when I received it, so I didn't understand what your problem was. >http://article.gmane.org/gmane.ietf.usenet.format:28154:raw >says that "something" between me and GmaNe "fixed" this, but >generally my UA is innocent if it comes to attempts of being >smart. >> Subject: =?UTF-8*de-DE-?B?RG9uYXVkYW1wZnNjaGlmZmZh?= >> =?UTF-8*de-DE-?B?aHJ0c2thcGl0w6Ruc23DvHR6ZQ==?= >Sure, you and me know this, but I'm talking about existing MUAs, >mail2news gateways, injection agents, servers, and newsreaders. Well if the sending agent cannot figure how to do it that way, then it will just have to leave it on one long line. Then it becomes the reading agent's problem :-) . >> BTW, your message achieved a Spammassassin score of 3 :-( . >1st UTF-8, 2nd B64, what was the 3rd ? SpamAssassin hits were FORGED_RCVD_HELO RCVD_BY_IP RCVD_NUMERIC_HELO SUBJECT_EXCESS_BASE64 SUBJECT_NOVOWEL -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 From owner-ietf-usefor@mail.imc.org Sat Apr 9 22:13:39 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id WAA23947 for ; Sat, 9 Apr 2005 22:13:39 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3A2CTAx040128 for ; Sat, 9 Apr 2005 19:12:29 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3A2CTGt040127 for ietf-usefor-skb; Sat, 9 Apr 2005 19:12:29 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtp814.mail.ukl.yahoo.com (smtp814.mail.ukl.yahoo.com [217.12.12.204]) by above.proper.com (8.12.11/8.12.9) with SMTP id j3A2CR8F040115 for ; Sat, 9 Apr 2005 19:12:28 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from unknown (HELO host81-144-72-58.midband.mdip.bt.net) (ietf-usefor@imc.org@81.144.72.58 with poptime) by smtp814.mail.ukl.yahoo.com with SMTP; 10 Apr 2005 02:12:21 -0000 Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3A2CA802775 for ietf-usefor@imc.org; Sun, 10 Apr 2005 03:12:10 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20643 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Path header delimiters Message-ID: X-Newsreader: NN version 6.5.2 (NOV) Date: Sat, 9 Apr 2005 21:06:31 GMT Lines: 138 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Our Chair has ruled that this matter it still open, so here is a summary of where I think we are at. Many years ago, when this WG was first formed, we were concerned that malefactors regularly preloaded the Path header with all sorts of sites (usually culled for some other article), in an attempt to disguise the point at which it had been injected. We therefore proposed to modify the Path header so that relaying agents would be able to make the following assertions when they add a new path-identity to the Path-header: #1 I am the injecting site. #2 I have checked the identity of the previous site, and I believe the path-identity inserted by that site to be correct. #3 I have checked the identity of the previous site, and I do not believe the path-identity claimed by that site; here is what I believe to be the true identity of that site. #4 I have made no checks on the identity of the previous site. The Path header is to News as the sequence of Received headers is to Email, and Received headers currently convey much the same information, and very useful they are when trying to trace the origin of some spam or other malefaction. I think we are still agreed that such a feature in the Path header is desirable - the point at issue is how to convey this information in a backwards compatible manner. The original suggestion was to introduce the delimiters '%', '/', and '?' in addition to the usual '!', it being apparently the case that RFC 1036 already allowed all these characters, and many more, to be used as delimiters, and it appeared that existing implementations did indeed accept them. However, RFC 1036 is somewhat vague on exactly what the allowed set of characters was. A few months ago, Bruce Lilly pointed out that maybe some implementations were not so accomodating, and pointed to the implementation of BNews (written by the author of RFC 1036) which actually accepted a rather small set. Following from that, various alternative schemes were proposed, as follows. In the following running example injector.com always uses #1 new-site.com always uses #2 or #3 good-site.com always uses #2 or #3 old-site always uses #4 dodgy.com was a bogus identity actually inserted by mallet.com A. Current draft: ----------------- Uses '%' for #1, '/' for #2, '?' for #3 amd '!' for #4 Path: good-site.com/mallet.com?dodgy.com!old-site.com! new-site.com/injector.com%not-for-mail B. Henry's proposal : ------------------------------------------------------------------ Uses '@' for #1, ',' for #2, ' ' for #3 amd '!' for #4, since it is clear from RFC 1036 that all of those are intended to be usable as delimiters. Path: good-site.com,mallet.com dodgy.com!old-site.com! new-site.com,injector.com@not-for-mail Observe that the ' ' delimiter turns up rather conveniently as a separator between the correct and bogus identities of mallet.com. One would need to discuss whether FWS as well as SP should delimit this case. C. The Diablo scheme -------------------- I still have not been able to find documentation on this, but from observed instances it appears to work as follows: Path: good-site.com!mallet.com.MISMATCH!dodgy.com!old-site.com! new-site.com!injector.com.POSTED!not-for-mail I see 2 problems with this one: 1: Any site which peers with injector.com (e.g. new-site.com) would normally scan the received Path for occurrences of "injector.com", and would send the article back to injector.com if it was not found (which, of course, it isn't here because it recorded itself as "injector.com.POSTED"). 2: It provides no distinction betwen cases #2 and #4, which rather defeats the object of the whole exercise. D. Another possible scheme -------------------------- If you want to avoid all delimiters other than '!', and to overcome the problems with the Diablo scheme, then here is one which relies on special keywords "M", "MISMATCH" and "POSTED" in places where the current syntax would expect a path-identity. Path: good-site.com!M!mallet.com!MISMATCH!dodgy.com!old-site.com! new-site.com!M!injector.com!POSTED!not-for-mail It makes the Path a little longer, but not unacceptably so, and assumes that those keywords will never represent real sites. E. A variant on scheme D ------------------------ Instead of the keyword 'M' to indicate case #2, just place two '#' delimiters in succession, giving: Path: good-site.com!!mallet.com!MISMATCH!dodgy.com!old-site.com! new-site.com!!injector.com!POSTED!not-for-mail This avoids any delimiter other than '!', but it assumes that two delimiters in succession will not cause any trouble. But it is clear that RFC 1036 permits such usage, because it uses two delimiters in succession in some of its own examples. My own view is that we should adopt scheme E. It is much easier to recognize what is meant, rather than having to remember exactly what '%', '/', and '?' mean. In the discussion since I posted that last list of outstanding issues, Frank Ellerman agreed that we should go with this scheme (having carefully examined those examples in RFC 1036), and nobody suggested anything different. So that scheme is the front runner at the moment. I could easily propose modifications to the existing texts to incorporate that scheme, but I don't want to embark on that unless we are agreed it is the way to go. So does anyone else want to comment, and can I assume that, if there are no objections, we go with scheme E? -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 From owner-ietf-usefor@mail.imc.org Sun Apr 10 11:41:04 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id LAA27673 for ; Sun, 10 Apr 2005 11:41:04 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3AFe3J2023250 for ; Sun, 10 Apr 2005 08:40:03 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3AFe3ws023249 for ietf-usefor-skb; Sun, 10 Apr 2005 08:40:03 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from ciao.gmane.org (main.gmane.org [80.91.229.2]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3AFe0Cj023235 for ; Sun, 10 Apr 2005 08:40:01 -0700 (PDT) (envelope-from usenet-format@gmane.org) Received: from list by ciao.gmane.org with local (Exim 4.43) id 1DKeUi-0000pl-70 for ietf-usefor@imc.org; Sun, 10 Apr 2005 17:37:04 +0200 Received: from 212.82.251.245 ([212.82.251.245]) by main.gmane.org with esmtp (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Sun, 10 Apr 2005 17:37:04 +0200 Received: from nobody by 212.82.251.245 with local (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Sun, 10 Apr 2005 17:37:04 +0200 X-Injected-Via-Gmane: http://gmane.org/ To: ietf-usefor@imc.org From: Frank Ellermann Subject: Broken Mesage-ID syntax (was: Issues outstanding) Date: Sun, 10 Apr 2005 17:38:33 +0200 Organization: Lines: 53 Message-ID: <42594879.3B92@xyzzy.claranet.de> References: <4255153E.3080500@isode.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Complaints-To: usenet@sea.gmane.org X-Gmane-NNTP-Posting-Host: 212.82.251.245 X-Mailer: Mozilla 3.0 (OS/2; U) Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Content-Transfer-Encoding: 7bit Alexey Melnikov wrote: > NO-WS-CTL have been removed. Can people check that the new > syntax is Ok? It is not. The "pseudo-consensus" between Charles and me was to replace msg-id-core by msg-id everywhere, and to fix the dot-atom-text issue with leading / trailing dots plus dot-dot. Two (Charles + me) is a bad number, could somebody like you please volunteer as third for a proper "rough consensus" about the last open msg-id point, a better name for the "LHS @ RHS" productions ? I've interpreted an unrelated article from Bruce elsewhere as support for "unique @ domain" productions, but that's obviously dubious, > I suspect that some minor issues raised by Frank are yet to > be addressed. The msg-id and dot stuff isn't minor, the former is important, the latter is erroneous. Here's the latest ABNF for a msg-id: | msg-id = "<" unique "@" mdomain ">" | unique = dot-atom-text / ( DQUOTE unique-quote DQUOTE ) | unique-quote = ( "." [unique-part] ) / | ( [unique-part] "." ) / | ( [unique-part] unique-literal [unique-part] ) | unique-part = 1*( atext / "." / unique-literal ) | unique-literal = "(" / ")" / "," / ; all specials, minus ">", | "[" / "]" / "@" / ; minus DQUOTE, minus "\", | ":" / ";" / "<" / ; minus single ".", plus: | ".." / "\\" / ( "\" DQUOTE ) | mdomain = dot-atom-text / ("[" address-literal "]") | address-literal = 1*( %d33-61 / ; printable ASCII minus | %d63-90 / ; ">", "[", "\", "]" | %d94-126 / ; plus "\[", "\\, "\]" | "\[" / "\\" / "\]" ) Some weeks later the "no-pseudo-consensus" arrived at this: msg-id = "<" id-local "@" id-domain ">" msg-id = "<" msg-local "@" msg-domain ">" msg-id = "<" unique "@" mdomain ">" There it ended with another "pseudo-consensus" to wait for the next draft and a ruling from the chair (= you). The latter was of course a joke, but maybe you can toss a coin - in that case don't forget Charles' msg-id = "<" id-left "@" id-right ">" Bye, Frank From owner-ietf-usefor@mail.imc.org Sun Apr 10 22:13:57 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id WAA11772 for ; Sun, 10 Apr 2005 22:13:56 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3B2CbYZ068802 for ; Sun, 10 Apr 2005 19:12:37 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3B2CbXr068801 for ietf-usefor-skb; Sun, 10 Apr 2005 19:12:37 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtp809.mail.ukl.yahoo.com (smtp809.mail.ukl.yahoo.com [217.12.12.199]) by above.proper.com (8.12.11/8.12.9) with SMTP id j3B2Caes068769 for ; Sun, 10 Apr 2005 19:12:36 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from unknown (HELO host81-144-75-36.midband.mdip.bt.net) (ietf-usefor@imc.org@81.144.75.36 with poptime) by smtp809.mail.ukl.yahoo.com with SMTP; 11 Apr 2005 02:12:28 -0000 Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3B2CB209324 for ietf-usefor@imc.org; Mon, 11 Apr 2005 03:12:11 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20647 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: http://www.ietf.org/internet-drafts/draft-ietf-usefor-useage-01.txt Content-Type: text/plain; charset=iso-8859-1 Message-ID: Content-Transfer-Encoding: 8bit X-Newsreader: NN version 6.5.2 (NOV) References: Mime-Version: 1.0 Date: Sun, 10 Apr 2005 22:21:14 GMT Lines: 370 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Content-Transfer-Encoding: 8bit In "Dr John Stockton" writes: >I've been reading >http://www.ietf.org/internet-drafts/draft-ietf-usefor-useage-01.txt, >March 2005. >I don't think I'm qualified to address the mailing list of the Usenet >Format Working Group; If this group were confined to "qualified" people, there would be no group :-). I have copied your message to the group, and am CCing this reply to you, but you will need to subscribe if you want to see any further responses. >"1.2. Objectives >... > NOTE: The extreme irritation caused to other readers by such > violations is not to be underestimated; ..." >And by false accusations of such violations! Yes, but don`t let's over-egg the pudding :-) . >"2.2. Textual Notations >... > NOTE: While such explanatory notes may seem superfluous in > principle, they often help the less-than-omniscient reader > understand the true intent of the specification in cases where > the wording is not entirely clear." >The wording should always be entirely clear, in the opinion of the >author! But the meaning may well not be entirely obvious to such a >reader. Yes. It's a nice quote by Henry Spencer from Son-of-1036 - too good not to use it. >"3.1. The Well-Behaved Posting Agent > The implementor of a posting agent SHOULD make it possible for a > suitably perseverent poster to generate any article, ..." >I don't believe that "perseverent" is a real word, except maybe in the >full OED; "perseverant" is in Chambers' Dictionary; OK, I will follow Chambers. >Last sentence of 3.1 contains "imppose". Fixed. >"3.1.1. Construction of Headers >... > Posting agents SHOULD permit the poster to include headers of > arbitrary length (and MUST permit at least 79 characters)." >You must mean "header lines" or "characters per line" ? No, I meant headers. The next sentence tells them to fold, preferably before 79, but certainly before 998. >... > "NOTE: ... standard 80-column screen" >Since use of text-only screens is now rare, how about something more >general, such as 80-column display, or viewport ? OK, I now say "display". >... > "... WSP ..." >Undefined in useage-01;... Defined in Usefor, by way of RFC 2822. >"3.1.1.1. Date >... > Date: Sat, 26 May 2001 11:13:00 -0500 (EST)" >To rub it in, you could observe (after checking) that EST is used in >Australia, probably (+1000). If someone can quote the Autralian rule authoritatively, and what "EST" stands for there, I might even be tempted to do just that. >I'd like to see full rigorous ISO 8601 dates being allowed, with day-of- >week as an optional but recommended comment - > Date : 2001-05-26 11:13:00 -0500 (EST) (Sat) It would be nice, but the email/news Date format is too entrenched to change it now. >In a time example, I'd choose an hour greater than 12, to expand the >experience of US readers. I think US readers are sufficiently familiar with the 24 hour clock as seen in Date headers. But I have changed it anyway. >"3.1.1.2. From" >Contains the word "ro". Oops! > "NOTE: ..." >ISTM that it might be better to bow to /force majeure/, and indicate, >maybe by reference to another document, what is and what is not >reasonable common practice. In particular, that the form used, after >removal of any ".invalid", MUST NOT be a possible address for someone >else, present or reasonable future. I am not sure. Sometimes people just add ".invalid" to their own genuine address. Not always spammer proof, but most spammers are stupid (Rule #3, or something) - I get massive amounts of spam directed at the Message-ID of my articles from 3 years back. The main purpose of that sentence was to encourage people NOT to munge their addresses WITHOUT adding .invalid as well. >"3.1.1.4. Subject >... > regognized" ! Oops! >"3.1.1.7. Organization >... > ... unless ... and unless ... " >ISTM that in English that's syntactically unreasonable! "unless ... or >... "? No, I meant 'and'. Applying De Morgan's rule, I said "DIScouage UNLESS acceptable AND UNLESS useful", which is the same as "ENcourage IF acceptable OR useful" (or would be if there wasn't a possible middle to exclude). >"3.1.2.1. Signatures > A "personal signature" is a short closing text automatically added" >ISTM that it is not necessarily automatic. OK. How about "added (usually automatically)"? >Since sigs are not normally quoted, you *might* agree with the >Implementors of Turnpike that sig lines can have as many as 79 >characters each. See my own signature below :-) . >"3.1.2.3. Content-Transfer-Encoding >... > moreover, Usenet articles are very likely to include trailing > whitespace in the form of a personal signature (3.1.2.1)." >3.1.2.1 is clear that the signature follows, and does not include, the >separator, which has a whitespace. OK. 'in the form of the "-- " which introduces a personal signature'. >"3.2.1.1. Subject >... > 1. Although the "Re" (which is an abbreviation for the Latin "In re", > meaning "in the matter of", and not an abbreviation of "Reference" > as is sometimes erroneously supposed) may be understood by English > speakers, and indeed by speakers of most European Languages, its > use in a newsgroup where articles were customarily written in > Arabic, or Hindi, or Chinese would be less than helpful." >That seems disrespectful to Eastern intelligence. If they can manage to >use News, they should be able to recognise the odd alien term (however >displayed) occurring always in a specific context. Apparently a lot or orientals are very unfamiliar with the Latin alphabet, but nevertheless are beginning to use mail and even news in their own languages. Granted they cannot avoid it 100% yet. There are domain-names (even if puny-coded), local-parts and newsgroup-names and much else that is still in ASCII, but those issues are being worked on (though slowly). >"3.2.1.1.1. Examples >... > Software can always recognize > that such changes have occurred from the References header." >I don't see that software can do that. ISTM that what you mean is >something different. No. Every followup MUST have a References header, so if software sees a References header it knows it is a followup (or something to be treated like a followup). So it can adjust its display accordingly (e.g. by threading); it does not need to see the "Re: ". From the software writer's POV, "Re: " is just some awkward characters that he has to detect and ignore in order to get his threading (well, some forms of threading) to work properly. >"3.2.1.3. Mail-Copies-To Has now been removed from all our drafts. >If it is permissible for a mail article to contain a newsgroups line >(I've seen it), then ISTM that a mail reading agent should give clear >warning. The intention is that the meaning of a Newsgroups header, if it is seen in an email message, is to indicate that it was also posted to that newsgroup (but replies to that email message should not include such a header unless they are posted to the newsgroup as well). >"3.2.1.4. References >... > carefully put their > by precursors.]" >"there" ! Oops! But it was not in text that was meant to go into the final version. >"3.2.2.1. Quoting and Attributions >... > SHOULD be so dintinguished ..." - distinguished. Genuine Oops! that one. >... > " The followup agent SHOULD also precede the quoted content by an > "attribution line" (however, ..." >I'd like to see a terminological amendment, to remove any taint of a >suggestion that the attribution is a single physical line. Point taken. s/"attribution line"/"attribution"/ and consequential changes elsewhere. >" The attribution MAY contain also a single (the one > from which the followup is being made), the precursor's message > identifier and/or the precursor's Date and Time." >'single' .. usually; but it can be convenient to change that manually to >indicate the degree of cross-posting. Sure. You edit anything manually as you want. But nobody wants to see more than one in the attribution, and I have never seen a user agent that put more than one (in fact they often don't put any). >'Date and Time' - IMHO, the Zone of the date & time is needed and should >be mentioned; and there should be consideration of whether this should >be a direct copy of the original Date:, whether it may be translated >into the responding poster's (or ISO) notation, whether it may be >adjusted to GMT or responder's zone. I do dislike being attributed as >having written on such as 3/21/05, or during AM/PM. I think normal practice is to use the same date-time format as in the Date header. Does anyone know of a system that changes the zone from whatever was actually in the Date header of the precursor? Basically, we are trying to document current best practice here rather than invent anything new. >" o The various fields may be separated by arbitrary text " > ^ brief OK. "arbitrary (but brief) text". Yes, we have all seen some horrors that people have inserted in there, and there is not much we can do to stop them. But no harm in giving a hint and hoping they take it :-( . >"3.3.2. Presentation of Articles >3.3.2.1. Threading > 4. Construct a tree in " ... >ISTM that it *might* be useful for a reading agent to display an >indication, for example, of the difference in tree-depth between that >article and the most previously read article in that thread, presumably >as a signed number, possibly background-coloured for sign. Of course, >all reading-agents should make thread-structure capable of being fully >seen; but it's not necessarily obvious when going from article to >article. Just a thought. I think this is an area where newsreader implementors should be encouraged to experiment. If you don't like what your newsreqder does, then you should go out and buy a better one. I think the purpose of that whole section was to point out various possibilities, each of which has disadvantages if used on its own, and to encourage experimentation with hybrids. There are some agents out there that seem to do a pretty good job, and others that don't. >"3.3.2.2. Killfiles > Moreover, articles > crossposted to many newsgroups SHOULD be considered to have been read > once they have been seen in any of those groups." >But "many" is not the right word, surely. Cross-posting can be to 2 >groups, and 2 < a few < many. s/many/several/ >"3.3.3. Interpretation of Bodies >... > Tab (US-ASCII 9) SHOULD be interpreted as sufficient horizontal white > space to reach the next of a set of fixed positions (customarily set > at every 8th character)." >"... after every 8th character)." OK. >"4.1. Construction of Headers > According to [USEPRO], an injecting agent MAY add other headers not > already provided by the poster, but SHOULD NOT alter, delete, or > reorder any existing header." >It might be useful to attempt to post a copy of that to certain >moderated newsgroups, sic, as comp.lang.asm.x86 :-( . And, perhaps, to >strengthen SHOULD NOT, if possible. No. Moderation comes before injection. Usepro makes it clear that moderators have some discretion to establish a "house style". In a well-managed hierarchy, there will be charters and moderation rules and means to enforce them. OTOH, some charters overdo it with pages and pages of rules governing the moderation process (and you and I know which group I am talking about :-( ). Our drafts need to stand well back from such issues. >"6.1.1. The 'newgroup' and 'mvgroup' Control Messages" >... >And 'rmgroup' ???? Actually no. That section was concerned with the detailed contents of newgroup etc messages, and rmgroup messages don't have any details - it's all or nothing :-) . >"9.2. Construction of Bodies > Posters SHOULD avoid using control characters and escape sequences > except for tab (US-ASCII 9), formfeed (US-ASCII 12) and, possibly, > backspace (US-ASCII 8), for reasons already explained in section > 3.3.3." >I thought that CR & LF were control characters too. Sure, but the requirements for CRLF are well documented in Usefor (via RFC 2822). >"12. Contact Address >Editor > Charles. H. Lindsey" > ^ One wonders what the dot is for. One does indeed! >P.S. I have discovered that the document bears a non-functional E-mail >address; sending the above to :- >" Email: chl@clw.cs.man.ac.uk" >gave >"A message that you sent could not be delivered to one or more of its >recipients. Ah! Manchester have finally pulled the plug on my old email address. -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 From owner-ietf-usefor@mail.imc.org Sun Apr 10 22:13:59 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id WAA11792 for ; Sun, 10 Apr 2005 22:13:59 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3B2CaSE068794 for ; Sun, 10 Apr 2005 19:12:36 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3B2CaIv068793 for ietf-usefor-skb; Sun, 10 Apr 2005 19:12:36 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtp809.mail.ukl.yahoo.com (smtp809.mail.ukl.yahoo.com [217.12.12.199]) by above.proper.com (8.12.11/8.12.9) with SMTP id j3B2CZ32068768 for ; Sun, 10 Apr 2005 19:12:36 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from unknown (HELO host81-144-75-36.midband.mdip.bt.net) (ietf-usefor@imc.org@81.144.75.36 with poptime) by smtp809.mail.ukl.yahoo.com with SMTP; 11 Apr 2005 02:12:24 -0000 Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3B2CCG09328 for ietf-usefor@imc.org; Mon, 11 Apr 2005 03:12:12 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20648 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: Broken Mesage-ID syntax (was: Issues outstanding) Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: <4255153E.3080500@isode.com> <42594879.3B92@xyzzy.claranet.de> Date: Sun, 10 Apr 2005 23:19:09 GMT Lines: 74 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In <42594879.3B92@xyzzy.claranet.de> Frank Ellermann writes: >Alexey Melnikov wrote: >> NO-WS-CTL have been removed. Can people check that the new >> syntax is Ok? >It is not. The "pseudo-consensus" between Charles and me was >to replace msg-id-core by msg-id everywhere, and to fix the >dot-atom-text issue with leading / trailing dots plus dot-dot. Yes, I asked Ken to make that msg-id change, but he hasn't done it yet. The dot-atom-text issue is noted, but not fixed yet. >Two (Charles + me) is a bad number, could somebody like you >please volunteer as third for a proper "rough consensus" about >the last open msg-id point, a better name for the "LHS @ RHS" >productions ? I've interpreted an unrelated article from Bruce >elsewhere as support for "unique @ domain" productions, but >that's obviously dubious, You can count me in for getting rid of , but you can count me out for any name change for and . >> I suspect that some minor issues raised by Frank are yet to >> be addressed. >The msg-id and dot stuff isn't minor, the former is important, >the latter is erroneous. Here's the latest ABNF for a msg-id: >| msg-id = "<" unique "@" mdomain ">" >| unique = dot-atom-text / ( DQUOTE unique-quote DQUOTE ) >| unique-quote = ( "." [unique-part] ) / >| ( [unique-part] "." ) / >| ( [unique-part] unique-literal [unique-part] ) >| unique-part = 1*( atext / "." / unique-literal ) >| unique-literal = "(" / ")" / "," / ; all specials, minus ">", >| "[" / "]" / "@" / ; minus DQUOTE, minus "\", >| ":" / ";" / "<" / ; minus single ".", plus: >| ".." / "\\" / ( "\" DQUOTE ) >| mdomain = dot-atom-text / ("[" address-literal "]") >| address-literal = 1*( %d33-61 / ; printable ASCII minus >| %d63-90 / ; ">", "[", "\", "]" >| %d94-126 / ; plus "\[", "\\, "\]" >| "\[" / "\\" / "\]" ) Well not with those names :-( . But it is still not technically correct. It prevents the msg-id <"foo.bar.baz"@example.com> as required (because under RFC 2822 that is indistinguishable from ), but it also prevents <"@.@"@example.com> and there is no reason to do that (ugly though it may be). BTW, the "[[Adjacent dots should not be allowed]]" remark in the draft is not a correct description of the problem. -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 From owner-ietf-usefor@mail.imc.org Sun Apr 10 22:52:34 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id WAA13658 for ; Sun, 10 Apr 2005 22:52:33 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3B2oQ7U072190 for ; Sun, 10 Apr 2005 19:50:26 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3B2oQno072189 for ietf-usefor-skb; Sun, 10 Apr 2005 19:50:26 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from mail1.panix.com (mail1.panix.com [166.84.1.72]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3B2oP1b072183 for ; Sun, 10 Apr 2005 19:50:25 -0700 (PDT) (envelope-from sethb@panix.com) Received: from panix5.panix.com (panix5.panix.com [166.84.1.5]) by mail1.panix.com (Postfix) with ESMTP id 7932E58AA7 for ; Sun, 10 Apr 2005 22:50:24 -0400 (EDT) Received: (from sethb@localhost) by panix5.panix.com (8.11.6p3/8.8.8/PanixN1.1) id j3B2oOf20818; Sun, 10 Apr 2005 22:50:24 -0400 (EDT) Date: Sun, 10 Apr 2005 22:50:24 -0400 (EDT) Message-Id: <200504110250.j3B2oOf20818@panix5.panix.com> From: Seth Breidbart To: ietf-usefor@imc.org In-reply-to: (chl@clerew.man.ac.uk) Subject: Re: http://www.ietf.org/internet-drafts/draft-ietf-usefor-useage-01.txt References: Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: "Charles Lindsey" wrote: > In "Dr John Stockton" > writes: >> In particular, that the form used, after removal of any >>".invalid", MUST NOT be a possible address for someone else, present >>or reasonable future. > > I am not sure. Sometimes people just add ".invalid" to their own genuine > address. That's why it says "someone _else_". >> Moreover, articles crossposted to many newsgroups SHOULD be >> considered to have been read once they have been seen in any of >> those groups." > >>But "many" is not the right word, surely. Cross-posting can be to 2 >>groups, and 2 < a few < many. > > s/many/several/ How about "multiple"? Seth From owner-ietf-usefor@mail.imc.org Mon Apr 11 06:24:37 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id GAA29270 for ; Mon, 11 Apr 2005 06:24:37 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3BAN0oO010436 for ; Mon, 11 Apr 2005 03:23:00 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3BAN0th010435 for ietf-usefor-skb; Mon, 11 Apr 2005 03:23:00 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtps-vbr1.xs4all.nl (smtps-vbr1.xs4all.nl [194.109.24.19]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3BAMwsV010409 for ; Mon, 11 Apr 2005 03:22:59 -0700 (PDT) (envelope-from rvtol@isolution.nl) Received: from isop10 (velvet.isolution.nl [194.109.164.102]) (authenticated bits=0) by smtps-vbr1.xs4all.nl (8.12.11/8.12.11) with ESMTP id j3BAMp7s067131 (version=TLSv1/SSLv3 cipher=RC4-MD5 bits=128 verify=NO) for ; Mon, 11 Apr 2005 12:22:56 +0200 (CEST) (envelope-from rvtol@isolution.nl) Message-ID: <0f6c01c53e80$6d4372c0$0b01a8c0@isolution.nl> From: "Ruud H.G. van Tol" To: References: <200504110250.j3B2oOf20818@panix5.panix.com> Subject: Re: http://www.ietf.org/internet-drafts/draft-ietf-usefor-useage-01.txt Date: Mon, 11 Apr 2005 12:15:50 +0200 Organization: Chaos rules. MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-15" Content-Transfer-Encoding: 7bit X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 6.00.2800.1478 X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2800.1478 X-Virus-Scanned: by XS4ALL Virus Scanner Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Content-Transfer-Encoding: 7bit Seth Breidbart: > Charles Lindsey: >> Dr John Stockton: >>> In particular, that the form used, after removal of any >>> ".invalid", MUST NOT be a possible address for someone else, >>> present or reasonable future. >> I am not sure. Sometimes people just add ".invalid" to their own >> genuine address. > That's why it says "someone _else_". Same observation here. The "for" didn't emphasize the issue of ownership (or permission) though, so it was easy to read it otherwise, like the "someone else" being the spammert. -- Grtz, Ruud From owner-ietf-usefor@mail.imc.org Mon Apr 11 07:14:09 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id HAA02544 for ; Mon, 11 Apr 2005 07:14:08 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3BBClG3032398 for ; Mon, 11 Apr 2005 04:12:47 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3BBClZU032397 for ietf-usefor-skb; Mon, 11 Apr 2005 04:12:47 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtp805.mail.ukl.yahoo.com (smtp805.mail.ukl.yahoo.com [217.12.12.195]) by above.proper.com (8.12.11/8.12.9) with SMTP id j3BBCjfk032346 for ; Mon, 11 Apr 2005 04:12:46 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from unknown (HELO host81-144-71-41.midband.mdip.bt.net) (ietf-usefor@imc.org@81.144.71.41 with poptime) by smtp805.mail.ukl.yahoo.com with SMTP; 11 Apr 2005 11:12:39 -0000 Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3B9wOe13681 for ietf-usefor@imc.org; Mon, 11 Apr 2005 10:58:24 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20650 Newsgroups: local.usefor Path: clerew!clerew!x Received: from clerew.man.ac.uk (localhost [127.0.0.1]) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) with ESMTP id j3B9uWG13620 for ; Mon, 11 Apr 2005 10:56:33 +0100 (BST) Subject: Fwd: http://www.ietf.org/internet-drafts/draft-ietf-usefor-usefor-03.txt References: Message-ID: To: local.usefor@clerew.man.ac.uk From: "Charles Lindsey" Content-Type: text/plain; format=flowed; delsp=yes; charset=iso-8859-1 MIME-Version: 1.0 Content-Transfer-Encoding: 8bit Date: Mon, 11 Apr 2005 10:56:25 +0100 In-Reply-To: User-Agent: Opera M2(BETA1)/8.00 (SunOS, build 913) Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Content-Transfer-Encoding: 8bit ------- Forwarded message ------- From: "Dr John Stockton" To: "Charles. H. Lindsey" Subject: http://www.ietf.org/internet-drafts/draft-ietf-usefor-usefor-03.txt Date: Sun, 10 Apr 2005 19:44:55 +0100 I've also been reading http://www.ietf.org/internet-drafts/draft-ietf-usefor-usefor-03.txt, April 2005. "Changes since draft-ietf-usefor-usefor-01 ... o Only allow "UT and "GMT" in Date header; disallow all other ." AIUI, it is UTC and not UT that is distributed by time servers, in which case a label UT would in practice be incorrect in principle. UT != UTC. Anyway, IIRC, UT is GMT, except when it is the University of Texas. Note that while input of UTC should allow 23:59:60, in this context output should not generate it, since an error of <= one second is better than a risk of a receiving agent getting upset about it. A posting agent can just wait for a leap second to have finished. However, I see that the tenor of 3.1.2 is incompatible with the introduction of UTC. "1.1 Basic Concepts "Netnews" is a set of protocols for generating, storing and retrieving news "articles" (which are a subset of Email messages)" They are surely not a subset of E-mail messages. They might be described as similar. But E-mail is now commonly HTML & News should not be; I'd omit the parenthetic analogy. "1.2 Scope ... An best ..." Ouch. "1.5 Definitions ... When an article is posted to more than one newsgroup, it is said to be "crossposted"; ..." 'When a single [copy of an ] article ...' ??? "2.2 Headers" and elsewhere. The common understanding of "Header[s]" is "That cabalistic stuff before the message starts", whereas Header is evidently used here to refer to a logical line starting with such as "Date: ". ISTM that, clear definitions notwithstanding, there is an undue risk of reader error unless there is a very clear written distinction in all references to the entire heading material, the logical (de-folded) lines, and the physical (to CRLF) lines. That part got me confused. "3.1.2 Date" ISTM that "GMT" is useful to the human reader, as indicating a choice of the world standard time rather than the local one. My Turnpike here uses +0000 in Winter Time, which carries an expectation that it will use +0100 in Summer Time, and that the relevant agent is in the London time zone. GMT would carry no such implication. Then agents such as Google should be encouraged to use GMT rather than -0800, if they cannot use the author's zone. In other words, I suggest un-deprecation of GMT (and/or UT/UTC). Presumably absence of a numeric offset indication, while not allowed, would be treated as +0000. "3.3.1 Lines" This may not be needed for the transmission of News; but it is very useful for Kill Rules. Therefore, IMHO it should be compulsory, not obsolete (unless replaced by, say, 'Bytes'). "5. Security Considerations ... Agents that generate message-ids for news articles SHOULD ensure that they are unpredictable." It would be undesirable to lose the ability to kill-rule on partial message-ID, or to kill spam addressed to message-IDs using partial ID, or to search news databases on partial ID. Therefore, I suggest something like "incompletely predictable". The right hand side in fact should be predictable. Example : the first two of the last four characters of a Turnpike message-ID are, for a given setup, constant over long periods of time; mine are currently Fw. Therefore, it could be useful for me to kill all mail to .*fw..@, provided that I have no user called halfwit@ or similar. AIUI, newer Turnpike allows the user to choose the last few characters of the left part, which is even better as a kill target. Regards, -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133   Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk      Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9      Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 From owner-ietf-usefor@mail.imc.org Mon Apr 11 10:39:22 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id KAA18370 for ; Mon, 11 Apr 2005 10:39:21 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3BEaqTJ003987 for ; Mon, 11 Apr 2005 07:36:52 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3BEaq4n003986 for ietf-usefor-skb; Mon, 11 Apr 2005 07:36:52 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from ciao.gmane.org (main.gmane.org [80.91.229.2]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3BEamHX003967 for ; Mon, 11 Apr 2005 07:36:50 -0700 (PDT) (envelope-from usenet-format@gmane.org) Received: from list by ciao.gmane.org with local (Exim 4.43) id 1DKzxX-0006do-Ce for ietf-usefor@imc.org; Mon, 11 Apr 2005 16:32:15 +0200 Received: from c-134-88-77.hh.dial.de.ignite.net ([62.134.88.77]) by main.gmane.org with esmtp (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Mon, 11 Apr 2005 16:32:15 +0200 Received: from nobody by c-134-88-77.hh.dial.de.ignite.net with local (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Mon, 11 Apr 2005 16:32:15 +0200 X-Injected-Via-Gmane: http://gmane.org/ To: ietf-usefor@imc.org From: Frank Ellermann Subject: Re: Broken Message-ID syntax Date: Mon, 11 Apr 2005 16:32:03 +0200 Organization: Lines: 44 Message-ID: <425A8A63.268A@xyzzy.claranet.de> References: <4255153E.3080500@isode.com> <42594879.3B92@xyzzy.claranet.de> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Complaints-To: usenet@sea.gmane.org X-Gmane-NNTP-Posting-Host: c-134-88-77.hh.dial.de.ignite.net X-Mailer: Mozilla 3.0 (OS/2; U) Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Content-Transfer-Encoding: 7bit Charles Lindsey wrote: > you can count me out for any name change for and > . That's why we need a third opinion like Alexejs's about this point, or a better question. A recent joke on ietf.general: /\ /\ __/ \__ ___/ \___/\___ _/\__/\__/\_ consensus rough consensus bad question [ABNF] > Well not with those names :-( . As long as the RHS contains "domain" and the "address-literal" is an "address-literal" you could beutify it, all productions starting with "id-" could be a good idea (id-local, id-domain, etc.) > It prevents the msg-id > <"foo.bar.baz"@example.com> > as required Yes, that's the idea: Leading dot or trailing dot or unique-literal <=> quote No leading/trailing dot and no unique-literal <=> don't quote > it also prevents > <"@.@"@example.com> "@" is a "unique literal" => quote. Working as designed. "@" is no atext => no dot-atom-text => unquoted is illegal. Where's the problem ? > the "[[Adjacent dots should not be allowed]]" remark in > the draft is not a correct description of the problem. In the proposed "unique" ABNF adjacent dots are handled by ".." in "unique-literal". Ugly like hell, but who cares if it's correct. Bye, Frank From owner-ietf-usefor@mail.imc.org Mon Apr 11 11:52:39 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id LAA24001 for ; Mon, 11 Apr 2005 11:52:39 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3BFpKbG019098 for ; Mon, 11 Apr 2005 08:51:20 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3BFpKk1019097 for ietf-usefor-skb; Mon, 11 Apr 2005 08:51:20 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from spsystems.net (spsystems.net [216.126.83.115]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3BFpJqN019080 for ; Mon, 11 Apr 2005 08:51:19 -0700 (PDT) (envelope-from henry@spsystems.net) Received: from spsystems.net (henry@localhost [127.0.0.1]) by spsystems.net (8.12.10/8.12.10) with ESMTP id j3BFpEVO022083; Mon, 11 Apr 2005 11:51:14 -0400 (EDT) Received: (from henry@localhost) by spsystems.net (8.12.10/8.12.10/Submit) id j3BFpEmU022082; Mon, 11 Apr 2005 11:51:14 -0400 (EDT) Date: Mon, 11 Apr 2005 11:51:13 -0400 (EDT) From: Henry Spencer To: Usefor Mailing List Subject: Re: Path header delimiters In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: On Sat, 9 Apr 2005, Charles Lindsey wrote: > E. A variant on scheme D > Path: good-site.com!!mallet.com!MISMATCH!dodgy.com!old-site.com! > new-site.com!!injector.com!POSTED!not-for-mail > This avoids any delimiter other than '!', but it assumes that two > delimiters in succession will not cause any trouble. But it is clear that > RFC 1036 permits such usage... This sounds good to me. The one slight amendment I would suggest would be to use "AKA" -- Also Known As -- instead of MISMATCH, making it a bit clearer just what's going on. (I struggled to find a concise way of writing "claimed to be", which is the underlying idea, and this was the best I could think of.) I support scheme E even if this suggestion is not accepted. Henry Spencer henry@spsystems.net From owner-ietf-usefor@mail.imc.org Mon Apr 11 12:14:03 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id MAA25616 for ; Mon, 11 Apr 2005 12:14:02 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3BGChVj023522 for ; Mon, 11 Apr 2005 09:12:43 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3BGChfa023520 for ietf-usefor-skb; Mon, 11 Apr 2005 09:12:43 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtp802.mail.ukl.yahoo.com (smtp802.mail.ukl.yahoo.com [217.12.12.139]) by above.proper.com (8.12.11/8.12.9) with SMTP id j3BGCgPA023491 for ; Mon, 11 Apr 2005 09:12:42 -0700 (PDT) (envelope-from chl@clerew.man.ac.uk) Received: from unknown (HELO host81-144-77-126.midband.mdip.bt.net) (ietf-usefor@imc.org@81.144.77.126 with poptime) by smtp802.mail.ukl.yahoo.com with SMTP; 11 Apr 2005 16:12:35 -0000 Received: (from chl@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3BCChM14516; Mon, 11 Apr 2005 13:12:43 +0100 (BST) Date: Mon, 11 Apr 2005 13:12:43 +0100 (BST) From: Charles Lindsey Message-Id: <200504111212.j3BCChM14516@clerew.man.ac.uk> To: ietf-usefor@imc.org Cc: "Dr John Stockton" Subject: Re: Fwd: http://www.ietf.org/internet-drafts/draft-ietf-usefor-usefor-03.txt References: X-Newsreader: NN version 6.5.2 (NOV) Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In local.usefor you write: >------- Forwarded message ------- >From: "Dr John Stockton" >To: "Charles. H. Lindsey" >Subject: >http://www.ietf.org/internet-drafts/draft-ietf-usefor-usefor-03.txt >Date: Sun, 10 Apr 2005 19:44:55 +0100 >I've also been reading >http://www.ietf.org/internet-drafts/draft-ietf-usefor-usefor-03.txt, >April 2005. I think most of these points are for Ken Murchison to answer. So I shall just make a few comments. >"Changes since draft-ietf-usefor-usefor-01 >... > o Only allow "UT and "GMT" in Date header; disallow all other zone>." >AIUI, it is UTC and not UT that is distributed by time servers, in which >case a label UT would in practice be incorrect in principle. UT != UTC. >Anyway, IIRC, UT is GMT, except when it is the University of Texas. The only reason these two s remain in the draft is that they are in current widespread use (well, GMT is anyway). So they must continue to be accepted, even though they MUST NOT be generated in future. Generally speaking, neither news nor email software cares in the least about the subtleties of UTC vs GMT. Times, as they appear in Date headers, are just approximations to the time the article was written, and cannot be assumed to be accurate to better than a few minutes. >Note that while input of UTC should allow 23:59:60,... That format for time is already permitted by RFC 2822, though I doubt any operating system ever produces it in real practice. >"1.1 Basic Concepts > "Netnews" is a set of protocols for generating, storing and > retrieving news "articles" (which are a subset of Email messages)" >They are surely not a subset of E-mail messages. They might be >described as similar. But E-mail is now commonly HTML & News should not >be; I'd omit the parenthetic analogy. Yes, it might be better to say "(whose format is a subset of that for Email messages)". Ken? >"1.5 Definitions >... > When an article is posted to more than one newsgroup, it > is said to be "crossposted"; ..." >'When a single [copy of an ] article ...' ??? I am not convinced that needs changing. >"2.2 Headers" and elsewhere. >The common understanding of "Header[s]" is "That cabalistic stuff before >the message starts", whereas Header is evidently used here to refer to a >logical line starting with such as "Date: ". We took a conscious decision to use the terminology in widespread use, rather than that defined in RFC 2822. The new NNTP draft is following that same approach. So the "cabalistic stuff" is the "Headers" (plural) and the Date header, for example, is a "Header" (singular). >ISTM that, clear definitions notwithstanding, there is an undue risk of >reader error unless there is a very clear written distinction in all >references to the entire heading material, the logical (de-folded) >lines, and the physical (to CRLF) lines. That part got me confused. I think (hope) that the draft always makes it clear from the context when a single "header line" is being discussed. >"3.1.2 Date" >ISTM that "GMT" is useful to the human reader, as indicating a choice of >the world standard time rather than the local one. My Turnpike here >uses +0000 in Winter Time, which carries an expectation that it will use >+0100 in Summer Time, and that the relevant agent is in the London time >zone. GMT would carry no such implication. Then agents such as Google >should be encouraged to use GMT rather than -0800, if they cannot use >the author's zone. In other words, I suggest un-deprecation of GMT >(and/or UT/UTC). I don't think there has ever been any expectation that you can guess, from somebody's Date header, whether that somebody lives in a location where daylight saving applies. And I believe Madrid also uses +0000 in Winter (and if Madrid doesn't, then Lisbon surely will - and what about Casablanca)? Generally speaking, the poster's timezone, as written, should be preserved as the article propagates or gets stored (indeed, gratuitous changes of headers are forbidden by Usepro). I find it useful to know whether someone really was taking the trouble to stay up late to respond to my messages. >Presumably absence of a numeric offset indication, while not allowed, >would be treated as +0000. Possibly. >"3.3.1 Lines" >This may not be needed for the transmission of News; but it is very >useful for Kill Rules. Therefore, IMHO it should be compulsory, not >obsolete (unless replaced by, say, 'Bytes'). I think the WG agreed early on that this header had passed its sell-by-date, though I do not think we intended it to be declared "obsolete" as the current Usefor draft implies. I think "obsolescent" is the right term. At least the draft does define exactly which lines are supposed to get counted, which had always been a bit vague hitherto. >"5. Security Considerations >... > Agents that generate message-ids for news articles > SHOULD ensure that they are unpredictable." >It would be undesirable to lose the ability to kill-rule on partial >message-ID, or to kill spam addressed to message-IDs using partial ID, >or to search news databases on partial ID. Therefore, I suggest >something like "incompletely predictable". The right hand side in fact >should be predictable. RFC 2822 RECOMMENDS (which is a pretty strong term) that the RHS should be the domain of the sender (or of his ISP), and Usefor inherits that RECCOMENDATION. -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 From owner-ietf-usefor@mail.imc.org Mon Apr 11 12:14:08 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id MAA25634 for ; Mon, 11 Apr 2005 12:14:08 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3BGCdGc023504 for ; Mon, 11 Apr 2005 09:12:39 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3BGCdrc023503 for ietf-usefor-skb; Mon, 11 Apr 2005 09:12:39 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtp802.mail.ukl.yahoo.com (smtp802.mail.ukl.yahoo.com [217.12.12.139]) by above.proper.com (8.12.11/8.12.9) with SMTP id j3BGCcVS023474 for ; Mon, 11 Apr 2005 09:12:38 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from unknown (HELO host81-144-77-126.midband.mdip.bt.net) (ietf-usefor@imc.org@81.144.77.126 with poptime) by smtp802.mail.ukl.yahoo.com with SMTP; 11 Apr 2005 16:12:32 -0000 Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3BGCEZ15275 for ietf-usefor@imc.org; Mon, 11 Apr 2005 17:12:14 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20651 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: http://www.ietf.org/internet-drafts/draft-ietf-usefor-useage-01.txt Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: <200504110250.j3B2oOf20818@panix5.panix.com> Date: Mon, 11 Apr 2005 12:14:32 GMT Lines: 29 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In <200504110250.j3B2oOf20818@panix5.panix.com> Seth Breidbart writes: >"Charles Lindsey" wrote: >> In "Dr John Stockton" >> writes: >>> Moreover, articles crossposted to many newsgroups SHOULD be >>> considered to have been read once they have been seen in any of >>> those groups." >> >>>But "many" is not the right word, surely. Cross-posting can be to 2 >>>groups, and 2 < a few < many. >> >> s/many/several/ >How about "multiple"? I think "several" is good enough. Anyone else preferring "multiple"? -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 From owner-ietf-usefor@mail.imc.org Mon Apr 11 13:35:55 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id NAA02824 for ; Mon, 11 Apr 2005 13:35:54 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3BHXvoP040743 for ; Mon, 11 Apr 2005 10:33:57 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3BHXvlI040742 for ietf-usefor-skb; Mon, 11 Apr 2005 10:33:57 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from a.mail.peak.org (a.mail.peak.org [69.59.192.41]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3BHXu7E040718 for ; Mon, 11 Apr 2005 10:33:56 -0700 (PDT) (envelope-from stanley@peak.org) Received: from a.shell.peak.org ([69.59.192.81]) by a.mail.peak.org (8.12.10/8.12.8) with ESMTP id j3BHXokK038935 (version=TLSv1/SSLv3 cipher=EDH-RSA-DES-CBC3-SHA bits=168 verify=NO) for ; Mon, 11 Apr 2005 10:33:51 -0700 (PDT) Date: Mon, 11 Apr 2005 10:33:50 -0700 (PDT) From: John Stanley X-X-Sender: stanley@a.shell.peak.org To: ietf-usefor@imc.org Subject: Re: Differences between RFC 2822 and Usefor Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Spam-Score: 0 () X-Scanned-By: MIMEDefang 2.39 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: "Charles Lindsey" : > Differences between RFC 2822 and Usefor. > ---------------------------------------- >2.11 A References header MUST be provided for followups (as opposed to >SHOULD be provided for replies in RFC 2822). Would you care to explain where you are getting this difference? In draft-ietf-usefor-usefor-03.txt, the References header is defined thusly: >3.2.1 References > The References header is the same as that specified in Section 3.6.4 > of [RFC2822] with the added restrictions detailed in Section 2.2 and > those listed below: Section 2.2 does not say anything about specific headers (and nothing at all about References), and the remaining part of 3.2.1 says nothing about followups and MUST. RFC2822 says "SHOULD". There is no section 2.11 at all, much less one that creates a difference between RFC2822 and USEFOR. So, it appears to me that the artificial requirement for References has been removed from the draft, as would be proper based upon a strict reading of RFC2119 and when a MUST is appropriate. So be it. That's one way of solving the debate between "followups MUST/non-followups MUST NOT" and "followups MUST/non-followups MAY". Not the way I would have picked, but one that is certainly RFC2119 conforming. Of course, now that this change has been made, it will require a full justification for putting back an extraneous (w.r.t. RFC2119) MUST condition. There is no interoperability issue, nothing breaks when it isn't there, so it will take some special arguments to justify replacing it. From owner-ietf-usefor@mail.imc.org Mon Apr 11 13:44:43 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id NAA03433 for ; Mon, 11 Apr 2005 13:44:43 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3BHhaEF043044 for ; Mon, 11 Apr 2005 10:43:36 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3BHhaje043043 for ietf-usefor-skb; Mon, 11 Apr 2005 10:43:36 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from b.mail.peak.org (b.mail.peak.org [69.59.192.42]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3BHhaJ0043019 for ; Mon, 11 Apr 2005 10:43:36 -0700 (PDT) (envelope-from stanley@peak.org) Received: from a.shell.peak.org ([69.59.192.81]) by b.mail.peak.org (8.12.10/8.12.8) with ESMTP id j3BHhUTF066946 (version=TLSv1/SSLv3 cipher=EDH-RSA-DES-CBC3-SHA bits=168 verify=NO) for ; Mon, 11 Apr 2005 10:43:30 -0700 (PDT) Date: Mon, 11 Apr 2005 10:43:30 -0700 (PDT) From: John Stanley X-X-Sender: stanley@a.shell.peak.org To: ietf-usefor@imc.org Subject: Re: Issues outstanding Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Spam-Score: 0 () X-Scanned-By: MIMEDefang 2.39 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: "Charles Lindsey" : >(Oddly, I noticed that John Stanley was in favour of Posted-And-Mailed - Why is this so odd, Charles? I was one of the people who worked on the draft that tried to define PAM and MCT. Do you think I would have spent my time on that were I opposed to them? >he is quite right, but if Mail- Copies-To has to wait, then >Posted-And-Mailed must wait too.) If our task is to document existing practice, and if you were correct in writing: >It is still not clear (to me) what the objection to keeping them is, and I >see no merit at all in #2b (since these headers are in moderately common >use, and the "experiment" has, in effect, been done). as Alexy quoted you as saying, then it would be incorrect NOT to include them. Are they not "current practice", are they NOT in moderately common use, or are we sticking our heads in the sand and ignoring the task we are supposed to accomplish here? I notice that Alexy did not debate the validitiy of your statement, he just decided the issue is closed and we will not document existing practice here. From owner-ietf-usefor@mail.imc.org Mon Apr 11 14:47:04 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id OAA07990 for ; Mon, 11 Apr 2005 14:47:04 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3BIjjOX055327 for ; Mon, 11 Apr 2005 11:45:45 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3BIjjSW055326 for ietf-usefor-skb; Mon, 11 Apr 2005 11:45:45 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtp1.Stanford.EDU (smtp1.Stanford.EDU [171.67.16.123]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3BIjiff055319 for ; Mon, 11 Apr 2005 11:45:44 -0700 (PDT) (envelope-from rra@stanford.edu) Received: from windlord.stanford.edu (windlord.Stanford.EDU [171.64.19.147]) by smtp1.Stanford.EDU (8.12.11/8.12.11) with SMTP id j3BIjhW3009007 for ; Mon, 11 Apr 2005 11:45:43 -0700 Received: (qmail 32689 invoked by uid 1000); 11 Apr 2005 18:45:43 -0000 To: ietf-usefor@imc.org, "Dr John Stockton" Subject: Re: Fwd: http://www.ietf.org/internet-drafts/draft-ietf-usefor-usefor-03.txt In-Reply-To: <200504111212.j3BCChM14516@clerew.man.ac.uk> (Charles Lindsey's message of "Mon, 11 Apr 2005 13:12:43 +0100 (BST)") References: <200504111212.j3BCChM14516@clerew.man.ac.uk> From: Russ Allbery Organization: The Eyrie Date: Mon, 11 Apr 2005 11:45:43 -0700 Message-ID: <87psx1gnjc.fsf@windlord.stanford.edu> User-Agent: Gnus/5.1007 (Gnus v5.10.7) XEmacs/21.4 (Jumbo Shrimp, linux) MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Charles Lindsey writes: > "Dr John Stockton" writes: >> "3.3.1 Lines" >> This may not be needed for the transmission of News; but it is very >> useful for Kill Rules. Therefore, IMHO it should be compulsory, not >> obsolete (unless replaced by, say, 'Bytes'). Kill Rules are based on overview information, not on article headers except in unusual situations (since killing on full headers is very slow). Overview contains its own line count that has nothing to do with an article header; this is specified by the NNTP protocol. You will find, if you analyze articles in the wild, that a significant percentage of Lines headers are wrong. > I think the WG agreed early on that this header had passed its > sell-by-date, though I do not think we intended it to be declared > "obsolete" as the current Usefor draft implies. I believe that we did. We intended it to be declared obsolete in the sense that nothing should generate it and nothing should use it. -- Russ Allbery (rra@stanford.edu) From owner-ietf-usefor@mail.imc.org Mon Apr 11 14:59:45 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id OAA08835 for ; Mon, 11 Apr 2005 14:59:44 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3BIwH36057762 for ; Mon, 11 Apr 2005 11:58:19 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3BIwHgK057761 for ietf-usefor-skb; Mon, 11 Apr 2005 11:58:17 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from ciao.gmane.org (main.gmane.org [80.91.229.2]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3BIwEQw057740 for ; Mon, 11 Apr 2005 11:58:15 -0700 (PDT) (envelope-from usenet-format@gmane.org) Received: from list by ciao.gmane.org with local (Exim 4.43) id 1DL43W-0004CA-5n for ietf-usefor@imc.org; Mon, 11 Apr 2005 20:54:42 +0200 Received: from 212.82.251.33 ([212.82.251.33]) by main.gmane.org with esmtp (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Mon, 11 Apr 2005 20:54:42 +0200 Received: from nobody by 212.82.251.33 with local (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Mon, 11 Apr 2005 20:54:42 +0200 X-Injected-Via-Gmane: http://gmane.org/ To: ietf-usefor@imc.org From: Frank Ellermann Subject: Re: Path header delimiters Date: Mon, 11 Apr 2005 20:56:47 +0200 Organization: Lines: 12 Message-ID: <425AC86F.2004@xyzzy.claranet.de> References: Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Complaints-To: usenet@sea.gmane.org X-Gmane-NNTP-Posting-Host: 212.82.251.33 X-Mailer: Mozilla 3.0 (OS/2; U) Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Content-Transfer-Encoding: 7bit Henry Spencer wrote: > I would suggest would be to use "AKA" -- Also Known As -- > instead of MISMATCH Two problems: MISMATCH is already "common practice", and we hope that no UUCP host MISMATCH existed, and that no TLD with this name will be introduced. For "AKA" I'm less sure, funny host names are AI, IO, PH, PN, TM, and TW - in theory these hosts could be news servers. Bye, Frank From owner-ietf-usefor@mail.imc.org Mon Apr 11 15:10:35 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id PAA10337 for ; Mon, 11 Apr 2005 15:10:34 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3BJ9auO059661 for ; Mon, 11 Apr 2005 12:09:36 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3BJ9ab7059660 for ietf-usefor-skb; Mon, 11 Apr 2005 12:09:36 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from spsystems.net (spsystems.net [216.126.83.115]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3BJ9abd059650 for ; Mon, 11 Apr 2005 12:09:36 -0700 (PDT) (envelope-from henry@spsystems.net) Received: from spsystems.net (henry@localhost [127.0.0.1]) by spsystems.net (8.12.10/8.12.10) with ESMTP id j3BJ9WVO024343; Mon, 11 Apr 2005 15:09:32 -0400 (EDT) Received: (from henry@localhost) by spsystems.net (8.12.10/8.12.10/Submit) id j3BJ9WTO024342; Mon, 11 Apr 2005 15:09:32 -0400 (EDT) Date: Mon, 11 Apr 2005 15:09:32 -0400 (EDT) From: Henry Spencer To: Usefor Mailing List Subject: Re: Path header delimiters In-Reply-To: <425AC86F.2004@xyzzy.claranet.de> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: On Mon, 11 Apr 2005, Frank Ellermann wrote: > > I would suggest would be to use "AKA" -- Also Known As -- > > instead of MISMATCH > > Two problems: MISMATCH is already "common practice"... As a standalone item in the path content? I think not. There is some vaguely-related existing practice of appending it to server names, which might be considered a minor point in its favor. > and we > hope that no UUCP host MISMATCH existed, and that no TLD with > this name will be introduced. For "AKA" I'm less sure... We can quite reasonably have the same hopes for AKA. Remember also that it doesn't greatly matter whether AKA is potentially a valid server name, because such a server can be known by a nickname (e.g. "news.aka") for news purposes. This would be a nuisance for its proprietors, yes, but not a problem for the net as a whole. Henry Spencer henry@spsystems.net From owner-ietf-usefor@mail.imc.org Mon Apr 11 16:02:12 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id QAA15727 for ; Mon, 11 Apr 2005 16:02:12 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3BK02ZE069521 for ; Mon, 11 Apr 2005 13:00:02 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3BK02Zp069520 for ietf-usefor-skb; Mon, 11 Apr 2005 13:00:02 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from ciao.gmane.org (main.gmane.org [80.91.229.2]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3BK00VW069503 for ; Mon, 11 Apr 2005 13:00:01 -0700 (PDT) (envelope-from usenet-format@gmane.org) Received: from list by ciao.gmane.org with local (Exim 4.43) id 1DL51h-0005A7-Fi for ietf-usefor@imc.org; Mon, 11 Apr 2005 21:56:53 +0200 Received: from 212.82.251.33 ([212.82.251.33]) by main.gmane.org with esmtp (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Mon, 11 Apr 2005 21:56:53 +0200 Received: from nobody by 212.82.251.33 with local (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Mon, 11 Apr 2005 21:56:53 +0200 X-Injected-Via-Gmane: http://gmane.org/ To: ietf-usefor@imc.org From: Frank Ellermann Subject: Re: Path header delimiters Date: Mon, 11 Apr 2005 21:58:39 +0200 Organization: Lines: 25 Message-ID: <425AD6EF.448A@xyzzy.claranet.de> References: <425AC86F.2004@xyzzy.claranet.de> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Complaints-To: usenet@sea.gmane.org X-Gmane-NNTP-Posting-Host: 212.82.251.33 X-Mailer: Mozilla 3.0 (OS/2; U) Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Content-Transfer-Encoding: 7bit Henry Spencer wrote: [MISMATCH] > As a standalone item in the path content? I think not. Probably I was wrong, a quick check confirmed your version ".MISMATCH" added to the host name, not "!MISMATCH". > This would be a nuisance for its proprietors, yes, but not > a problem for the net as a whole. Also true, the former host TV apparently found a better name. A potential host AKA after the introduction of TLD AKA could also do this as news server. Besides ...!news.aka!AKA!... would be fun. More fun: Any news about publishing s-o-1036 as RfC 4036 ? It's apparently not yet in the RfC editor queue. Less fun: Charles and I need a volunteer for some kind of "rough consensus" about the names (sic!) of some msg-id ABNF productions, could you please toss a coin or add a comment ? Bye, Frank From owner-ietf-usefor@mail.imc.org Tue Apr 12 05:33:17 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id FAA12653 for ; Tue, 12 Apr 2005 05:33:17 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3C9W4ur043058 for ; Tue, 12 Apr 2005 02:32:04 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3C9W4H3043057 for ietf-usefor-skb; Tue, 12 Apr 2005 02:32:04 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from rufus.isode.com (rufus.isode.com [62.3.217.251]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3C9W1WJ043010 for ; Tue, 12 Apr 2005 02:32:03 -0700 (PDT) (envelope-from alexey.melnikov-usefor@isode.com) Received: from [192.168.0.3] ([62.3.217.253]) by rufus.isode.com via TCP (internal) with ESMTPA; Tue, 12 Apr 2005 10:31:55 +0100 Message-ID: <425B958A.9010107@isode.com> Date: Tue, 12 Apr 2005 10:31:54 +0100 From: Alexey Melnikov User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.7.2) Gecko/20040804 Netscape/7.2 (ax) X-Accept-Language: en-us, en MIME-Version: 1.0 To: Russ Allbery CC: ietf-usefor@imc.org, Dr John Stockton Subject: Re: Fwd: http://www.ietf.org/internet-drafts/draft-ietf-usefor-usefor-03.txt References: <200504111212.j3BCChM14516@clerew.man.ac.uk> <87psx1gnjc.fsf@windlord.stanford.edu> In-Reply-To: <87psx1gnjc.fsf@windlord.stanford.edu> Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Content-Transfer-Encoding: 7bit Russ Allbery wrote: >Charles Lindsey writes: > > >>"Dr John Stockton" writes: >> >> >>>"3.3.1 Lines" >>> >>> >>>This may not be needed for the transmission of News; but it is very >>>useful for Kill Rules. Therefore, IMHO it should be compulsory, not >>>obsolete (unless replaced by, say, 'Bytes'). >>> >>> > >Kill Rules are based on overview information, not on article headers >except in unusual situations (since killing on full headers is very slow). >Overview contains its own line count that has nothing to do with an >article header; this is specified by the NNTP protocol. > >You will find, if you analyze articles in the wild, that a significant >percentage of Lines headers are wrong. > > >>I think the WG agreed early on that this header had passed its >>sell-by-date, though I do not think we intended it to be declared >>"obsolete" as the current Usefor draft implies. >> >> >I believe that we did. We intended it to be declared obsolete in the >sense that nothing should generate it and nothing should use it. > > Exactly. I don't think any clarification in the document is needed. From owner-ietf-usefor@mail.imc.org Tue Apr 12 07:14:05 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id HAA19776 for ; Tue, 12 Apr 2005 07:14:05 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3CBCfS1078142 for ; Tue, 12 Apr 2005 04:12:41 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3CBCfp2078141 for ietf-usefor-skb; Tue, 12 Apr 2005 04:12:41 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtp811.mail.ukl.yahoo.com (smtp811.mail.ukl.yahoo.com [217.12.12.201]) by above.proper.com (8.12.11/8.12.9) with SMTP id j3CBCceg078094 for ; Tue, 12 Apr 2005 04:12:40 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from unknown (HELO host81-144-77-234.midband.mdip.bt.net) (ietf-usefor@imc.org@81.144.77.234 with poptime) by smtp811.mail.ukl.yahoo.com with SMTP; 12 Apr 2005 11:12:31 -0000 Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3CBCFj21322 for ietf-usefor@imc.org; Tue, 12 Apr 2005 12:12:15 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20662 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: Differences between RFC 2822 and Usefor Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: Date: Tue, 12 Apr 2005 11:08:04 GMT Lines: 62 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In John Stanley writes: >"Charles Lindsey" : >> Differences between RFC 2822 and Usefor. >> ---------------------------------------- >>2.11 A References header MUST be provided for followups (as opposed to >>SHOULD be provided for replies in RFC 2822). >Would you care to explain where you are getting this difference? In >draft-ietf-usefor-usefor-03.txt, the References header is defined thusly: The "SHOULD" in question is in the first paragraph of section 3.6.4 of RFC 2822. In consequence of that SHOULD, many MUAs still regard providing a References header as an "optional extra", with the consequence that attempts to display threading in mailing lists (including this one) usually succeed only in showing broken threads. The "MUST" in question arises from the long held consensus in this Working Group that the References header in News is NOT an "optional extra". And you yourself have been at the forefront in upholding that consensus. There is a disagreement as to how exactly that requirement should be expressed, and as to how the wording should reflect the use of the References header in the case of multipart FAQs and the like, but there is no disagreement at all that it MUST be provided in cases where a poster follows up (responds/replies/whatever) to an article from an earlier poster. That clear requirement has been in all our drafts up to article-13. >>3.2.1 References >> The References header is the same as that specified in Section 3.6.4 >> of [RFC2822] with the added restrictions detailed in Section 2.2 and >> those listed below: >Section 2.2 does not say anything about specific headers (and nothing at >all about References), and the remaining part of 3.2.1 says nothing about >followups and MUST. RFC2822 says "SHOULD". The text regarding the References header in the new usefor-03 draft is just plain WRONG. Yes. the bit about section 2.2 covers the obligatory SP after the ':', as required in all headers, and the "listed below" refers to the avoidance of comments (MUST accept but do not generate yet). But the requirement for the header to be present for followups has been omitted, and it needs to be put back (modulo the relatively minor matter of the precise wording). And there is also an error in the syntax of the References header, which Frank Ellermann pointed out some while back, and which needs to be corrected. -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 From owner-ietf-usefor@mail.imc.org Tue Apr 12 08:02:55 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id IAA22579 for ; Tue, 12 Apr 2005 08:02:55 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3CC1IF8095225 for ; Tue, 12 Apr 2005 05:01:18 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3CC1IoO095223 for ietf-usefor-skb; Tue, 12 Apr 2005 05:01:18 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtp801.mail.ukl.yahoo.com (smtp801.mail.ukl.yahoo.com [217.12.12.138]) by above.proper.com (8.12.11/8.12.9) with SMTP id j3CC1Ee3095146 for ; Tue, 12 Apr 2005 05:01:17 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from unknown (HELO host81-144-77-103.midband.mdip.bt.net) (ietf-usefor@imc.org@81.144.77.103 with poptime) by smtp801.mail.ukl.yahoo.com with SMTP; 12 Apr 2005 12:01:08 -0000 Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3CC0tQ22216 for ietf-usefor@imc.org; Tue, 12 Apr 2005 13:00:55 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20665 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: Path header delimiters Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: Date: Tue, 12 Apr 2005 11:36:53 GMT Lines: 32 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In Henry Spencer writes: >On Mon, 11 Apr 2005, Frank Ellermann wrote: >> > I would suggest would be to use "AKA" -- Also Known As -- >> > instead of MISMATCH >> >> Two problems: MISMATCH is already "common practice"... >As a standalone item in the path content? I think not. There is some >vaguely-related existing practice of appending it to server names, which >might be considered a minor point in its favor. You are damning it with faint praise :-) . Actually, I think I prefer MISMATCH, not only because it is close to some existing practice, but also because it is a conspicuous word that will shout out at you "there is something fishy about this Path" whenever you see it. Anyway, it a minor detail as regards scheme E, which seems to be gaining added consenus. Does anyone else want to play? -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 From owner-ietf-usefor@mail.imc.org Tue Apr 12 08:03:08 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id IAA22636 for ; Tue, 12 Apr 2005 08:03:08 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3CC1MUb095262 for ; Tue, 12 Apr 2005 05:01:22 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3CC1MUZ095261 for ietf-usefor-skb; Tue, 12 Apr 2005 05:01:22 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtp801.mail.ukl.yahoo.com (smtp801.mail.ukl.yahoo.com [217.12.12.138]) by above.proper.com (8.12.11/8.12.9) with SMTP id j3CC1FQI095153 for ; Tue, 12 Apr 2005 05:01:17 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from unknown (HELO host81-144-77-103.midband.mdip.bt.net) (ietf-usefor@imc.org@81.144.77.103 with poptime) by smtp801.mail.ukl.yahoo.com with SMTP; 12 Apr 2005 12:01:09 -0000 Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3CC0sr22211 for ietf-usefor@imc.org; Tue, 12 Apr 2005 13:00:54 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20664 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: Broken Message-ID syntax Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: <4255153E.3080500@isode.com> <42594879.3B92@xyzzy.claranet.de> <425A8A63.268A@xyzzy.claranet.de> Date: Tue, 12 Apr 2005 11:31:32 GMT Lines: 41 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In <425A8A63.268A@xyzzy.claranet.de> Frank Ellermann writes: >Charles Lindsey wrote: >> It prevents the msg-id >> <"foo.bar.baz"@example.com> >> as required >Yes, that's the idea: >Leading dot or trailing dot or unique-literal <=> quote >No leading/trailing dot and no unique-literal <=> don't quote >> it also prevents >> <"@.@"@example.com> >"@" is a "unique literal" => quote. Working as designed. >"@" is no atext => no dot-atom-text => unquoted is illegal. >Where's the problem ? The problem is that anything with an "@" in it (or various other weird characters) clearly MUST be quoted, but your syntax does not allow me to say <"@.@"@example.com> (though it would allow <"@..@"@example.com>). >In the proposed "unique" ABNF adjacent dots are handled by >".." in "unique-literal". Ugly like hell, but who cares if >it's correct. And I suspect it will be even uglier after the "@.@" bug is fixed :-( . But let us get a working correct syntax first, and then we can see whether there is a neater way to express it. -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 From owner-ietf-usefor@mail.imc.org Tue Apr 12 08:48:11 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id IAA22580 for ; Tue, 12 Apr 2005 08:02:55 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3CC1Jtq095236 for ; Tue, 12 Apr 2005 05:01:19 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3CC1JS4095235 for ietf-usefor-skb; Tue, 12 Apr 2005 05:01:19 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtp801.mail.ukl.yahoo.com (smtp801.mail.ukl.yahoo.com [217.12.12.138]) by above.proper.com (8.12.11/8.12.9) with SMTP id j3CC1Isa095179 for ; Tue, 12 Apr 2005 05:01:18 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from unknown (HELO host81-144-77-103.midband.mdip.bt.net) (ietf-usefor@imc.org@81.144.77.103 with poptime) by smtp801.mail.ukl.yahoo.com with SMTP; 12 Apr 2005 12:01:11 -0000 Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3CC0rq22205 for ietf-usefor@imc.org; Tue, 12 Apr 2005 13:00:53 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20663 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: Fwd: http://www.ietf.org/internet-drafts/draft-ietf-usefor-usefor-03.txt Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: <200504111212.j3BCChM14516@clerew.man.ac.uk> <87psx1gnjc.fsf@windlord.stanford.edu> Date: Tue, 12 Apr 2005 11:23:40 GMT Lines: 61 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In <87psx1gnjc.fsf@windlord.stanford.edu> Russ Allbery writes: >Charles Lindsey writes: >> "Dr John Stockton" writes: >>> "3.3.1 Lines" >>> This may not be needed for the transmission of News; but it is very >>> useful for Kill Rules. Therefore, IMHO it should be compulsory, not >>> obsolete (unless replaced by, say, 'Bytes'). >Kill Rules are based on overview information, not on article headers >except in unusual situations (since killing on full headers is very slow). >Overview contains its own line count that has nothing to do with an >article header; this is specified by the NNTP protocol. I think it is up to individual implementors of reading agents whether to provide kill file facilities on all headers, or only on those in the overview; some do, some don't. >You will find, if you analyze articles in the wild, that a significant >percentage of Lines headers are wrong. Indeed, but any killfile rule is likely to be expressed as "kill all articles where Lines > 1000", in which case absolute accuracy of the line count hardly matters. But I agree that there are better ways of killing overly long articles. >> I think the WG agreed early on that this header had passed its >> sell-by-date, though I do not think we intended it to be declared >> "obsolete" as the current Usefor draft implies. >I believe that we did. We intended it to be declared obsolete in the >sense that nothing should generate it and nothing should use it. The wording actually used, which has not changed since our very early drafts, was This header is to be regarded as obsolete, and it will likely be removed entirely in a future version of this standard. In the meantime, its use is deprecated. Which to me indicates "It is not obsolete yet, but it soon will be, so it is not really a good idea to use it". I think we did it that way because we wanted to document the correct way to do the count (because implementations regularly did it wrong), but there was still a widespread usage of it, which was likely to continue for some while. To my mind, the proper word to describe that situation is "obsolescent", rather than "obsolete". -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 From owner-ietf-usefor@mail.imc.org Tue Apr 12 12:02:10 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id MAA10381 for ; Tue, 12 Apr 2005 12:02:09 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3CG0elD031831 for ; Tue, 12 Apr 2005 09:00:40 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3CG0edL031830 for ietf-usefor-skb; Tue, 12 Apr 2005 09:00:40 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from rufus.isode.com (rufus.isode.com [62.3.217.251]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3CG0cUH031822 for ; Tue, 12 Apr 2005 09:00:39 -0700 (PDT) (envelope-from alexey.melnikov-usefor@isode.com) Received: from [172.16.2.185] (shiny.isode.com [62.3.217.250]) by rufus.isode.com via TCP (internal) with ESMTPA; Tue, 12 Apr 2005 17:00:35 +0100 Message-ID: <425BF0A3.3020708@isode.com> Date: Tue, 12 Apr 2005 17:00:35 +0100 From: Alexey Melnikov User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.7.2) Gecko/20040804 Netscape/7.2 (ax) X-Accept-Language: en-us, en MIME-Version: 1.0 To: Charles Lindsey CC: ietf-usefor@imc.org Subject: Re: Differences between RFC 2822 and Usefor References: In-Reply-To: Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Content-Transfer-Encoding: 7bit Charles Lindsey wrote: >>>3.2.1 References >>> >>> >>> The References header is the same as that specified in Section 3.6.4 >>> of [RFC2822] with the added restrictions detailed in Section 2.2 and >>> those listed below: >>> >>> >>Section 2.2 does not say anything about specific headers (and nothing at >>all about References), and the remaining part of 3.2.1 says nothing about >>followups and MUST. RFC2822 says "SHOULD". >> >> > >The text regarding the References header in the new usefor-03 draft is >just plain WRONG. > >Yes. the bit about section 2.2 covers the obligatory SP after the ':', as >required in all headers, and the "listed below" refers to the avoidance of >comments (MUST accept but do not generate yet). > >But the requirement for >the header to be present for followups has been omitted, and it needs to >be put back (modulo the relatively minor matter of the precise wording). > > No it is not, because this is not a USEFOR business to say anything about followups. Any given article doesn't become valid/invalid because of presence or lack of References. USEFOR doesn't deal with a thread of messages, this is a protocol issue. >And there is also an error in the syntax of the References header, which >Frank Ellermann pointed out some while back, and which needs to be >corrected. > This should be fixed or at least discussed. From owner-ietf-usefor@mail.imc.org Tue Apr 12 13:01:16 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id NAA16018 for ; Tue, 12 Apr 2005 13:01:14 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3CGxsKS037542 for ; Tue, 12 Apr 2005 09:59:54 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3CGxsQ2037541 for ietf-usefor-skb; Tue, 12 Apr 2005 09:59:54 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from mail1.panix.com (mail1.panix.com [166.84.1.72]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3CGxrQA037532 for ; Tue, 12 Apr 2005 09:59:54 -0700 (PDT) (envelope-from sethb@panix.com) Received: from panix5.panix.com (panix5.panix.com [166.84.1.5]) by mail1.panix.com (Postfix) with ESMTP id 936CA58B3B for ; Tue, 12 Apr 2005 12:59:52 -0400 (EDT) Received: (from sethb@localhost) by panix5.panix.com (8.11.6p3/8.8.8/PanixN1.1) id j3CGxq628085; Tue, 12 Apr 2005 12:59:52 -0400 (EDT) Date: Tue, 12 Apr 2005 12:59:52 -0400 (EDT) Message-Id: <200504121659.j3CGxq628085@panix5.panix.com> From: Seth Breidbart To: ietf-usefor@imc.org In-reply-to: (chl@clerew.man.ac.uk) Subject: Re: Path header delimiters References: Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: > Anyway, it a minor detail as regards scheme E, which seems to be > gaining added consenus. Does anyone else want to play? Sure. consensus++ Seth From owner-ietf-usefor@mail.imc.org Tue Apr 12 14:53:49 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id OAA25043 for ; Tue, 12 Apr 2005 14:53:49 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3CIqeQQ047274 for ; Tue, 12 Apr 2005 11:52:40 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3CIqefV047273 for ietf-usefor-skb; Tue, 12 Apr 2005 11:52:40 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtp3.Stanford.EDU (smtp3.Stanford.EDU [171.67.16.138]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3CIqd33047261 for ; Tue, 12 Apr 2005 11:52:39 -0700 (PDT) (envelope-from rra@stanford.edu) Received: from windlord.stanford.edu (windlord.Stanford.EDU [171.64.19.147]) by smtp3.Stanford.EDU (8.12.11/8.12.11) with SMTP id j3CIqc7b001663 for ; Tue, 12 Apr 2005 11:52:38 -0700 Received: (qmail 22549 invoked by uid 1000); 12 Apr 2005 18:52:38 -0000 To: Dr John Stockton Cc: ietf-usefor@imc.org Subject: Re: Fwd: http://www.ietf.org/internet-drafts/draft-ietf-usefor-usefor-03.txt In-Reply-To: <4XACZ2FxV$WCFwUM@merlyn.demon.co.uk> (John Stockton's message of "Tue, 12 Apr 2005 17:21:05 +0100") References: <200504111212.j3BCChM14516@clerew.man.ac.uk> <87psx1gnjc.fsf@windlord.stanford.edu> <4XACZ2FxV$WCFwUM@merlyn.demon.co.uk> From: Russ Allbery Organization: The Eyrie Date: Tue, 12 Apr 2005 11:52:37 -0700 Message-ID: <87oecjx1xm.fsf@windlord.stanford.edu> User-Agent: Gnus/5.1007 (Gnus v5.10.7) XEmacs/21.4 (Jumbo Shrimp, linux) MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: John Stockton writes: > I use Turnpike. It has kill-rules acting on article headers. With a > general rule using a RegExp one can kill or accept by a wide variety of > criteria. But in the controlling dialogue of the version that I use > there is specific provision for "Number of lines ... (as recorded in the > header)" - enter a number, rather than compose a RegExp - evidently, an > Internet Expert must have considered Lines to be useful information.. > I find this useful in newsgroups where a few users tend to post whole > web pages, computer programs, or essays. Yes. I'm not disagreeing about the usefulness of that sort of killfile rule. I'm saying that Turnpike should apply that rule against the overview information, which is guaranteed to be correct. You'll get exactly the same results, faster, and more reliably. If it's using the Lines header in the article, it's getting occasionally bogus results right now. The standard is intended to tell Turnpike and other news readers that do this to stop and use the overview information instead. Declaring this to be a bug in Turnpike is fully intentional. >> I believe that we did. We intended it to be declared obsolete in the >> sense that nothing should generate it and nothing should use it. > That will not meet with universal agreement. It generally does as soon as people understand what I mean when I say that a better version of that information is already in the overview data. -- Russ Allbery (rra@stanford.edu) From owner-ietf-usefor@mail.imc.org Tue Apr 12 15:01:01 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id PAA25487 for ; Tue, 12 Apr 2005 15:01:00 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3CIxVr1047677 for ; Tue, 12 Apr 2005 11:59:31 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3CIxV1x047676 for ietf-usefor-skb; Tue, 12 Apr 2005 11:59:31 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtp1.Stanford.EDU (smtp1.Stanford.EDU [171.67.16.123]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3CIxUIp047669 for ; Tue, 12 Apr 2005 11:59:30 -0700 (PDT) (envelope-from rra@stanford.edu) Received: from windlord.stanford.edu (windlord.Stanford.EDU [171.64.19.147]) by smtp1.Stanford.EDU (8.12.11/8.12.11) with SMTP id j3CIxTo8009174 for ; Tue, 12 Apr 2005 11:59:30 -0700 Received: (qmail 22722 invoked by uid 1000); 12 Apr 2005 18:59:29 -0000 To: ietf-usefor@imc.org Subject: Re: Fwd: http://www.ietf.org/internet-drafts/draft-ietf-usefor-usefor-03.txt In-Reply-To: (Charles Lindsey's message of "Tue, 12 Apr 2005 11:23:40 GMT") References: <200504111212.j3BCChM14516@clerew.man.ac.uk> <87psx1gnjc.fsf@windlord.stanford.edu> From: Russ Allbery Organization: The Eyrie Date: Tue, 12 Apr 2005 11:59:29 -0700 Message-ID: <87fyxvx1m6.fsf@windlord.stanford.edu> User-Agent: Gnus/5.1007 (Gnus v5.10.7) XEmacs/21.4 (Jumbo Shrimp, linux) MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Charles Lindsey writes: > Russ Allbery writes: >> Kill Rules are based on overview information, not on article headers >> except in unusual situations (since killing on full headers is very >> slow). Overview contains its own line count that has nothing to do >> with an article header; this is specified by the NNTP protocol. > I think it is up to individual implementors of reading agents whether to > provide kill file facilities on all headers, or only on those in the > overview; some do, some don't. Except for Lines. Lines should always be applied against overview. > Indeed, but any killfile rule is likely to be expressed as "kill all > articles where Lines > 1000", in which case absolute accuracy of the > line count hardly matters. But I agree that there are better ways of > killing overly long articles. Like EXACTLY THAT KILLFILE RULE, but applied against the guaranteed-accurate overview information. > The wording actually used, which has not changed since our very early > drafts, was > This header is to be regarded as obsolete, and it will likely be > removed entirely in a future version of this standard. In the > meantime, its use is deprecated. > Which to me indicates "It is not obsolete yet, but it soon will be, so > it is not really a good idea to use it". I think we did it that way > because we wanted to document the correct way to do the count (because > implementations regularly did it wrong), but there was still a > widespread usage of it, which was likely to continue for some while. > To my mind, the proper word to describe that situation is "obsolescent", > rather than "obsolete". I think I see what you're saying -- it's not obsolete in the sense that Path is obsolete (where news servers actually reject messages containing the header). I'm fine with the wording; I guess I don't really care whether that means "obsolete" or "obsolescent" as long as the key point is expressed: don't use the header and don't bother creating it. -- Russ Allbery (rra@stanford.edu) From owner-ietf-usefor@mail.imc.org Tue Apr 12 15:04:50 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id PAA25867 for ; Tue, 12 Apr 2005 15:04:49 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3CJ464C047876 for ; Tue, 12 Apr 2005 12:04:06 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3CJ466b047875 for ietf-usefor-skb; Tue, 12 Apr 2005 12:04:06 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtp1.Stanford.EDU (smtp1.Stanford.EDU [171.67.16.123]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3CJ45vX047868 for ; Tue, 12 Apr 2005 12:04:05 -0700 (PDT) (envelope-from rra@stanford.edu) Received: from windlord.stanford.edu (windlord.Stanford.EDU [171.64.19.147]) by smtp1.Stanford.EDU (8.12.11/8.12.11) with SMTP id j3CJ45Lt011240 for ; Tue, 12 Apr 2005 12:04:05 -0700 Received: (qmail 23592 invoked by uid 1000); 12 Apr 2005 19:04:04 -0000 To: ietf-usefor@imc.org Subject: Re: Path header delimiters In-Reply-To: (Charles Lindsey's message of "Tue, 12 Apr 2005 11:36:53 GMT") References: From: Russ Allbery Organization: The Eyrie Date: Tue, 12 Apr 2005 12:04:04 -0700 Message-ID: <87br8jx1ej.fsf@windlord.stanford.edu> User-Agent: Gnus/5.1007 (Gnus v5.10.7) XEmacs/21.4 (Jumbo Shrimp, linux) MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Charles Lindsey writes: > Anyway, it a minor detail as regards scheme E, which seems to be gaining > added consenus. Does anyone else want to play? Giving semantic meaning to a doubled delimiter makes me nervous, just because that's something one generally doesn't do in protocols of this type. But that's just a vague, untargetted concern and I can't think of concrete problems that it would solve. It's the best solution I've seen so far, so I'm happy to go with it unless someone comes up with something better. I do think that this is an important problem for us to solve. -- Russ Allbery (rra@stanford.edu) From owner-ietf-usefor@mail.imc.org Tue Apr 12 15:22:23 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id PAA28118 for ; Tue, 12 Apr 2005 15:22:23 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3CJLb6o048838 for ; Tue, 12 Apr 2005 12:21:37 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3CJLbw9048837 for ietf-usefor-skb; Tue, 12 Apr 2005 12:21:37 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from anchor-post-34.mail.demon.net (anchor-post-34.mail.demon.net [194.217.242.92]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3CJLakt048829 for ; Tue, 12 Apr 2005 12:21:36 -0700 (PDT) (envelope-from richard@highwayman.com) Received: from gti.noc.demon.net ([195.11.55.101] helo=happyday.al.cl.cam.ac.uk) by anchor-post-34.mail.demon.net with esmtp (Exim 4.42) id 1DLQx5-000MnJ-Cv; Tue, 12 Apr 2005 19:21:35 +0000 Message-ID: Date: Tue, 12 Apr 2005 20:20:10 +0100 To: Russ Allbery Cc: Dr John Stockton , ietf-usefor@imc.org From: Richard Clayton Subject: Re: Fwd: http://www.ietf.org/internet-drafts/draft-ietf-usefor-usefor-03.txt References: <200504111212.j3BCChM14516@clerew.man.ac.uk> <87psx1gnjc.fsf@windlord.stanford.edu> <4XACZ2FxV$WCFwUM@merlyn.demon.co.uk> <87oecjx1xm.fsf@windlord.stanford.edu> In-Reply-To: <87oecjx1xm.fsf@windlord.stanford.edu> MIME-Version: 1.0 X-Mailer: Turnpike Integrated Version 5.02 M Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 In message <87oecjx1xm.fsf@windlord.stanford.edu>, Russ Allbery writes > >John Stockton writes: > >> I use Turnpike. It has kill-rules acting on article headers. With a >> general rule using a RegExp one can kill or accept by a wide variety of >> criteria. But in the controlling dialogue of the version that I use >> there is specific provision for "Number of lines ... (as recorded in the >> header)" - enter a number, rather than compose a RegExp - evidently, an >> Internet Expert must have considered Lines to be useful information.. > >> I find this useful in newsgroups where a few users tend to post whole >> web pages, computer programs, or essays. > >Yes. I'm not disagreeing about the usefulness of that sort of killfile >rule. I'm saying that Turnpike should apply that rule against the >overview information, which is guaranteed to be correct. which it can already do :) .... assuming of course that its heuristics (and configuration) mean that it is indeed fetching the overview info and it isn't just using HEAD commands (that will always "work") ... in the real world, clients currently have to guess (or be told) what facilities to use; and tend to err in favour of features that are bound to be available... ... hence the value of NNTP v2 :) >If it's using the Lines header in the article, it's getting occasionally >bogus results right now. The standard is intended to tell Turnpike and >other news readers that do this to stop and use the overview information >instead. Declaring this to be a bug in Turnpike is fully intentional. John should add USEXOVER=YES to the appropriate [NEWS xxx] part of his configuration file :) and the "bug" will not bite him - -- richard @ highwayman . com "Nothing seems the same Still you never see the change from day to day And no-one notices the customs slip away" -----BEGIN PGP SIGNATURE----- Version: PGPsdk version 1.7.1 iQA/AwUBQlwfapoAxkTY1oPiEQJrRwCfW7Md/LVN74iV2bP1SbCfHaZmqs0An376 8uo4K/bZlJr1+0lljqx46KYU =smJq -----END PGP SIGNATURE----- From owner-ietf-usefor@mail.imc.org Tue Apr 12 20:01:03 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id UAA26868 for ; Tue, 12 Apr 2005 20:01:02 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3CNxlQ5066798 for ; Tue, 12 Apr 2005 16:59:47 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3CNxlYm066797 for ietf-usefor-skb; Tue, 12 Apr 2005 16:59:47 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from ciao.gmane.org (main.gmane.org [80.91.229.2]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3CNxiDr066783 for ; Tue, 12 Apr 2005 16:59:45 -0700 (PDT) (envelope-from usenet-format@gmane.org) Received: from list by ciao.gmane.org with local (Exim 4.43) id 1DLVEu-00086t-KU for ietf-usefor@imc.org; Wed, 13 Apr 2005 01:56:16 +0200 Received: from 212.82.251.36 ([212.82.251.36]) by main.gmane.org with esmtp (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Wed, 13 Apr 2005 01:56:16 +0200 Received: from nobody by 212.82.251.36 with local (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Wed, 13 Apr 2005 01:56:16 +0200 X-Injected-Via-Gmane: http://gmane.org/ To: ietf-usefor@imc.org From: Frank Ellermann Subject: Re: Broken Message-ID syntax Date: Wed, 13 Apr 2005 01:57:50 +0200 Organization: Lines: 47 Message-ID: <425C607E.1B5A@xyzzy.claranet.de> References: <4255153E.3080500@isode.com> <42594879.3B92@xyzzy.claranet.de> <425A8A63.268A@xyzzy.claranet.de> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Complaints-To: usenet@sea.gmane.org X-Gmane-NNTP-Posting-Host: 212.82.251.36 X-Mailer: Mozilla 3.0 (OS/2; U) Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Content-Transfer-Encoding: 7bit Charles Lindsey wrote: > anything with an "@" in it (or various other weird > characters) clearly MUST be quoted, ACK > your syntax does not allow me to say <"@.@"@example.com> NAK | msg-id = "<" unique "@" mdomain ">" | unique = dot-atom-text / ( DQUOTE unique-quote DQUOTE ) Skipping some details for your msg-id <"@.@"@example.com> I get example.com ~ mdomain "@.@" ~ ( DQUOTE unique-quote DQUOTE ) Modulo DQUOTE that's @.@ ~ unique-quote | unique-quote = ( "." [unique-part] ) / | ( [unique-part] "." ) / | ( [unique-part] unique-literal [unique-part] ) @.@ doesn't start with a dot, it doesn't end with a dot, so let's try: @.@ ~ [unique-part] unique-literal [unique-part] Module @ ~ unique-literal it can match in two different ways: @ ~ unique-literal @. ~ unique-part .@ ~ unique-part @ ~ unique-literal | unique-part = 1*( atext / "." / unique-literal ) Note the dot "." in unique-part, this should work as desiged. > though it would allow <"@..@"@example.com> Any number of dots, zero, one, two, or more. > it will be even uglier after the "@.@" bug is fixed :-( . First I want to see a bug before we make it uglier than it is. It's already ambiguous, it can't get worse, unless it's plain wrong. Bye, Frank From owner-ietf-usefor@mail.imc.org Wed Apr 13 12:13:54 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id MAA09766 for ; Wed, 13 Apr 2005 12:13:54 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3DGCbfZ072659 for ; Wed, 13 Apr 2005 09:12:37 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3DGCbPu072658 for ietf-usefor-skb; Wed, 13 Apr 2005 09:12:37 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtp803.mail.ukl.yahoo.com (smtp803.mail.ukl.yahoo.com [217.12.12.140]) by above.proper.com (8.12.11/8.12.9) with SMTP id j3DGCaOX072623 for ; Wed, 13 Apr 2005 09:12:37 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from unknown (HELO host81-144-75-78.midband.mdip.bt.net) (ietf-usefor@imc.org@81.144.75.78 with poptime) by smtp803.mail.ukl.yahoo.com with SMTP; 13 Apr 2005 16:12:30 -0000 Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3DGCDE00006 for ietf-usefor@imc.org; Wed, 13 Apr 2005 17:12:13 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20676 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: Differences between RFC 2822 and Usefor Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: <425BF0A3.3020708@isode.com> Date: Wed, 13 Apr 2005 15:59:00 GMT Lines: 55 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In <425BF0A3.3020708@isode.com> Alexey Melnikov writes: >Charles Lindsey wrote: >>The text regarding the References header in the new usefor-03 draft is >>just plain WRONG. >>But the requirement for >>the header to be present for followups has been omitted, and it needs to >>be put back (modulo the relatively minor matter of the precise wording). >> >> >No it is not, because this is not a USEFOR business to say anything >about followups. Any given article doesn't become valid/invalid because >of presence or lack of References. USEFOR doesn't deal with a thread of >messages, this is a protocol issue. Then why is RFC 2822 allowed to say that the References header SHOULD be present in the corresponding email situation? And why does the word "followup" occur 10 times in the Usefor draft? And why does the Usefor draft contain the words None of the headers appearing in this section is required to appear in every article but some of them are required in certain types of article, such as followups. And it says The References header is the same as that specified in Section 3.6.4 of [RFC2822] with the added restrictions detailed in Section 2.2 and those listed below: Which clearly implies that we have signed up to everything in RFC 2822 Section 3.6.4 except as detailed otherwise, which means that we have signed up to that "SHOULD", which we haven't. Netnews departs from RFC 2822 in various ways, which we document. Changing that "SHOULD" to a "MUST" is one of them (the identical SHOULD->MUST for Message-ID is another, but I do not see any objection to saying that). I understand that we intend to document these differences in an Appendix to Usefor, but how can we document differences that we are not allowed to describe? -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 From owner-ietf-usefor@mail.imc.org Wed Apr 13 12:13:55 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id MAA09764 for ; Wed, 13 Apr 2005 12:13:54 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3DGCaWA072649 for ; Wed, 13 Apr 2005 09:12:36 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3DGCaMI072648 for ietf-usefor-skb; Wed, 13 Apr 2005 09:12:36 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtp803.mail.ukl.yahoo.com (smtp803.mail.ukl.yahoo.com [217.12.12.140]) by above.proper.com (8.12.11/8.12.9) with SMTP id j3DGCZ2l072621 for ; Wed, 13 Apr 2005 09:12:35 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from unknown (HELO host81-144-75-78.midband.mdip.bt.net) (ietf-usefor@imc.org@81.144.75.78 with poptime) by smtp803.mail.ukl.yahoo.com with SMTP; 13 Apr 2005 16:12:29 -0000 Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3DGCBV29988 for ietf-usefor@imc.org; Wed, 13 Apr 2005 17:12:11 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20674 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: Path header delimiters Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: <87br8jx1ej.fsf@windlord.stanford.edu> Date: Wed, 13 Apr 2005 15:24:17 GMT Lines: 31 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In <87br8jx1ej.fsf@windlord.stanford.edu> Russ Allbery writes: >Charles Lindsey writes: >> Anyway, it a minor detail as regards scheme E, which seems to be gaining >> added consenus. Does anyone else want to play? >Giving semantic meaning to a doubled delimiter makes me nervous, just >because that's something one generally doesn't do in protocols of this >type. But that's just a vague, untargetted concern and I can't think of >concrete problems that it would solve. >It's the best solution I've seen so far, so I'm happy to go with it unless >someone comes up with something better. I do think that this is an >important problem for us to solve. OK, that makes it quite a smooth consensus compared to some we have had :-) . I shall write some text to incorporate it next week (no time at the moment - visiting grandchildren and a daughter getting married at the weekend). -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 From owner-ietf-usefor@mail.imc.org Wed Apr 13 13:03:25 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id MAA09765 for ; Wed, 13 Apr 2005 12:13:54 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3DGCXNP072632 for ; Wed, 13 Apr 2005 09:12:33 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3DGCX7A072631 for ietf-usefor-skb; Wed, 13 Apr 2005 09:12:33 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtp803.mail.ukl.yahoo.com (smtp803.mail.ukl.yahoo.com [217.12.12.140]) by above.proper.com (8.12.11/8.12.9) with SMTP id j3DGCVux072613 for ; Wed, 13 Apr 2005 09:12:32 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from unknown (HELO host81-144-75-78.midband.mdip.bt.net) (ietf-usefor@imc.org@81.144.75.78 with poptime) by smtp803.mail.ukl.yahoo.com with SMTP; 13 Apr 2005 16:12:25 -0000 Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3DGCCp29996 for ietf-usefor@imc.org; Wed, 13 Apr 2005 17:12:12 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20675 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: Fwd: http://www.ietf.org/internet-drafts/draft-ietf-usefor-usefor-03.txt Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: <200504111212.j3BCChM14516@clerew.man.ac.uk> <87psx1gnjc.fsf@windlord.stanford.edu> <87fyxvx1m6.fsf@windlord.stanford.edu> Date: Wed, 13 Apr 2005 15:27:14 GMT Lines: 21 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In <87fyxvx1m6.fsf@windlord.stanford.edu> Russ Allbery writes: >I think I see what you're saying -- it's not obsolete in the sense that >Path is obsolete (where news servers actually reject messages containing >the header). I'm fine with the wording; I guess I don't really care >whether that means "obsolete" or "obsolescent" as long as the key point is >expressed: don't use the header and don't bother creating it. Eh? Since when has Path been obsolete. Are you confusing it with some other ancient header? -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 From owner-ietf-usefor@mail.imc.org Wed Apr 13 13:03:26 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id MAA09763 for ; Wed, 13 Apr 2005 12:13:54 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3DGCZwv072641 for ; Wed, 13 Apr 2005 09:12:35 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3DGCZ7T072640 for ietf-usefor-skb; Wed, 13 Apr 2005 09:12:35 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtp803.mail.ukl.yahoo.com (smtp803.mail.ukl.yahoo.com [217.12.12.140]) by above.proper.com (8.12.11/8.12.9) with SMTP id j3DGCYCa072616 for ; Wed, 13 Apr 2005 09:12:34 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from unknown (HELO host81-144-75-78.midband.mdip.bt.net) (ietf-usefor@imc.org@81.144.75.78 with poptime) by smtp803.mail.ukl.yahoo.com with SMTP; 13 Apr 2005 16:12:28 -0000 Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3DGCDV00010 for ietf-usefor@imc.org; Wed, 13 Apr 2005 17:12:13 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20677 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: Broken Message-ID syntax Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: <4255153E.3080500@isode.com> <42594879.3B92@xyzzy.claranet.de> <425A8A63.268A@xyzzy.claranet.de> <425C607E.1B5A@xyzzy.claranet.de> Date: Wed, 13 Apr 2005 16:04:57 GMT Lines: 34 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In <425C607E.1B5A@xyzzy.claranet.de> Frank Ellermann writes: >Module @ ~ unique-literal it can match in two different ways: >@ ~ unique-literal @. ~ unique-part >.@ ~ unique-part @ ~ unique-literal >| unique-part = 1*( atext / "." / unique-literal ) >Note the dot "." in unique-part, this should work as desiged. Yes, you have convinced me now. So there remains the issue of the naming of the syntax rules, which I think needs a ruling from the Chair. Also, the wording needs to explain this additional problem (and also the reason for exlcusion of NO-WS-CONTROL). At the very least, the three examples given should be changed to: <"ab.cd"@example.com> <"ab.\cd"@example.com> -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 From owner-ietf-usefor@mail.imc.org Wed Apr 13 16:04:20 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id QAA29252 for ; Wed, 13 Apr 2005 16:04:19 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3DK2sfk086901 for ; Wed, 13 Apr 2005 13:02:54 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3DK2s7b086900 for ietf-usefor-skb; Wed, 13 Apr 2005 13:02:54 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtp3.Stanford.EDU (smtp3.Stanford.EDU [171.67.16.138]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3DK2sig086894 for ; Wed, 13 Apr 2005 13:02:54 -0700 (PDT) (envelope-from rra@stanford.edu) Received: from windlord.stanford.edu (windlord.Stanford.EDU [171.64.19.147]) by smtp3.Stanford.EDU (8.12.11/8.12.11) with SMTP id j3DK2r0q031092 for ; Wed, 13 Apr 2005 13:02:53 -0700 Received: (qmail 13475 invoked by uid 1000); 13 Apr 2005 20:02:53 -0000 To: ietf-usefor@imc.org Subject: Re: Fwd: http://www.ietf.org/internet-drafts/draft-ietf-usefor-usefor-03.txt In-Reply-To: (Charles Lindsey's message of "Wed, 13 Apr 2005 15:27:14 GMT") References: <200504111212.j3BCChM14516@clerew.man.ac.uk> <87psx1gnjc.fsf@windlord.stanford.edu> <87fyxvx1m6.fsf@windlord.stanford.edu> From: Russ Allbery Organization: The Eyrie Date: Wed, 13 Apr 2005 13:02:53 -0700 Message-ID: <87mzs2a1hu.fsf@windlord.stanford.edu> User-Agent: Gnus/5.1007 (Gnus v5.10.7) XEmacs/21.4 (Jumbo Shrimp, linux) MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Charles Lindsey writes: > Russ Allbery writes: >> I think I see what you're saying -- it's not obsolete in the sense that >> Path is obsolete (where news servers actually reject messages containing >> the header). I'm fine with the wording; I guess I don't really care >> whether that means "obsolete" or "obsolescent" as long as the key point is >> expressed: don't use the header and don't bother creating it. > Eh? Since when has Path been obsolete. Are you confusing it with some > other ancient header? Er, sorry, not Path, Received. Bleh. -- Russ Allbery (rra@stanford.edu) From owner-ietf-usefor@mail.imc.org Wed Apr 13 16:36:12 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id QAA05267 for ; Wed, 13 Apr 2005 16:36:12 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3DKYM4S089463 for ; Wed, 13 Apr 2005 13:34:22 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3DKYMFk089462 for ietf-usefor-skb; Wed, 13 Apr 2005 13:34:22 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from ciao.gmane.org (main.gmane.org [80.91.229.2]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3DKYKOs089453 for ; Wed, 13 Apr 2005 13:34:20 -0700 (PDT) (envelope-from usenet-format@gmane.org) Received: from list by ciao.gmane.org with local (Exim 4.43) id 1DLoUS-00012v-2F for ietf-usefor@imc.org; Wed, 13 Apr 2005 22:29:36 +0200 Received: from 212.82.251.217 ([212.82.251.217]) by main.gmane.org with esmtp (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Wed, 13 Apr 2005 22:29:36 +0200 Received: from nobody by 212.82.251.217 with local (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Wed, 13 Apr 2005 22:29:36 +0200 X-Injected-Via-Gmane: http://gmane.org/ To: ietf-usefor@imc.org From: Frank Ellermann Subject: Fixed (was: Broken Message-ID syntax) Date: Wed, 13 Apr 2005 22:25:21 +0200 Organization: Lines: 96 Message-ID: <425D8031.D57@xyzzy.claranet.de> References: <4255153E.3080500@isode.com> <42594879.3B92@xyzzy.claranet.de> <425A8A63.268A@xyzzy.claranet.de> <425C607E.1B5A@xyzzy.claranet.de> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Complaints-To: usenet@sea.gmane.org X-Gmane-NNTP-Posting-Host: 212.82.251.217 X-Mailer: Mozilla 3.0 (OS/2; U) Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Content-Transfer-Encoding: 7bit Charles Lindsey wrote: > Yes, you have convinced me now. Great, we have a msg-id syntax. > there remains the issue of the naming of the syntax rules, We're talking about seven names. So far we agree on msg-id, and we disagree on names for LHS, RHS, and address-literal. Maybe we can solve it for the remaining three names without bothering Alexey, Henry, or Ken: 1 - "unique-quote", the LHS if it must be quoted. If you hate "unique" a very traditional name would be "local". For a post-modern touch we could take "id". Maybe "quote" was too sloppy and "quoted" is better. Six possibilities: unique-quote local-quote id-quote unique-quoted local-quoted id-quoted 2 - "unique-part", the syntactically irrelevant stuff for the question "to quote or not to quote". Obviously we'd want the same prefix as in (1). Maybe "text" is better if we don't need it in (3): unique-part local-part id-part unique-text local-text id-text 3 - "unique-literal", something that requires quoting. The "literal" is somewhat dubious, 2822 and usefor-03 use it for the RHS, but here we are in the LHS Let's say that "literal" was a bad idea. Depending on (2) "text" could be fine. I also like "special" (as in your "mqspecial"): unique-text local-text id-text unique-special local-special id-special JFTR the three names where we are unable to find a compromise: 4 - "address-literal" vs. "no-fold-literal". You want the same name as in 2822 with a different syntax. For this reason I want a different name reflecting the semantics. We've established that the semantics is a "domain-literal" in RfC 2822, but that's already used for an 2822-construct with CFWS and FWS. A related source is RfC 2821, there the name is "address-literal". IMHO a typical case of "if it quacks like a duck...". 5 - "mdomain" vs. "id-right". As in (4), you want the name as in RfC 2822 with a different syntax. I want a name that reflects the semantics "domain". We've spent hours, days, and weeks with the prolem of the RHS in a Message-ID, because you said that it's no domain. I've shown that that's _not_ what 2822 means. AFAIK Bruce also disagrees with you. STD 11, RfC 1036, and s-o-1036 disagree with you. Abusing foreign namespaces is a serious problem. If you don't like "mdomain" you could pick "msg-domain" or "id-domain". But "id-right" is OUT as far as I'm concerned. 6 - "unique" vs. "id-left". A similar case as in (5). Maybe you feel that "unique" is too traditional, and "local" is even more traditional. OTOH it's true, so we could only beautify it a bit: msg-unique id-unique msg-local id-local We can use the prefix "id-" in (6), (5), (3), (2), and (1). Or the prefix "msg-" in (6), (5), we already have it in msg-id. > which I think needs a ruling from the Chair. Alexey as King Canute decreeing that the RHS is NOT a domain ? I don't think so. > the wording needs to explain this additional problem Not necessarily, leading / trailing / adjacent dots aren't new, it's exactly the same problem as in RfC 2822. > also the reason for exlcusion of NO-WS-CONTROL Maybe, but please state them all: It was never used, it never worked, it would be a PITA in news-URLs, it wasn't allowed in any news standard. More than one line of text, so that we have a line between these lines saying "2822 GOT IT WRONG - PERIOD." Bye, Frank From owner-ietf-usefor@mail.imc.org Wed Apr 13 17:23:35 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id RAA08347 for ; Wed, 13 Apr 2005 17:23:35 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3DLM9la096060 for ; Wed, 13 Apr 2005 14:22:09 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3DLM9vi096059 for ietf-usefor-skb; Wed, 13 Apr 2005 14:22:09 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from a.mail.peak.org (a.mail.peak.org [69.59.192.41]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3DLM8NS096047 for ; Wed, 13 Apr 2005 14:22:09 -0700 (PDT) (envelope-from stanley@peak.org) Received: from a.shell.peak.org ([69.59.192.81]) by a.mail.peak.org (8.12.10/8.12.8) with ESMTP id j3DLM2mO051869 (version=TLSv1/SSLv3 cipher=EDH-RSA-DES-CBC3-SHA bits=168 verify=NO) for ; Wed, 13 Apr 2005 14:22:03 -0700 (PDT) Date: Wed, 13 Apr 2005 14:22:02 -0700 (PDT) From: John Stanley X-X-Sender: stanley@a.shell.peak.org To: ietf-usefor@imc.org Subject: Re: Differences between RFC 2822 and Usefor (fwd) Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Spam-Score: 0 () X-Scanned-By: MIMEDefang 2.39 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Third attempt at one message. This "open list" does not seem to be accepting messages ... "Charles Lindsey" : >The "SHOULD" in question is in the first paragraph of section 3.6.4 of RFC >2822. I know where the SHOULD comes from. RFC2822 isn't "in question", it is an existing standard. The question is where you see a MUST in USEFOR. I don't see one. It simply isn't there. >The "MUST" in question arises from the long held consensus in this Working >Group that the References header in News is NOT an "optional extra". Not according to draft-usefor-03. According to draft-usefor-03, References is defined to be the same as in RFC2822, with the addition of a few bits that are not mandates for using it. This consensus you claim exists has a very odd way of being expressed in our products. Almost as if it doesn't really exist. The fact that I am the only one who even commented on it being gone is a pretty good sign that nobody cares that it is gone. >And you yourself have been at the forefront in upholding that consensus. Hardly. I've repeatedly said "IF we are going to say that References qualifies for RFC2119 mandates, THEN we should not water it down by saying that non-followups MAY contain them." That first clause is called a "conditional". If we DON'T say that References qualifies for such a mandate (and I've argued that it is hard to justify one, if you recall), then the situation is different. Like it is now. We currently do NOT say it is mandatory. Once we say that, the rest of the argument about WHEN it is mandatory goes away. >That clear requirement has been in all our drafts up to article-13. It is no longer a requirement at all in our draft. As you have been quick to tell me when I object to changes YOU make, things change. Our current draft does NOT make it a requirement; it is lunacy to claim that this lack of mandate reflects a "consensus" that there is a mandate, or that the complete lack of comment about the change (other than your expressed belief that a missing mandate is still a mandate) is a better sign of consensus than your word for it. At this point, the mandate is not part of the draft, and putting one in would (or at least, should, in an honest system) require a strong demonstration of interoperability issues. Since news systems simply do not care if the header is in the article or not, there is no way to demonstrate an interoperability issue, and thus RFC2119 language is not appropriate. I've accepted the change, why is the editor whose name is on it as author trying to claim it doesn't exist? >The text regarding the References header in the new usefor-03 draft is >just plain WRONG. Says you. If it was wrong, it should not have been sent off as our work product without a single person here seeing it. Well, ok, maybe someone here did see it, but it sure wasn't announced for review prior to it becoming official. >But the requirement for >the header to be present for followups has been omitted, and it needs to >be put back (modulo the relatively minor matter of the precise wording). You need to PROVE that NEED before you make a change to the draft. You saying "we need this" doesn't meet the requirements of RFC2119 and you know it. Were that all it takes, then we would be documenting PAM and MCT and making THEM mandatory. But, no. It doesn't work that way. You're taking all mention of those headers OUT of the drafts that still refer to them. >And there is also an error in the syntax of the References header, which >Frank Ellermann pointed out some while back, and which needs to be >corrected. Please stop trying to equate a syntax error with a change you want that has nothing to do with syntax. Making References optional (in accordance with RFC2822) is a different kind of change, and it's been made. Making it MANDATORY needs strong justification, and I'll await your attempts at same. "Because it was that way before" doesn't work when I object to a change you decide to make, don't pretend that you can use it as an excuse. From owner-ietf-usefor@mail.imc.org Wed Apr 13 17:41:09 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id RAA09381 for ; Wed, 13 Apr 2005 17:41:08 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3DLdldE097279 for ; Wed, 13 Apr 2005 14:39:47 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3DLdliX097278 for ietf-usefor-skb; Wed, 13 Apr 2005 14:39:47 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from b.mail.peak.org (b.mail.peak.org [69.59.192.42]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3DLdkD5097271 for ; Wed, 13 Apr 2005 14:39:46 -0700 (PDT) (envelope-from stanley@peak.org) Received: from a.shell.peak.org ([69.59.192.81]) by b.mail.peak.org (8.12.10/8.12.8) with ESMTP id j3DLde0j001422 (version=TLSv1/SSLv3 cipher=EDH-RSA-DES-CBC3-SHA bits=168 verify=NO) for ; Wed, 13 Apr 2005 14:39:41 -0700 (PDT) Date: Wed, 13 Apr 2005 14:39:40 -0700 (PDT) From: John Stanley X-X-Sender: stanley@a.shell.peak.org To: ietf-usefor@imc.org Subject: Re: Differences between RFC 2822 and Usefor Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Spam-Score: 0 () X-Scanned-By: MIMEDefang 2.39 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: "Charles Lindsey" : >And why does the Usefor draft contain the words > None of the headers appearing in this section is required to appear > in every article but some of them are required in certain types of > article, such as followups. Because someone (an editor, perhaps?) overlooked this statement when removing the mandate for References? Remove the "such as" clause and the problem is fixed. Do it and lets move on. Problem solved. >Which clearly implies that we have signed up to everything in RFC 2822 >Section 3.6.4 except as detailed otherwise, which means that we have >signed up to that "SHOULD", which we haven't. Clearly the text in usefor-03 says we HAVE signed up for RFC2822, since it says that. "as defined in ... with these limited exceptions" means "as defined in" with a few exceptions, none of which is a mandate for use in any kind of article, only mandates about what it contains when it is used. >Netnews departs from RFC 2822 in various ways, which we document. Yep. And one of those ways is no longer a mandatory References header in this nebulous "followup" thing. Undocumented "MUSTS" are not MUSTS. "Charles says..." is not RFC2119 justification. The side effect is now that THIS requirement is gone, there is no way of identifying what is and is not a followup no matter WHAT definition of followup may be used, so there is no reason to argue about what is and is not a followup anymore. If you think that not being able to identify what is a followup isn't important (i.e., it's ok for non-followups to contain References headers) then you cannot possibly think it is important enough to identify followups that the header designed for that purpose has to be mandatory. Be consistent -- either it's important or it isn't. Don't think that the mandate can be slipped back into USEPRO, since any mandate in USEPRO would contradict USEFOR, and that would look stupid. If USEFOR says it's optional, then USEPRO cannot claim that it is an interoperability issue and is thus not optional. From owner-ietf-usefor@mail.imc.org Mon Apr 18 00:06:13 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id AAA18406 for ; Mon, 18 Apr 2005 00:06:12 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3I44L7o039856 for ; Sun, 17 Apr 2005 21:04:21 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3I44KKC039855 for ietf-usefor-skb; Sun, 17 Apr 2005 21:04:20 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from spsystems.net (spsystems.net [216.126.83.115]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3I44K0A039848 for ; Sun, 17 Apr 2005 21:04:20 -0700 (PDT) (envelope-from henry@spsystems.net) Received: from spsystems.net (henry@localhost [127.0.0.1]) by spsystems.net (8.12.10/8.12.10) with ESMTP id j3I43vVO001814; Mon, 18 Apr 2005 00:03:57 -0400 (EDT) Received: (from henry@localhost) by spsystems.net (8.12.10/8.12.10/Submit) id j3I43pU4001813; Mon, 18 Apr 2005 00:03:51 -0400 (EDT) Date: Mon, 18 Apr 2005 00:03:51 -0400 (EDT) From: Henry Spencer To: Usefor Mailing List Subject: Re: Path header delimiters In-Reply-To: <425AD6EF.448A@xyzzy.claranet.de> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: On Mon, 11 Apr 2005, Frank Ellermann wrote: > More fun: Any news about publishing s-o-1036 as RfC 4036 ? Despite good intentions, I've been distracted and haven't gotten to the point of making a submission. Early next month, I hope. (No chance between now and then, too much to do.) > Less fun: Charles and I need a volunteer for some kind of > "rough consensus" about the names (sic!) of some msg-id ABNF > productions, could you please toss a coin or add a comment ? I've been hoping that stuff would sort itself out, because I haven't been paying enough attention to have an opinion on it. Henry Spencer henry@spsystems.net From owner-ietf-usefor@mail.imc.org Mon Apr 18 06:03:57 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id GAA02451 for ; Mon, 18 Apr 2005 06:03:57 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3IA1MCd078552 for ; Mon, 18 Apr 2005 03:01:22 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3IA1Mf7078551 for ietf-usefor-skb; Mon, 18 Apr 2005 03:01:22 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from rufus.isode.com (rufus.isode.com [62.3.217.251]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3IA1JCi078517 for ; Mon, 18 Apr 2005 03:01:20 -0700 (PDT) (envelope-from alexey.melnikov-usefor@isode.com) Received: from [192.168.0.3] ([62.3.217.253]) by rufus.isode.com via TCP (internal) with ESMTPA; Mon, 18 Apr 2005 11:01:00 +0100 Message-ID: <4262CF9B.40207@isode.com> Date: Sun, 17 Apr 2005 22:05:31 +0100 From: Alexey Melnikov User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.7.2) Gecko/20040804 Netscape/7.2 (ax) X-Accept-Language: en-us, en To: Charles Lindsey CC: ietf-usefor@imc.org Subject: Re: Differences between RFC 2822 and Usefor References: <425BF0A3.3020708@isode.com> In-Reply-To: MIME-version: 1.0 Content-type: text/plain; charset="ISO-8859-1"; format="flowed" Content-transfer-encoding: 7bit Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Content-Transfer-Encoding: 7bit Charles Lindsey wrote: >In <425BF0A3.3020708@isode.com> Alexey Melnikov writes: > >>Charles Lindsey wrote: >> >> >>>The text regarding the References header in the new usefor-03 draft is >>>just plain WRONG. >>> >>> >>>But the requirement for >>>the header to be present for followups has been omitted, and it needs to >>>be put back (modulo the relatively minor matter of the precise wording). >>> >>> >>No it is not, because this is not a USEFOR business to say anything >>about followups. Any given article doesn't become valid/invalid because >>of presence or lack of References. USEFOR doesn't deal with a thread of >>messages, this is a protocol issue. >> >> > >Then why is RFC 2822 allowed to say that the References header SHOULD be >present in the corresponding email situation? > Because RFC 2822 is very much like a combined USEFOR+USEPRO. All I am trying to say is that USEFOR is not the proper place to state any requirement that doesn't apply to all articles. >And why does the word "followup" occur 10 times in the Usefor draft? > 6 of them are used when defining different terms, 3 are used when describing Followup-to (naturally), 1 is used in the sentence you've quoted below. IMHO, this is quite reasonable. I would like to avoid any use of the term in USEFOR, but this is just not practical. >And why does the Usefor draft contain the words > > None of the headers appearing in this section is required to appear > in every article but some of them are required in certain types of > article, such as followups. > Which is also followed by the following sentence: Further discussion of these requirements appears in [USEPRO] and [USEAGE]. This is a clear indication that USEPRO is a proper place to define any additional requirements. >And it says > > The References header is the same as that specified in Section 3.6.4 > of [RFC2822] with the added restrictions detailed in Section 2.2 and > those listed below: > >Which clearly implies that we have signed up to everything in RFC 2822 >Section 3.6.4 except as detailed otherwise, which means that we have >signed up to that "SHOULD", which we haven't. > > You can add it to USEPRO to "Duties of a Followup Agent" as per above. >Netnews departs from RFC 2822 in various ways, which we document. Changing >that "SHOULD" to a "MUST" is one of them (the identical SHOULD->MUST for >Message-ID is another, but I do not see any objection to saying that). >I understand that we intend to document these differences in an Appendix >to Usefor, but how can we document differences that we are not allowed to >describe? > > From owner-ietf-usefor@mail.imc.org Wed Apr 20 18:58:28 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id SAA02930 for ; Wed, 20 Apr 2005 18:58:27 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3KMvHnL078173 for ; Wed, 20 Apr 2005 15:57:17 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3KMvHhi078172 for ietf-usefor-skb; Wed, 20 Apr 2005 15:57:17 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from probity.mcc.ac.uk (probity.mcc.ac.uk [130.88.200.94]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3KMvG6W078160 for ; Wed, 20 Apr 2005 15:57:16 -0700 (PDT) (envelope-from chl@clerew.man.ac.uk) Received: from dialup027.mcc.ac.uk ([130.88.69.27]) by probity.mcc.ac.uk with esmtp (Exim 4.20) id 1DOO8A-000CWp-UH for ietf-usefor@imc.org; Wed, 20 Apr 2005 23:57:15 +0100 Received: (from chl@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3KMZAq15354; Wed, 20 Apr 2005 23:35:10 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20689 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: Differences between RFC 2822 and Usefor Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: Date: Wed, 20 Apr 2005 20:06:14 GMT Lines: 30 MIME-Version: 1.0 X-Scanner: exiscan for exim4 (http://duncanthrax.net/exiscan/) *1DOO8A-000CWp-UH*Hqf6Bbhi9qw* Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In John Stanley writes: >The side effect is now that THIS requirement is gone, there is no way of >identifying what is and is not a followup no matter WHAT definition of >followup may be used, so there is no reason to argue about what is and is >not a followup anymore. If you think that not being able to identify what >is a followup isn't important (i.e., it's ok for non-followups to contain >References headers) then you cannot possibly think it is important enough >to identify followups that the header designed for that purpose has to be >mandatory. Be consistent -- either it's important or it isn't. No, the requirement has not gone. >Don't think that the mandate can be slipped back into USEPRO, since any >mandate in USEPRO would contradict USEFOR, and that would look stupid. If >USEFOR says it's optional, then USEPRO cannot claim that it is an >interoperability issue and is thus not optional. The mandate in question is already in USEPRO. And always has been. -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 From owner-ietf-usefor@mail.imc.org Wed Apr 20 18:58:29 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id SAA02948 for ; Wed, 20 Apr 2005 18:58:29 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3KMvI8Z078193 for ; Wed, 20 Apr 2005 15:57:18 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3KMvITF078192 for ietf-usefor-skb; Wed, 20 Apr 2005 15:57:18 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from probity.mcc.ac.uk (probity.mcc.ac.uk [130.88.200.94]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3KMvHNl078183 for ; Wed, 20 Apr 2005 15:57:17 -0700 (PDT) (envelope-from chl@clerew.man.ac.uk) Received: from dialup027.mcc.ac.uk ([130.88.69.27]) by probity.mcc.ac.uk with esmtp (Exim 4.20) id 1DOO8C-000CWp-FD for ietf-usefor@imc.org; Wed, 20 Apr 2005 23:57:16 +0100 Received: (from chl@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3KMXf015340; Wed, 20 Apr 2005 23:33:41 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20687 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: Fwd: http://www.ietf.org/internet-drafts/draft-ietf-usefor-usefor-03.txt Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: <200504111212.j3BCChM14516@clerew.man.ac.uk> <87psx1gnjc.fsf@windlord.stanford.edu> <87fyxvx1m6.fsf@windlord.stanford.edu> <87mzs2a1hu.fsf@windlord.stanford.edu> Date: Wed, 20 Apr 2005 19:45:12 GMT Lines: 32 MIME-Version: 1.0 X-Scanner: exiscan for exim4 (http://duncanthrax.net/exiscan/) *1DOO8C-000CWp-FD*4wwlrdD5M7M* Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In <87mzs2a1hu.fsf@windlord.stanford.edu> Russ Allbery writes: >Charles Lindsey writes: >> Russ Allbery writes: >>> I think I see what you're saying -- it's not obsolete in the sense that >>> Path is obsolete (where news servers actually reject messages containing >>> the header). I'm fine with the wording; I guess I don't really care >>> whether that means "obsolete" or "obsolescent" as long as the key point is >>> expressed: don't use the header and don't bother creating it. >> Eh? Since when has Path been obsolete. Are you confusing it with some >> other ancient header? >Er, sorry, not Path, Received. Bleh. Ah! But in that case I don't think servers should be rejecting such headers. If the article has been wandering in any out of both email and news systems, then it is sometimes useful to have the full history recorded (though I might agree that if there was a mailing list involved I might not want to know what happened before it reached the mailing list). -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 From owner-ietf-usefor@mail.imc.org Wed Apr 20 18:58:51 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id SAA02977 for ; Wed, 20 Apr 2005 18:58:51 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3KMvJ97078201 for ; Wed, 20 Apr 2005 15:57:19 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3KMvJIQ078200 for ietf-usefor-skb; Wed, 20 Apr 2005 15:57:19 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from probity.mcc.ac.uk (probity.mcc.ac.uk [130.88.200.94]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3KMvINk078191 for ; Wed, 20 Apr 2005 15:57:18 -0700 (PDT) (envelope-from chl@clerew.man.ac.uk) Received: from dialup027.mcc.ac.uk ([130.88.69.27]) by probity.mcc.ac.uk with esmtp (Exim 4.20) id 1DOO8D-000CWp-FM for ietf-usefor@imc.org; Wed, 20 Apr 2005 23:57:17 +0100 Received: (from chl@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3KMWsv15334; Wed, 20 Apr 2005 23:32:54 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20686 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: Fixed (was: Broken Message-ID syntax) Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: <4255153E.3080500@isode.com> <42594879.3B92@xyzzy.claranet.de> <425A8A63.268A@xyzzy.claranet.de> <425C607E.1B5A@xyzzy.claranet.de> <425D8031.D57@xyzzy.claranet.de> Date: Wed, 20 Apr 2005 19:42:16 GMT Lines: 38 MIME-Version: 1.0 X-Scanner: exiscan for exim4 (http://duncanthrax.net/exiscan/) *1DOO8D-000CWp-FM*QJF8NzufRBI* Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In <425D8031.D57@xyzzy.claranet.de> Frank Ellermann writes: >Charles Lindsey wrote: >> Yes, you have convinced me now. >Great, we have a msg-id syntax. >> there remains the issue of the naming of the syntax rules, >We're talking about seven names. So far we agree on msg-id, >and we disagree on names for LHS, RHS, and address-literal. >Maybe we can solve it for the remaining three names without >bothering Alexey, Henry, or Ken: Essentially, we either use the same names as RFC 2822 uses, saying that "the following syntax rules replace the corresponding rules in RFC 2822" (giving us id-left, id-right, no-fold-quote, no-fold literal, etc), or else we depart from the RFC 2822 names entirely, except for the msg-id at the top, in which case your names are as good as any. I don't see that there is anything in-between. My preference is to retain the RFC 2822 naming, which accords with the general principle that we do not change anything in RFC 2822 that we do not need to change. If our Chair says that we can change these names, then let it be so. Alexey? -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 From owner-ietf-usefor@mail.imc.org Wed Apr 20 19:11:45 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id TAA03900 for ; Wed, 20 Apr 2005 19:11:44 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3KMvHYw078182 for ; Wed, 20 Apr 2005 15:57:17 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3KMvHcP078181 for ietf-usefor-skb; Wed, 20 Apr 2005 15:57:17 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from probity.mcc.ac.uk (probity.mcc.ac.uk [130.88.200.94]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3KMvGsY078161 for ; Wed, 20 Apr 2005 15:57:17 -0700 (PDT) (envelope-from chl@clerew.man.ac.uk) Received: from dialup027.mcc.ac.uk ([130.88.69.27]) by probity.mcc.ac.uk with esmtp (Exim 4.20) id 1DOO8B-000CWp-KZ for ietf-usefor@imc.org; Wed, 20 Apr 2005 23:57:15 +0100 Received: (from chl@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3KMYVY15348; Wed, 20 Apr 2005 23:34:31 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20690 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: Differences between RFC 2822 and Usefor Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: <425BF0A3.3020708@isode.com> <4262CF9B.40207@isode.com> Date: Wed, 20 Apr 2005 20:12:34 GMT Lines: 41 MIME-Version: 1.0 X-Scanner: exiscan for exim4 (http://duncanthrax.net/exiscan/) *1DOO8B-000CWp-KZ*E4nNcJZLgk2* Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In <4262CF9B.40207@isode.com> Alexey Melnikov writes: >Charles Lindsey wrote: >>And why does the Usefor draft contain the words >> >> None of the headers appearing in this section is required to appear >> in every article but some of them are required in certain types of >> article, such as followups. >> >Which is also followed by the following sentence: > Further discussion of these requirements appears in [USEPRO] >and [USEAGE]. Which sentence should not be there IMHO. >This is a clear indication that USEPRO is a proper place to define any >additional requirements. >You can add it to USEPRO to "Duties of a Followup Agent" as per above. Yes, that is possible. Indeed it is already there, but might need a little more saying about it there, including pointing out the divergence from RFC 2822 and something about using it in non-followup situations such as mulipart FAQs and message/partial. Incidentally, Usefor also needs the semantics of this header explaining (such as the significance attaching to the order in which the precursors are listed), so maybe the mention of multipart FAQs and message/partial could be slipped in there. -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 From owner-ietf-usefor@mail.imc.org Wed Apr 20 19:14:06 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id TAA04293 for ; Wed, 20 Apr 2005 19:14:06 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3KMvLQN078219 for ; Wed, 20 Apr 2005 15:57:21 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3KMvL9L078218 for ietf-usefor-skb; Wed, 20 Apr 2005 15:57:21 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from probity.mcc.ac.uk (probity.mcc.ac.uk [130.88.200.94]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3KMvKiY078211 for ; Wed, 20 Apr 2005 15:57:20 -0700 (PDT) (envelope-from chl@clerew.man.ac.uk) Received: from dialup027.mcc.ac.uk ([130.88.69.27]) by probity.mcc.ac.uk with esmtp (Exim 4.20) id 1DOO8F-000CWp-Dk for ietf-usefor@imc.org; Wed, 20 Apr 2005 23:57:19 +0100 Received: (from chl@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3KMUe715315; Wed, 20 Apr 2005 23:30:40 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20688 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: Differences between RFC 2822 and Usefor (fwd) Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: Date: Wed, 20 Apr 2005 20:03:27 GMT Lines: 60 MIME-Version: 1.0 X-Scanner: exiscan for exim4 (http://duncanthrax.net/exiscan/) *1DOO8F-000CWp-Dk*h0tR2iM6i9I* Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In John Stanley writes: >"Charles Lindsey" : >>The "MUST" in question arises from the long held consensus in this Working >>Group that the References header in News is NOT an "optional extra". >Not according to draft-usefor-03. According to draft-usefor-03, >References is defined to be the same as in RFC2822, with the addition of a >few bits that are not mandates for using it. This consensus you claim >exists has a very odd way of being expressed in our products. Almost as if >it doesn't really exist. >The fact that I am the only one who even commented on it being gone is a >pretty good sign that nobody cares that it is gone. No, I knew that it wasn't there, and there are a lot of other things that should be in the Usefor draft and that have not made it there yet. But I do not want to be seen arguing with my co-editor in public if it can be avoided. Especially on issues that do not appear to be the subject of controversy. >>That clear requirement has been in all our drafts up to article-13. >It is no longer a requirement at all in our draft. Indeed, but a draft is only a draft. There has been no WG decision to change the requirement for a References header in followups. Our Chair has taken the view that Usefor is not the place to say it, but that Usepro IS the proper place (and indeed there is already a corresponding MUST in Usepro, though it may need a little more amplification). I do not agree with the Chair on this one, but if nobody else speaks up and if his ruling stands, then Usepro is where it will be said (which also, incidentally, removes most of the cause of the disagreement there has been about how to word it all). But in that case, I will have to find somewhere in Usepro to say that the References header MAY also be used in other cases such as multipart FAQs, as well as pointing out the divergence from the SHOULD in RFC 2822. >At this point, the mandate is not part of the draft, and putting one in >would (or at least, should, in an honest system) require a strong >demonstration of interoperability issues. Since news systems simply do not >care if the header is in the article or not, there is no way to >demonstrate an interoperability issue, and thus RFC2119 language is not >appropriate. News readers that try to do threading care very much if it is missing. -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 From owner-ietf-usefor@mail.imc.org Wed Apr 20 19:47:10 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id TAA09537 for ; Wed, 20 Apr 2005 19:47:10 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3KNij22088559 for ; Wed, 20 Apr 2005 16:44:45 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3KNijiM088558 for ietf-usefor-skb; Wed, 20 Apr 2005 16:44:45 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtp2.Stanford.EDU (smtp2.Stanford.EDU [171.67.16.125]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3KNiiND088549 for ; Wed, 20 Apr 2005 16:44:45 -0700 (PDT) (envelope-from rra@stanford.edu) Received: from windlord.stanford.edu (windlord.Stanford.EDU [171.64.19.147]) by smtp2.Stanford.EDU (8.12.11/8.12.11) with SMTP id j3KNiirY017262 for ; Wed, 20 Apr 2005 16:44:44 -0700 Received: (qmail 7383 invoked by uid 1000); 20 Apr 2005 23:44:43 -0000 To: ietf-usefor@imc.org Subject: Re: Fwd: http://www.ietf.org/internet-drafts/draft-ietf-usefor-usefor-03.txt In-Reply-To: (Charles Lindsey's message of "Wed, 20 Apr 2005 19:45:12 GMT") References: <200504111212.j3BCChM14516@clerew.man.ac.uk> <87psx1gnjc.fsf@windlord.stanford.edu> <87fyxvx1m6.fsf@windlord.stanford.edu> <87mzs2a1hu.fsf@windlord.stanford.edu> From: Russ Allbery Organization: The Eyrie Date: Wed, 20 Apr 2005 16:44:43 -0700 Message-ID: <87wtqxui6s.fsf@windlord.stanford.edu> User-Agent: Gnus/5.1007 (Gnus v5.10.7) XEmacs/21.4 (Jumbo Shrimp, linux) MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Charles Lindsey writes: > Ah! But in that case I don't think servers should be rejecting such > headers. If the article has been wandering in any out of both email and > news systems, then it is sometimes useful to have the full history > recorded (though I might agree that if there was a mailing list involved > I might not want to know what happened before it reached the mailing > list). Then I don't understand the distinction you're drawing between obsolete and obsolescent. As near as I could tell, the only difference in practice was that servers were rejecting the former headers, but now you're saying that they shouldn't even do that. -- Russ Allbery (rra@stanford.edu) From owner-ietf-usefor@mail.imc.org Wed Apr 20 20:16:56 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id UAA12241 for ; Wed, 20 Apr 2005 20:16:55 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3L0ESqY095750 for ; Wed, 20 Apr 2005 17:14:28 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3L0ESxh095749 for ietf-usefor-skb; Wed, 20 Apr 2005 17:14:28 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from ciao.gmane.org (main.gmane.org [80.91.229.2]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3L0EQX0095742 for ; Wed, 20 Apr 2005 17:14:27 -0700 (PDT) (envelope-from usenet-format@gmane.org) Received: from list by ciao.gmane.org with local (Exim 4.43) id 1DOPGU-0002Cv-QZ for ietf-usefor@imc.org; Thu, 21 Apr 2005 02:09:54 +0200 Received: from 212.82.251.239 ([212.82.251.239]) by main.gmane.org with esmtp (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Thu, 21 Apr 2005 02:09:54 +0200 Received: from nobody by 212.82.251.239 with local (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Thu, 21 Apr 2005 02:09:54 +0200 X-Injected-Via-Gmane: http://gmane.org/ To: ietf-usefor@imc.org From: Frank Ellermann Subject: References (was: Differences between RFC 2822 and Usefor (fwd)) Date: Thu, 21 Apr 2005 02:09:25 +0200 Organization: Lines: 46 Message-ID: <4266EF35.21A5@xyzzy.claranet.de> References: Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Complaints-To: usenet@sea.gmane.org X-Gmane-NNTP-Posting-Host: 212.82.251.239 X-Mailer: Mozilla 3.0 (OS/2; U) Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Content-Transfer-Encoding: 7bit Charles Lindsey wrote: > I do not want to be seen arguing with my co-editor in public > if it can be avoided. It would be rather nice if your co-editor says something in public, like "hi there" for starters. > There has been no WG decision to change the requirement for > a References header in followups. I'm completely lost with this stuff. What's the precise problem ? Something without References is never a follow-up. No References => no follow-up. If that's the case we obviously also have: Follow-up => references. Based on this premise we don't have (and IMHO don't want): References => follow-up That's elementary logic as far as I'm concerned. We don't need your co-editor or Alexey for this issue, I'm 100% sure. Now to the claim that it's exactly the same as in RfC 2822: That's not exactly true. We truncate References, RfC 2822 does not, that's a Usepro issue. We separate References by at least one WSP, RfC 2822 does not, that's a Usefor issue. RfC 2822 fixes In-Reply-To without References, we'd expect the same from a mail2news gateway, but don't need it within news. It doesn't hurt, but it's not required. So what's the problem ? It's an optional header field as in RfC 2822, and its presence does not automatically mean "follow-up", it could be also something else. > I do not agree with the Chair on this one I'm not even sure where you disagree. Is it related to anything I said above ? Bye, Frank From owner-ietf-usefor@mail.imc.org Wed Apr 20 20:26:32 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id UAA13108 for ; Wed, 20 Apr 2005 20:26:30 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3L0Os0L097583 for ; Wed, 20 Apr 2005 17:24:54 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3L0OsHd097582 for ietf-usefor-skb; Wed, 20 Apr 2005 17:24:54 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from b.mail.peak.org (b.mail.peak.org [69.59.192.42]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3L0Orf5097573 for ; Wed, 20 Apr 2005 17:24:53 -0700 (PDT) (envelope-from stanley@peak.org) Received: from a.shell.peak.org ([69.59.192.81]) by b.mail.peak.org (8.12.10/8.12.8) with ESMTP id j3L0OlI5021878 (version=TLSv1/SSLv3 cipher=EDH-RSA-DES-CBC3-SHA bits=168 verify=NO) for ; Wed, 20 Apr 2005 17:24:48 -0700 (PDT) Date: Wed, 20 Apr 2005 17:24:47 -0700 (PDT) From: John Stanley X-X-Sender: stanley@a.shell.peak.org To: ietf-usefor@imc.org Subject: Re: Differences between RFC 2822 and Usefor (fwd) Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Spam-Score: 0 () X-Scanned-By: MIMEDefang 2.39 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: "Charles Lindsey" : >There has been no WG decision to >change the requirement for a References header in followups. Not according to USEFOR-03. >Our Chair has >taken the view that Usefor is not the place to say it, but that Usepro IS >the proper place (and indeed there is already a corresponding MUST in >Usepro, though it may need a little more amplification). USEFOR-03 says it is defined as found in RFC2822, and is optional. Optional means "not mandatory". There is no MUST. >But in that case, I will have to find somewhere in Usepro to say that the >References header MAY also be used in other cases such as multipart FAQs, If it is optional, then it is optional. You don't need to say it MAY be used when USEFOR-03 already says it may be used. Like I already said, once you remove the ability to detect followups by making References optional, it doesn't matter when it may or may not be used. In fact, it may be used anytime (that's what "optional" means), even apparently when there is nothing at all for it to convey. By changing the requirement for References, we've solve a whole passle of things that should be be CHANGED BACK to the way they were before you decided to take it upon yourself to change them. Like "non-followups MAY contain" References headers, which, despite your repeated claims to the contrary, is a major change that nobody asked for. >News readers that try to do threading care very much if it is missing. That is not an interoperability issue, and you know it. It doesn't cause the news reader to break, unless it is a VERY poorly designed news reader. How does it deal with non-followups today, which do not contain References headers, or broken followups, which also do not contain them, if they cannot deal with any article tomorrow that does not contain one? If they break today, they are patently broken and need to be fixed. I know of none that are that broken. Do you? And, by the way, it's software. Until it gains sentience, it doesn't care about anything. >No, the requirement has not gone. Read the draft-03. It is gone. Explicit language that makes References optional. "As defined in RFC2822". No MUST. A MUST that doesn't exist is not a requirement. >The mandate in question is already in USEPRO. And always has been. But is missing completely in USEFOR, where the header is defined. It's gone. It's time to stop pretending that it's there, or that it wasn't a deliberate change to remove it. I'm ok with it gone; why are you suddenly trying to say that it needs to be put back? You didn't say that prior to my commenting on it, it must have been ok -- you had ample time to comment on the draft prior to my speaking up, and did not. From owner-ietf-usefor@mail.imc.org Wed Apr 20 20:33:08 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id UAA13583 for ; Wed, 20 Apr 2005 20:33:07 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3L0W4x9098581 for ; Wed, 20 Apr 2005 17:32:04 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3L0W4YW098580 for ietf-usefor-skb; Wed, 20 Apr 2005 17:32:04 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtp2.Stanford.EDU (smtp2.Stanford.EDU [171.67.16.125]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3L0W38A098574 for ; Wed, 20 Apr 2005 17:32:03 -0700 (PDT) (envelope-from rra@stanford.edu) Received: from windlord.stanford.edu (windlord.Stanford.EDU [171.64.19.147]) by smtp2.Stanford.EDU (8.12.11/8.12.11) with SMTP id j3L0W2g5031344 for ; Wed, 20 Apr 2005 17:32:02 -0700 Received: (qmail 8989 invoked by uid 1000); 21 Apr 2005 00:32:02 -0000 To: ietf-usefor@imc.org Subject: Re: References In-Reply-To: <4266EF35.21A5@xyzzy.claranet.de> (Frank Ellermann's message of "Thu, 21 Apr 2005 02:09:25 +0200") References: <4266EF35.21A5@xyzzy.claranet.de> From: Russ Allbery Organization: The Eyrie Date: Wed, 20 Apr 2005 17:32:02 -0700 Message-ID: <87mzrtt1fh.fsf@windlord.stanford.edu> User-Agent: Gnus/5.1007 (Gnus v5.10.7) XEmacs/21.4 (Jumbo Shrimp, linux) MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Frank Ellermann writes: > Based on this premise we don't have (and IMHO don't want): > References => follow-up > That's elementary logic as far as I'm concerned. We don't > need your co-editor or Alexey for this issue, I'm 100% sure. Sounds right to me. If it doesn't have References, it's not a follow-up, but in practice people use References for things other than follow-ups, and I don't see any reason to ban those uses. So the question is whether we want to artificially define all those things as follow-ups as well, or not worry about it and just say that References is required for follow-ups but may also be used for other things. The second seems clearer and simpler to me. > Now to the claim that it's exactly the same as in RfC 2822: > That's not exactly true. We truncate References, RfC 2822 > does not, that's a Usepro issue. We separate References by > at least one WSP, RfC 2822 does not, that's a Usefor issue. Yup. > RfC 2822 fixes In-Reply-To without References, we'd expect > the same from a mail2news gateway, but don't need it within > news. It doesn't hurt, but it's not required. Yup. > So what's the problem ? It's an optional header field as > in RfC 2822, and its presence does not automatically mean > "follow-up", it could be also something else. Exactly. -- Russ Allbery (rra@stanford.edu) From owner-ietf-usefor@mail.imc.org Wed Apr 20 20:57:51 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id UAA15506 for ; Wed, 20 Apr 2005 20:57:49 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3L0uhi2002254 for ; Wed, 20 Apr 2005 17:56:43 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3L0uh1E002253 for ietf-usefor-skb; Wed, 20 Apr 2005 17:56:43 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtp2.Stanford.EDU (smtp2.Stanford.EDU [171.67.16.125]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3L0ugPx002247 for ; Wed, 20 Apr 2005 17:56:42 -0700 (PDT) (envelope-from rra@stanford.edu) Received: from windlord.stanford.edu (windlord.Stanford.EDU [171.64.19.147]) by smtp2.Stanford.EDU (8.12.11/8.12.11) with SMTP id j3L0ugCD005051 for ; Wed, 20 Apr 2005 17:56:42 -0700 Received: (qmail 9461 invoked by uid 1000); 21 Apr 2005 00:56:41 -0000 To: ietf-usefor@imc.org Subject: Re: Differences between RFC 2822 and Usefor (fwd) In-Reply-To: (John Stanley's message of "Wed, 20 Apr 2005 17:24:47 -0700 (PDT)") References: From: Russ Allbery Organization: The Eyrie Date: Wed, 20 Apr 2005 17:56:41 -0700 Message-ID: <87ekd5t0ae.fsf@windlord.stanford.edu> User-Agent: Gnus/5.1007 (Gnus v5.10.7) XEmacs/21.4 (Jumbo Shrimp, linux) MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: John Stanley writes: > USEFOR-03 says it is defined as found in RFC2822, and is > optional. Optional means "not mandatory". There is no MUST. I think it's obvious that the References header has to be optional in that sense, since you can construct a well-formed Usenet message that doesn't have a References header. I'll repeat my earlier suggestion here, which is to simply define followups in such a way as to say that if something doesn't have a References header, it's not a followup. That neatly avoids the whole issue without needing to worry about MUSTs and SHOULDs. It becomes a definitional issue. If it doesn't have References, it's not a followup. > Like I already said, once you remove the ability to detect followups by > making References optional, it doesn't matter when it may or may not be > used. In fact, it may be used anytime (that's what "optional" means), > even apparently when there is nothing at all for it to convey. > By changing the requirement for References, we've solve a whole passle > of things that should be be CHANGED BACK to the way they were before you > decided to take it upon yourself to change them. Like "non-followups MAY > contain" References headers, which, despite your repeated claims to the > contrary, is a major change that nobody asked for. Well, to be fair, I think it's a reasonable way of dealing with the issue. I'm not sure if that means I was asking for it, but I can see the logic behind it. The way I look at this, there are two separate issues that we keep mixing together: * What a news reader should do when constructing a posted reply to a particular post. * How a news reader should display messages with References headers. The former is what we mean by a followup, and by definition it has to build a References header. If it doesn't, it's doing something other than posting a followup. The latter is a separate issue, a UI issue. It is expected that messages connected by References will be shown in a thread, subject to various user configuration and choices like whether changing the Subject should result in a separate thread in the user's display (something that should be a configuration option, not something the standard should be mandating either way, as there are people who prefer it both ways and there is no interoperability issue). There are various reasons why sometimes people like to connect messages that are not posted replies to each other, and that do not, or only barely do, fit into the notion of a followup. I don't think there's really anything else we can say that's more specific than that. -- Russ Allbery (rra@stanford.edu) From owner-ietf-usefor@mail.imc.org Thu Apr 21 05:46:53 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id FAA11055 for ; Thu, 21 Apr 2005 05:46:53 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3L9jQv7023930 for ; Thu, 21 Apr 2005 02:45:26 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3L9jQY4023929 for ietf-usefor-skb; Thu, 21 Apr 2005 02:45:26 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from mail.gmx.net (pop.gmx.de [213.165.64.20]) by above.proper.com (8.12.11/8.12.9) with SMTP id j3L9jOXT023851 for ; Thu, 21 Apr 2005 02:45:25 -0700 (PDT) (envelope-from mail@sebastian-brocks.de) Received: (qmail invoked by alias); 21 Apr 2005 09:45:18 -0000 Received: from xdsl-213-196-200-53.netcologne.de (EHLO [192.168.1.2]) [213.196.200.53] by mail.gmx.net (mp029) with SMTP; 21 Apr 2005 11:45:18 +0200 X-Authenticated: #1840277 Message-ID: <4267762D.6020902@sebastian-brocks.de> Date: Thu, 21 Apr 2005 11:45:17 +0200 From: Sebastian Brocks User-Agent: Mozilla/5.0 (X11; U; Linux i686; de-DE; rv:1.7.6) Gecko/20050404 Thunderbird/1.0.2 Mnenhy/0.7.2.0 X-Accept-Language: de-DE, de, en-us, en MIME-Version: 1.0 To: ietf-usefor@imc.org Subject: Re: Differences between RFC 2822 and Usefor (fwd) References: <87ekd5t0ae.fsf@windlord.stanford.edu> In-Reply-To: <87ekd5t0ae.fsf@windlord.stanford.edu> X-Enigmail-Version: 0.90.0.0 X-Enigmail-Supports: pgp-inline, pgp-mime Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit X-Y-GMX-Trusted: 0 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Content-Transfer-Encoding: 7bit Russ Allbery schrieb: > I don't think there's really anything else we can say that's more specific > than that. I agree. greetings, Sebastian From owner-ietf-usefor@mail.imc.org Thu Apr 21 12:14:40 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id MAA12890 for ; Thu, 21 Apr 2005 12:14:40 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3LGDYT7084539 for ; Thu, 21 Apr 2005 09:13:34 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3LGDYri084537 for ietf-usefor-skb; Thu, 21 Apr 2005 09:13:34 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from lon-mail-2.gradwell.net (lon-mail-2.gradwell.net [193.111.201.126]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3LGDXsK084531 for ; Thu, 21 Apr 2005 09:13:33 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from host81-144-72-42.midband.mdip.bt.net ([81.144.72.42]) by lon-mail-2.gradwell.net with esmtp (Gradwell gwh-smtpd 1.181) id 4267d12b.858f.9 for ietf-usefor@imc.org; Thu, 21 Apr 2005 17:13:31 +0100 (envelope-sender ) Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3LGCYU23014 for ietf-usefor@imc.org; Thu, 21 Apr 2005 17:12:34 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20698 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: Differences between RFC 2822 and Usefor (fwd) Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: Date: Thu, 21 Apr 2005 15:55:38 GMT Lines: 29 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In John Stanley writes: >"Charles Lindsey" : >>Our Chair has >>taken the view that Usefor is not the place to say it, but that Usepro IS >>the proper place (and indeed there is already a corresponding MUST in >>Usepro, though it may need a little more amplification). >USEFOR-03 says it is defined as found in RFC2822, and is optional. >Optional means "not mandatory". There is no MUST. No, USEPRO-03 does not say it is optional. It is silent on the issue (and our Chair confirms this by saying that RFC 2822 contains both format and protocol material, and so if you want to see whether some protocol material in RFC 2822 still applies, then you have to look at USEPRO). I am not sure I like that way of looking at it, but it is what our Chair says he wants. -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 From owner-ietf-usefor@mail.imc.org Thu Apr 21 12:14:41 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id MAA12891 for ; Thu, 21 Apr 2005 12:14:40 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3LGDZO7084549 for ; Thu, 21 Apr 2005 09:13:35 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3LGDZst084548 for ietf-usefor-skb; Thu, 21 Apr 2005 09:13:35 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from lon-mail-2.gradwell.net (lon-mail-2.gradwell.net [193.111.201.126]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3LGDYhS084542 for ; Thu, 21 Apr 2005 09:13:35 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from host81-144-72-42.midband.mdip.bt.net ([81.144.72.42]) by lon-mail-2.gradwell.net with esmtp (Gradwell gwh-smtpd 1.181) id 4267d12d.858f.b for ietf-usefor@imc.org; Thu, 21 Apr 2005 17:13:33 +0100 (envelope-sender ) Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3LGCXh23009 for ietf-usefor@imc.org; Thu, 21 Apr 2005 17:12:33 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20697 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: Fwd: http://www.ietf.org/internet-drafts/draft-ietf-usefor-usefor-03.txt Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: <200504111212.j3BCChM14516@clerew.man.ac.uk> <87psx1gnjc.fsf@windlord.stanford.edu> <87fyxvx1m6.fsf@windlord.stanford.edu> <87mzs2a1hu.fsf@windlord.stanford.edu> <87wtqxui6s.fsf@windlord.stanford.edu> Date: Thu, 21 Apr 2005 15:44:40 GMT Lines: 34 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In <87wtqxui6s.fsf@windlord.stanford.edu> Russ Allbery writes: >Charles Lindsey writes: >> Ah! But in that case I don't think servers should be rejecting such >> headers. If the article has been wandering in any out of both email and >> news systems, then it is sometimes useful to have the full history >> recorded (though I might agree that if there was a mailing list involved >> I might not want to know what happened before it reached the mailing >> list). >Then I don't understand the distinction you're drawing between obsolete >and obsolescent. As near as I could tell, the only difference in practice >was that servers were rejecting the former headers, but now you're saying >that they shouldn't even do that. Generally speaking, our draft has always said (and I think Usefor still says) that if you see a header that you do not recognize as a news header, you just leave it be and ignore it. Essentially the same applies in Mail. The only time you should be worrying is if you DO recognize it as a news header, and can see that it is malformed and likely to cause trouble elsewhere (and especially so if you are an injecting agent). -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 From owner-ietf-usefor@mail.imc.org Thu Apr 21 13:51:14 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id NAA21721 for ; Thu, 21 Apr 2005 13:51:13 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3LHnP0x011354 for ; Thu, 21 Apr 2005 10:49:25 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3LHnPfM011353 for ietf-usefor-skb; Thu, 21 Apr 2005 10:49:25 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from a.mail.peak.org (a.mail.peak.org [69.59.192.41]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3LHnODg011333 for ; Thu, 21 Apr 2005 10:49:24 -0700 (PDT) (envelope-from stanley@peak.org) Received: from a.shell.peak.org ([69.59.192.81]) by a.mail.peak.org (8.12.10/8.12.8) with ESMTP id j3LHnHfO031096 (version=TLSv1/SSLv3 cipher=EDH-RSA-DES-CBC3-SHA bits=168 verify=NO) for ; Thu, 21 Apr 2005 10:49:19 -0700 (PDT) Date: Thu, 21 Apr 2005 10:49:17 -0700 (PDT) From: John Stanley X-X-Sender: stanley@a.shell.peak.org To: ietf-usefor@imc.org Subject: References Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Spam-Score: 0 () X-Scanned-By: MIMEDefang 2.39 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Frank Ellermann : >I'm completely lost with this stuff. What's the precise >problem ? Something without References is never a follow-up. Since References is now optional, yes, something without References can be a followup. And something WITH a References can be something other than a followup. The current language says that followups SHOULD have that header, it does not prohibit it in anything else. >If that's the case we obviously also have: >Follow-up => references. Not obvious. The header is optional. See RFC2822. >Based on this premise we don't have (and IMHO don't want): >References => follow-up The only purpose for the References header is to identify followups and to what article the present one is a followup. Previous thought on the matter was that this function ('identify followups') was important enough to justify an RFC2119 mandate. It is, after all, how true threading works and the only way it can be accomplished. But threading is a display issue, not an interoperability one, so it was an abuse of the language to pretend it merited an RFC2119 "MUST". If you have some other function in mind, please enlighten us, and then justify any mandates you want to apply. What interoperability problem does making References mandatory solve? >Now to the claim that it's exactly the same as in RfC 2822: >That's not exactly true. Of course it isn't true. Nobody has said it is. Read draft-usefor-03 to see the differences. None of those differences deals with a mandate for use. Notice that "we" were able to insert the differences that "we" thought were important from RFC2822. "We" apparently didn't think a MUST was important; I agree. Let's move on. Russ Allbery : >So the question is whether we want to >artificially define all those things as follow-ups as well, or not worry >about it and just say that References is required for follow-ups but may >also be used for other things. RFC2822 says that References SHOULD be used in followups. This is not a requirement, it is only a recommendation. What is the justification for making References a requirement? What interoperability issue exists? What breaks if a References header is not in an article, and don't say "news readers", because news readers deal all the time with articles that don't have References headers. They have to, since currently the standard says that non-followups must not have them. I've yet to see a news reader that doesn't deal with non-followups; I suggest that if you have one, you get a better newsreader. And if you remove the one function that the header performs today ("identify a followup"), then why bother keeping that header? How do you differentiate between "this article is a followup", and "this article contains a References header for some other reason"? And if you say "who cares?", I'll say 1) I do, and 2) the existing standard does, and 3) if nobody does, then get rid of the header, since it performs no useful function. >> So what's the problem ? It's an optional header field as >> in RfC 2822, and its presence does not automatically mean >> "follow-up", it could be also something else. >Exactly. So the purpose of References headers is now deprecated. That's an argument for making them optional (as we have already done), not for making them a requirement. >I think it's obvious that the References header has to be optional in that >sense, since you can construct a well-formed Usenet message that doesn't >have a References header. Currently, you cannot construct a well-formed USENET message that is a followup without a References header. That is a direct effect of the language in RFC1036 that says "It is required for all follow-up messages". While the article may appear to be well-formed, it is technically in violation of the standards and thus is not valid. This is exactly the reason why I argued that a specific proposed moderated newsgroup that was intended to contain only followups should enforce that requirement, and, as I recall, you agreed with the technical argument but disagreed with doing that for social/political reasons. Correct me if I am wrong in my recollection. Once you say it is optional, you cannot also say it is mandatory. If USEFOR defines valid USENET format messages, and USEFOR says that a header is optional, then any message that does not contain that header can be valid. Another standard for the same medium that says "this article is not valid", when USEFOR says it is, is contradictory and broken. >I'll repeat my earlier suggestion here, which is to simply define >followups in such a way as to say that if something doesn't have a >References header, it's not a followup. We are not debating what is and is not a followup here. We are discussing the fact that our draft now defers the definition of References to RFC2822 with certain limited differences, none of which deal with changing the RFC2119 status of that header. References is now a SHOULD, not a MUST. If it is to be changed to a MUST, then justify the change. If you cannot justify the change, it ought not be made. This group has suffered too many times from unilateral changes made at the whim of the person who controls the text to allow it to continue. Charles says it needs to be changed back; Charles ought to justify that "need" or drop it. > * What a news reader should do when constructing a posted reply to a > particular post. >[That] is what we mean by a followup, and by definition it has to >build a References header. If it doesn't, it's doing something other than >posting a followup. True wrt RFC1036. Untrue wrt draft-usefor-03. SHOULD is a recommendation. However strong you think it is, it is still only a recommendation. From owner-ietf-usefor@mail.imc.org Thu Apr 21 15:05:46 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id PAA27072 for ; Thu, 21 Apr 2005 15:05:46 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3LJ4Pp4036310 for ; Thu, 21 Apr 2005 12:04:25 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3LJ4Po5036309 for ietf-usefor-skb; Thu, 21 Apr 2005 12:04:25 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtp3.Stanford.EDU (smtp3.Stanford.EDU [171.67.16.138]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3LJ4Pgn036303 for ; Thu, 21 Apr 2005 12:04:25 -0700 (PDT) (envelope-from rra@stanford.edu) Received: from windlord.stanford.edu (windlord.Stanford.EDU [171.64.19.147]) by smtp3.Stanford.EDU (8.12.11/8.12.11) with SMTP id j3LJ4NSv006253 for ; Thu, 21 Apr 2005 12:04:24 -0700 Received: (qmail 11849 invoked by uid 1000); 21 Apr 2005 19:04:23 -0000 To: ietf-usefor@imc.org Subject: Re: Fwd: http://www.ietf.org/internet-drafts/draft-ietf-usefor-usefor-03.txt In-Reply-To: (Charles Lindsey's message of "Thu, 21 Apr 2005 15:44:40 GMT") References: <200504111212.j3BCChM14516@clerew.man.ac.uk> <87psx1gnjc.fsf@windlord.stanford.edu> <87fyxvx1m6.fsf@windlord.stanford.edu> <87mzs2a1hu.fsf@windlord.stanford.edu> <87wtqxui6s.fsf@windlord.stanford.edu> From: Russ Allbery Organization: The Eyrie Date: Thu, 21 Apr 2005 12:04:23 -0700 Message-ID: <87d5so2bpk.fsf@windlord.stanford.edu> User-Agent: Gnus/5.1007 (Gnus v5.10.7) XEmacs/21.4 (Jumbo Shrimp, linux) MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Charles Lindsey writes: > Russ Allbery writes: >> Then I don't understand the distinction you're drawing between obsolete >> and obsolescent. As near as I could tell, the only difference in >> practice was that servers were rejecting the former headers, but now >> you're saying that they shouldn't even do that. > Generally speaking, our draft has always said (and I think Usefor still > says) that if you see a header that you do not recognize as a news > header, you just leave it be and ignore it. Essentially the same applies > in Mail. I got this part, but it doesn't answer my confusion above. What's the difference between obsolete and obsolescent? -- Russ Allbery (rra@stanford.edu) From owner-ietf-usefor@mail.imc.org Thu Apr 21 18:02:17 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id SAA20763 for ; Thu, 21 Apr 2005 18:02:16 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3LM0vKe075343 for ; Thu, 21 Apr 2005 15:00:57 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3LM0vXJ075341 for ietf-usefor-skb; Thu, 21 Apr 2005 15:00:57 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtp1.Stanford.EDU (smtp1.Stanford.EDU [171.67.16.123]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3LM0vgX075335 for ; Thu, 21 Apr 2005 15:00:57 -0700 (PDT) (envelope-from rra@stanford.edu) Received: from windlord.stanford.edu (windlord.Stanford.EDU [171.64.19.147]) by smtp1.Stanford.EDU (8.12.11/8.12.11) with SMTP id j3LM0tWn026241 for ; Thu, 21 Apr 2005 15:00:55 -0700 Received: (qmail 16756 invoked by uid 1000); 21 Apr 2005 22:00:54 -0000 To: ietf-usefor@imc.org Subject: Re: References In-Reply-To: (John Stanley's message of "Thu, 21 Apr 2005 10:49:17 -0700 (PDT)") References: From: Russ Allbery Organization: The Eyrie Date: Thu, 21 Apr 2005 15:00:54 -0700 Message-ID: <87y8bb23jd.fsf@windlord.stanford.edu> User-Agent: Gnus/5.1007 (Gnus v5.10.7) XEmacs/21.4 (Jumbo Shrimp, linux) MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: John Stanley writes: > RFC2822 says that References SHOULD be used in followups. This is not a > requirement, it is only a recommendation. What is the justification for > making References a requirement? What interoperability issue exists? > What breaks if a References header is not in an article, and don't say > "news readers", because news readers deal all the time with articles > that don't have References headers. They have to, since currently the > standard says that non-followups must not have them. I've yet to see a > news reader that doesn't deal with non-followups; I suggest that if you > have one, you get a better newsreader. I don't want to use any sort of SHOULD or MUST in this area (beyond the one we inherit from RFC 2822). I want to say that, in a Usenet context, a followup is *defined* to be a message posted in reply to another message with a References header pointing to the message being replied to. In other words, if your news posting software doesn't insert a References header, it's still posting well-formed messages. It's just not posting *followups*, by definition. (So if your software calls such messages followups, it has a serious definitional problem.) I think we're mostly agreeing. I want to take this approach since it avoids the whole question of whether things really break without References headers. > And if you remove the one function that the header performs today > ("identify a followup"), then why bother keeping that header? How do you > differentiate between "this article is a followup", and "this article > contains a References header for some other reason"? In practice, I don't think there's any need to distinguish, for UI purposes, between different reasons why an article cites a previous article with References. I just think it's a definitional stretch to call all such messages followups, since people normally associate the idea of followup with a reply and clearly not all of those articles are replies. > And if you say "who cares?", I'll say 1) I do, and 2) the existing > standard does, and 3) if nobody does, then get rid of the header, since > it performs no useful function. The function the References header provides is enabling construction of threads. This is closely related to the notion of followups but, to me, is not identical to it. > True wrt RFC1036. Untrue wrt draft-usefor-03. SHOULD is a > recommendation. However strong you think it is, it is still only a > recommendation. Right, I agree that what I'm pushing for is not what's in the current language. -- Russ Allbery (rra@stanford.edu) From owner-ietf-usefor@mail.imc.org Thu Apr 21 21:39:34 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id VAA08660 for ; Thu, 21 Apr 2005 21:39:33 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3M1bd2K036542 for ; Thu, 21 Apr 2005 18:37:39 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3M1bdRF036541 for ietf-usefor-skb; Thu, 21 Apr 2005 18:37:39 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from ciao.gmane.org (main.gmane.org [80.91.229.2]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3M1bbL5036530 for ; Thu, 21 Apr 2005 18:37:38 -0700 (PDT) (envelope-from usenet-format@gmane.org) Received: from list by ciao.gmane.org with local (Exim 4.43) id 1DOn2O-0008Lb-Ab for ietf-usefor@imc.org; Fri, 22 Apr 2005 03:32:56 +0200 Received: from 212.82.251.14 ([212.82.251.14]) by main.gmane.org with esmtp (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Fri, 22 Apr 2005 03:32:56 +0200 Received: from nobody by 212.82.251.14 with local (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Fri, 22 Apr 2005 03:32:56 +0200 X-Injected-Via-Gmane: http://gmane.org/ To: ietf-usefor@imc.org From: Frank Ellermann Subject: Re: References Date: Fri, 22 Apr 2005 03:36:29 +0200 Organization: Lines: 10 Message-ID: <4268551D.5434@xyzzy.claranet.de> References: <87y8bb23jd.fsf@windlord.stanford.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Complaints-To: usenet@sea.gmane.org X-Gmane-NNTP-Posting-Host: 212.82.251.14 X-Mailer: Mozilla 3.0 (OS/2; U) Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Content-Transfer-Encoding: 7bit Russ Allbery wrote: > people normally associate the idea of followup with a reply People: yes. Anybody who knows RfC 2822 and (s-o-) 1036: no. A followup is defined in usepro, a reply is defined in 2822, they are to great parts unrelated and / or different. There is not one string "followup" in the complete 2822 text. Bye. From owner-ietf-usefor@mail.imc.org Thu Apr 21 22:24:46 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id WAA10563 for ; Thu, 21 Apr 2005 22:24:46 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3M2Nav8051286 for ; Thu, 21 Apr 2005 19:23:36 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3M2NaQN051279 for ietf-usefor-skb; Thu, 21 Apr 2005 19:23:36 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from lon-mail-6.gradwell.net (lon-mail-6.gradwell.net [193.111.201.132]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3M2NXYi051229 for ; Thu, 21 Apr 2005 19:23:34 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from host81-144-73-224.midband.mdip.bt.net ([81.144.73.224]) by lon-mail-6.gradwell.net with esmtp (Gradwell gwh-smtpd 1.181) id 42686024.1595d.ff for ietf-usefor@imc.org; Fri, 22 Apr 2005 03:23:32 +0100 (envelope-sender ) Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3M2CBl28135 for ietf-usefor@imc.org; Fri, 22 Apr 2005 03:12:11 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20700 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: References Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: <4266EF35.21A5@xyzzy.claranet.de> <87mzrtt1fh.fsf@windlord.stanford.edu> Date: Thu, 21 Apr 2005 16:29:29 GMT Lines: 30 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In <87mzrtt1fh.fsf@windlord.stanford.edu> Russ Allbery writes: >Sounds right to me. >If it doesn't have References, it's not a follow-up, but in practice >people use References for things other than follow-ups, and I don't see >any reason to ban those uses. So the question is whether we want to >artificially define all those things as follow-ups as well, or not worry >about it and just say that References is required for follow-ups but may >also be used for other things. Exactly. Those other uses for the References header deserve a brief mention somewhere, probably in USEFOR, but it would indeed be somewhat artificial to define those cases as "followups", and I personally do not want to do that. >The second seems clearer and simpler to me. Yes, that now seems to be where we are headed. -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 From owner-ietf-usefor@mail.imc.org Thu Apr 21 23:11:34 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id WAA10562 for ; Thu, 21 Apr 2005 22:24:46 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3M2Na8u051285 for ; Thu, 21 Apr 2005 19:23:36 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3M2NanH051272 for ietf-usefor-skb; Thu, 21 Apr 2005 19:23:36 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from lon-mail-6.gradwell.net (lon-mail-6.gradwell.net [193.111.201.132]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3M2NYnn051234 for ; Thu, 21 Apr 2005 19:23:34 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from host81-144-73-224.midband.mdip.bt.net ([81.144.73.224]) by lon-mail-6.gradwell.net with esmtp (Gradwell gwh-smtpd 1.181) id 42686024.1595d.100 for ietf-usefor@imc.org; Fri, 22 Apr 2005 03:23:32 +0100 (envelope-sender ) Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3M2CAf28127 for ietf-usefor@imc.org; Fri, 22 Apr 2005 03:12:10 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20699 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: References (was: Differences between RFC 2822 and Usefor (fwd)) Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: <4266EF35.21A5@xyzzy.claranet.de> Date: Thu, 21 Apr 2005 16:22:39 GMT Lines: 62 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In <4266EF35.21A5@xyzzy.claranet.de> Frank Ellermann writes: >Charles Lindsey wrote: >> There has been no WG decision to change the requirement for >> a References header in followups. >I'm completely lost with this stuff. What's the precise >problem ? Something without References is never a follow-up. >No References => no follow-up. Which indeed follows from the fact that USEPRO says that a followup agent MUST incorporate a References header (and do it correctly, of course). >If that's the case we obviously also have: >Follow-up => references. A pure logical deduction indeed. But for the removal of all doubt, and to emphasise the difference from the treatment of "replies" in RFC 2822, we used to say that as well. Now we don't (because our Chair so decrees), but because it was redundant nothing has actually changed (John's claims to the contrary notwithstanding). >Based on this premise we don't have (and IMHO don't want): >References => follow-up But there is just one niggle here (yes, I agree we want sometimes to use References for multi-part FAQs and other special cases). How do we define the term "followup". Is is the the thing which is produced (as carefully defined in USEPRO) by a "followup agent", or is it anything that has a References header. I prefer the former (and you seem to have been using it in that sense), and John has been arguing for the latter. Technically, it makes no difference so long as we use the term consistently. >> I do not agree with the Chair on this one >I'm not even sure where you disagree. Is it related to >anything I said above ? It is because I preferred to mention it in both places, for the removal of all doubt as stated above, and also to provide a convenient place to introduce the special cases (FASQs etc) and to exaplain the semantics of the header properly. That still has to be done somewhere, and probably in USEFOR. -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 From owner-ietf-usefor@mail.imc.org Thu Apr 21 23:36:29 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id XAA15055 for ; Thu, 21 Apr 2005 23:36:28 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3M3ZNKO070504 for ; Thu, 21 Apr 2005 20:35:23 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3M3ZNl3070503 for ietf-usefor-skb; Thu, 21 Apr 2005 20:35:23 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from ciao.gmane.org (main.gmane.org [80.91.229.2]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3M3ZLJq070495 for ; Thu, 21 Apr 2005 20:35:22 -0700 (PDT) (envelope-from usenet-format@gmane.org) Received: from list by ciao.gmane.org with local (Exim 4.43) id 1DOosI-00022L-0N for ietf-usefor@imc.org; Fri, 22 Apr 2005 05:30:38 +0200 Received: from 212.82.251.14 ([212.82.251.14]) by main.gmane.org with esmtp (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Fri, 22 Apr 2005 05:30:37 +0200 Received: from nobody by 212.82.251.14 with local (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Fri, 22 Apr 2005 05:30:37 +0200 X-Injected-Via-Gmane: http://gmane.org/ To: ietf-usefor@imc.org From: Frank Ellermann Subject: Re: References Date: Fri, 22 Apr 2005 05:25:05 +0200 Organization: Lines: 37 Message-ID: <42686E91.42E2@xyzzy.claranet.de> References: <4266EF35.21A5@xyzzy.claranet.de> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Complaints-To: usenet@sea.gmane.org X-Gmane-NNTP-Posting-Host: 212.82.251.14 X-Mailer: Mozilla 3.0 (OS/2; U) Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Content-Transfer-Encoding: 7bit Charles Lindsey wrote: > How do we define the term "followup". | A "followup" is an article containing a response to the | contents of an earlier article (the followup's "precursor"). Found in usefor-03. It's used for things like "Followup-To: ". Just above the definition of "Followup agent" in chapter 1.5 "Definitions". > Is is the the thing which is produced (as carefully defined > in USEPRO) by a "followup agent" Sure. > or is it anything that has a References header. Not sure. If it's the RfC 2822 idea of a reply coming from a mail2news gateway it probably is some "followup", but behind a gateway you never know. Maybe it only had an In-Reply-To, the gateway tried to fix it, but it was a broken pipermail archive In-ReplyTo (the thread is then correct, but the "followup" is shown at the wrong place in the thread). > I prefer the former (and you seem to have been using it in > that sense) ACK. You could add "it has a References header field" to be absolutely sure: A "followup" is an article containing a response to the contents of an earlier article (the followup's "precursor" indicated in the "References" header field). Bye, Frank From owner-ietf-usefor@mail.imc.org Fri Apr 22 00:26:32 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id AAA18235 for ; Fri, 22 Apr 2005 00:26:31 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3M4PZl3079903 for ; Thu, 21 Apr 2005 21:25:35 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3M4PZQo079902 for ietf-usefor-skb; Thu, 21 Apr 2005 21:25:35 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtp3.Stanford.EDU (smtp3.Stanford.EDU [171.67.16.138]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3M4PZnj079896 for ; Thu, 21 Apr 2005 21:25:35 -0700 (PDT) (envelope-from rra@stanford.edu) Received: from windlord.stanford.edu (windlord.Stanford.EDU [171.64.19.147]) by smtp3.Stanford.EDU (8.12.11/8.12.11) with SMTP id j3M4PYKL006950 for ; Thu, 21 Apr 2005 21:25:34 -0700 Received: (qmail 32648 invoked by uid 1000); 22 Apr 2005 04:25:34 -0000 To: ietf-usefor@imc.org Subject: Re: References In-Reply-To: <42686E91.42E2@xyzzy.claranet.de> (Frank Ellermann's message of "Fri, 22 Apr 2005 05:25:05 +0200") References: <4266EF35.21A5@xyzzy.claranet.de> <42686E91.42E2@xyzzy.claranet.de> From: Russ Allbery Organization: The Eyrie Date: Thu, 21 Apr 2005 21:25:34 -0700 Message-ID: <87fyxja14x.fsf@windlord.stanford.edu> User-Agent: Gnus/5.1007 (Gnus v5.10.7) XEmacs/21.4 (Jumbo Shrimp, linux) MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Frank Ellermann writes: > ACK. You could add "it has a References header field" to be > absolutely sure: > A "followup" is an article containing a response to the > contents of an earlier article (the followup's "precursor" > indicated in the "References" header field). This is the definition that I'd go for. -- Russ Allbery (rra@stanford.edu) From owner-ietf-usefor@mail.imc.org Fri Apr 22 12:15:34 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id MAA01227 for ; Fri, 22 Apr 2005 12:15:33 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3MGEPT6019201 for ; Fri, 22 Apr 2005 09:14:25 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3MGEPFQ019195 for ietf-usefor-skb; Fri, 22 Apr 2005 09:14:25 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from lon-mail-6.gradwell.net (lon-mail-6.gradwell.net [193.111.201.132]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3MGENdq019148 for ; Fri, 22 Apr 2005 09:14:24 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from host81-144-64-140.midband.mdip.bt.net ([81.144.64.140]) by lon-mail-6.gradwell.net with esmtp (Gradwell gwh-smtpd 1.181) id 426922de.cea8.b7 for ietf-usefor@imc.org; Fri, 22 Apr 2005 17:14:22 +0100 (envelope-sender ) Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3MGCK102367 for ietf-usefor@imc.org; Fri, 22 Apr 2005 17:12:20 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20706 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: References Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: Date: Fri, 22 Apr 2005 13:29:11 GMT Lines: 106 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In John Stanley writes: >Frank Ellermann : >Since References is now optional, yes, No, they are NOT optional. Read Usepro. Everything that comes out of a followup agent MUST have one. Please stop deliberately distorting what everyone else is saying. > something without References can be >a followup. No. And something WITH a References can be something other than a >followup. Yes. >The only purpose for the References header is to identify followups and to >what article the present one is a followup. No, its primary purpose is to facilitate threading, but it is also useful to identify followups, as you say, and also to enable quick retrieval of precursors by clicking on them. And to stitch message/partials together. And maybe other things. > Previous thought on the matter >was that this function ('identify followups') was important enough to >justify an RFC2119 mandate. It is, after all, how true threading works and >the only way it can be accomplished. But threading is a display issue, >not an interoperability one, so it was an abuse of the language to >pretend it merited an RFC2119 "MUST". No. If you read USEPRO carefully, you will see that threading (or other means of presenting lists of articles for display) is an (optional) feature of the protocol; i.e. implememtors MAY provide it and, if they so choose, then they are entitled to assume that the headers they rely on for the purpose (References, Subject, Date) have been properly constructed, and this merits MUST or SHOULD wording in the appropriate places. >Russ Allbery : >>So the question is whether we want to >>artificially define all those things as follow-ups as well, or not worry >>about it and just say that References is required for follow-ups but may >>also be used for other things. >RFC2822 says that References SHOULD be used in followups. No it doesn't. It says they should be used for "Replies", which is not quite the same thing. However, followups and replies are sufficiently similar that if we put a MUST somewhere regarding followups, we need to draw attention to the difference from the treatment of replies by RFC 2822. >And if you remove the one function that the header performs today >("identify a followup"), then why bother keeping that header? How do you >differentiate between "this article is a followup", and "this article >contains a References header for some other reason"? And if you say "who >cares?", I'll say 1) I do, Well you are giving a good impression of not caring :-( . And even if you don't care, then I still do. > and 2) the existing standard does, Which is a good reason for keeping it so, and we should not change it without a clear agreement in this WG to do so. >>I think it's obvious that the References header has to be optional in that >>sense, since you can construct a well-formed Usenet message that doesn't >>have a References header. >Currently, you cannot construct a well-formed USENET message that is a >followup without a References header. That is a direct effect of the >language in RFC1036 that says "It is required for all follow-up messages". And it's a direct effect of what USEPRO says. >>I'll repeat my earlier suggestion here, which is to simply define >>followups in such a way as to say that if something doesn't have a >>References header, it's not a followup. >We are not debating what is and is not a followup here. Yes we are. And the opinion I am hearing (even from you now) is that it is whatever a followup agent produces. I think, though, I would also want to include the case where a user constructs a followup manually (possibly to more than one precursor, as you seem to do) and generates a suitable References header manually. -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 From owner-ietf-usefor@mail.imc.org Fri Apr 22 12:15:49 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id MAA01263 for ; Fri, 22 Apr 2005 12:15:49 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3MGEMLE019145 for ; Fri, 22 Apr 2005 09:14:22 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3MGEMri019144 for ietf-usefor-skb; Fri, 22 Apr 2005 09:14:22 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from lon-mail-6.gradwell.net (lon-mail-6.gradwell.net [193.111.201.132]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3MGELqK019137 for ; Fri, 22 Apr 2005 09:14:22 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from host81-144-64-140.midband.mdip.bt.net ([81.144.64.140]) by lon-mail-6.gradwell.net with esmtp (Gradwell gwh-smtpd 1.181) id 426922dc.cea8.b5 for ietf-usefor@imc.org; Fri, 22 Apr 2005 17:14:20 +0100 (envelope-sender ) Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3MGCIR02359 for ietf-usefor@imc.org; Fri, 22 Apr 2005 17:12:18 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20705 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: Fwd: http://www.ietf.org/internet-drafts/draft-ietf-usefor-usefor-03.txt Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: <200504111212.j3BCChM14516@clerew.man.ac.uk> <87psx1gnjc.fsf@windlord.stanford.edu> <87fyxvx1m6.fsf@windlord.stanford.edu> <87mzs2a1hu.fsf@windlord.stanford.edu> <87wtqxui6s.fsf@windlord.stanford.edu> <87d5so2bpk.fsf@windlord.stanford.edu> Date: Fri, 22 Apr 2005 12:09:06 GMT Lines: 27 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In <87d5so2bpk.fsf@windlord.stanford.edu> Russ Allbery writes: >I got this part, but it doesn't answer my confusion above. What's the >difference between obsolete and obsolescent? "Obsolete" means this is no longer an official Netnews header. It would be labelled as "obsolete" in the IANA Registy of headers. "Obsolescent" means that is is still an official Netnews header (just), but that its continued use is discouraged (for reasons that are given), and that it is likely to be declared fully obsolete in a future standard. Our earlier drafts uses the words "is to be regarded as obsolete", which falls slightly short of "is obsolete" and would IMO be better expressed as "obsolescent", if indeed that is the meaning we wish to convey (it was certainly the meaning I intended to convey in those drafts). -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 From owner-ietf-usefor@mail.imc.org Fri Apr 22 14:18:24 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id OAA09931 for ; Fri, 22 Apr 2005 14:18:24 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3MIH7HA058335 for ; Fri, 22 Apr 2005 11:17:07 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3MIH7Tb058334 for ietf-usefor-skb; Fri, 22 Apr 2005 11:17:07 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from b.mail.peak.org (b.mail.peak.org [69.59.192.42]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3MIH6Gh058278 for ; Fri, 22 Apr 2005 11:17:06 -0700 (PDT) (envelope-from stanley@peak.org) Received: from a.shell.peak.org ([69.59.192.81]) by b.mail.peak.org (8.12.10/8.12.8) with ESMTP id j3MIH0I5031160 (version=TLSv1/SSLv3 cipher=EDH-RSA-DES-CBC3-SHA bits=168 verify=NO) for ; Fri, 22 Apr 2005 11:17:01 -0700 (PDT) Date: Fri, 22 Apr 2005 11:16:59 -0700 (PDT) From: John Stanley X-X-Sender: stanley@a.shell.peak.org To: ietf-usefor@imc.org Subject: Re: Differences between RFC 2822 and Usefor (fwd) Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Spam-Score: 0 () X-Scanned-By: MIMEDefang 2.39 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: "Charles Lindsey" : > No, USEPRO-03 does not say it is optional. Read what I wrote, Charles, and respond to that. I said: >USEFOR-03 says it is defined as found in RFC2822, and is optional. USEFOR is where the definition appears, and that definition defers to RFC2822 for almost everything. One of the areas it does NOT change RFC2822 is in the requirement for use of the header, which says "SHOULD". There is no MUST, either in RFC2822 or USEFOR. The header is optional, even in followups. Now, I realize that YOU say that RFC2119 "SHOULD" is a requirement of some kind, but RFC2119 actually says that it is a recommendation, and "required" and "recommended" are two very different words with two very different meanings, even when one of them carries an additional adjective "STRONG". From owner-ietf-usefor@mail.imc.org Fri Apr 22 14:33:05 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id OAA10761 for ; Fri, 22 Apr 2005 14:33:04 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3MIVthl063985 for ; Fri, 22 Apr 2005 11:31:55 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3MIVtxs063984 for ietf-usefor-skb; Fri, 22 Apr 2005 11:31:55 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from b.mail.peak.org (b.mail.peak.org [69.59.192.42]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3MIVsBT063950 for ; Fri, 22 Apr 2005 11:31:54 -0700 (PDT) (envelope-from stanley@peak.org) Received: from a.shell.peak.org ([69.59.192.81]) by b.mail.peak.org (8.12.10/8.12.8) with ESMTP id j3MIVmI5037154 (version=TLSv1/SSLv3 cipher=EDH-RSA-DES-CBC3-SHA bits=168 verify=NO) for ; Fri, 22 Apr 2005 11:31:49 -0700 (PDT) Date: Fri, 22 Apr 2005 11:31:48 -0700 (PDT) From: John Stanley X-X-Sender: stanley@a.shell.peak.org To: ietf-usefor@imc.org Subject: Re: References Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Spam-Score: 0 () X-Scanned-By: MIMEDefang 2.39 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: "Charles Lindsey" : >Frank Ellermann writes: >>No References => no follow-up. >Which indeed follows from the fact that USEPRO says that a followup agent >MUST incorporate a References header (and do it correctly, of course). Unfortunately, RFC2822 says only SHOULD regarding the References header and followups, and USEFOR says that the References header is defined as found in RFC2822 (and for the pedants who cannot identify the context of this discussion, I'll add "other than a few syntax changes that don't change the requirements for usage, which is the context of this discussion here.) So References in a followup is only a SHOULD, and if USEPRO says otherwise, it contradicts both RFC2822 and USEFOR. While one of those we can modify for good reason, the other is the result of our modifications, and we did NOT modify the References header to make it mandatory in any use. >A pure logical deduction indeed. But for the removal of all doubt, and to >emphasise the difference from the treatment of "replies" in RFC 2822, we >used to say that as well. Now we don't (because our Chair so decrees), but >because it was redundant nothing has actually changed (John's claims to >the contrary notwithstanding). Stop putting words in my mouth, Charles. >Is is the the thing which is produced (as carefully defined in USEPRO) by >a "followup agent", or is it anything that has a References header. I >prefer the former (and you seem to have been using it in that sense), and >John has been arguing for the latter. Technically, it makes no difference >so long as we use the term consistently. Stop putting words in my mouth, Charles. You don't understand what I've said, so you are the last person who ought to be pretending you do. You being the editor and able to make willy-nilly changes to the drafts does NOT give you the right to put words specifically in MY mouth, even though you are putting words regularly into the mouths of the group as a whole. You are apparently incapable of differentiating between "poster intends a followup in what he sends" and "reader detects a followup in what he gets." The previous standards and drafts have said that the intent for something to be a followup by a poster REQUIRES a References header to be inserted (MUST), and otherwise that header is prohibited. That means that the READER can then determine what the poster intends to be a followup by looking for a References header. And yes, Charles, when you say "non-followups MAY contain a References header", you have removed that ability to DETECT at the reader's end of the pipe what is and is not intended to be a followup. That is a change, despite your repeated denials. From owner-ietf-usefor@mail.imc.org Fri Apr 22 15:54:08 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id PAA18143 for ; Fri, 22 Apr 2005 15:54:07 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3MJqjnU086990 for ; Fri, 22 Apr 2005 12:52:45 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3MJqjGV086989 for ietf-usefor-skb; Fri, 22 Apr 2005 12:52:45 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from a.mail.peak.org (a.mail.peak.org [69.59.192.41]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3MJqjkG086971 for ; Fri, 22 Apr 2005 12:52:45 -0700 (PDT) (envelope-from stanley@peak.org) Received: from a.shell.peak.org ([69.59.192.81]) by a.mail.peak.org (8.12.10/8.12.8) with ESMTP id j3MJqdfO024233 (version=TLSv1/SSLv3 cipher=EDH-RSA-DES-CBC3-SHA bits=168 verify=NO) for ; Fri, 22 Apr 2005 12:52:39 -0700 (PDT) Date: Fri, 22 Apr 2005 12:52:39 -0700 (PDT) From: John Stanley X-X-Sender: stanley@a.shell.peak.org To: ietf-usefor@imc.org Subject: Re: References Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Spam-Score: 0 () X-Scanned-By: MIMEDefang 2.39 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Russ Allbery : >I don't want to use any sort of SHOULD or MUST in this area (beyond the >one we inherit from RFC 2822). Ok. I'm happy with that. But we cannot then change our minds in mid-stream and put a SHOULD NOT or MUST into other pieces of the news standards, since that would contradict USEFOR and RFC2822. >I want to say that, in a Usenet context, a >followup is *defined* to be a message posted in reply to another message >with a References header pointing to the message being replied to. So what? Not "so what, you want to say", but "so what if a followup is defined that way?" You can't identify one as such using the RFC2822 headers, so why bother putting a special name on it? It's like defining a "narquark" as "an article posted by a person with a mole on his left shoulder." If you cannot identify "narquarks", you aren't able to treat them specially, so they are still just "articles" and get treated like any other. In other words, if it is important enough to DEFINE what a followup is (presumably so it can be treated specially), then it is important enough to provide a mechanism to identify followups at the point you want to apply that special treatment. The only person who really knows "this is a response to somethign else" is the poster. Without a References MUST/MUST NOT dichotomy, you cannot identify them once they leave his hands. You get "well, this sorta looks like a followup" and MAYBE it is, maybe it isn't. (And in case the argument that "it guesses right often enough", I'll point out that this is a TECHNICAL standard and TECHNICAL standards don't promote guessing, especially when they could promote fact just as easily.) >In other words, if your news posting software doesn't insert a References >header, it's still posting well-formed messages. It's just not posting >*followups*, by definition. Well, I guess we have to disagree here. I don't think you can isolate the concept of a "followup" from the intent of the poster. An article that someone posts using F in trn, where he puts in a new subject and entirely new content, but forgets to remove the References header, is not really a followup of any kind. The recipient would treat it as such based on the References header, but it really isn't. It's really an incorrectly formatted article (today) because it contains a header it is prohibited from containing based on its purpose. >In practice, I don't think there's any need to distinguish, for UI >purposes, between different reasons why an article cites a previous >article with References. Then "followup" is a meaningless concept, since it cannot be identified for special treatment. >The function the References header provides is enabling construction of >threads. This is closely related to the notion of followups but, to me, >is not identical to it. What is a "thread" if there is no "followup"? What is a "thread" when a "thread" may contain any collection of articles at all, not just ones that are somehow related in the real world? So, this looks like a complete turnaround in what I'm saying, huh? No followups, no threads, References meaningless? Nope. Here's the crux: Either we say "RFC1036 was right" regarding the References header and followups and propogate that, with all the implications, into the new standard, or we say "RFC1036 was wrong" and propogate that, with all the implications, into the new standard. This half of one, half of the other attitude is ridiculous. If RFC1036 is right, then Rererences is a MUST/MUST NOT header, and followup is a valuable concept because we can identify them at the recipient's end. If RFC1036 is wrong, then References is a MAY/SHOULD/SOMETIMES header and followup is meaningless because we cannot identify them. From owner-ietf-usefor@mail.imc.org Fri Apr 22 16:20:02 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id QAA26521 for ; Fri, 22 Apr 2005 16:20:01 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3MKIXhH091197 for ; Fri, 22 Apr 2005 13:18:33 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3MKIXm4091196 for ietf-usefor-skb; Fri, 22 Apr 2005 13:18:33 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from lentil.epix.net (lentil.epix.net [199.224.64.67]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3MKIWwI091189 for ; Fri, 22 Apr 2005 13:18:32 -0700 (PDT) (envelope-from mibsoft@epix.net) Received: from [192.168.2.11] (hrbg-66-33-227-136-pppoe.dsl.hrbg.epix.net [66.33.227.136]) by lentil.epix.net (8.12.10/2004120601/PL) with ESMTP id j3MKIMl8000281 for ; Fri, 22 Apr 2005 16:18:27 -0400 (EDT) Message-ID: <42695C10.2030408@epix.net> Date: Fri, 22 Apr 2005 16:18:24 -0400 From: "Forrest J. Cavalier III" User-Agent: Mozilla Thunderbird 0.7 (Windows/20040616) X-Accept-Language: en-us, en MIME-Version: 1.0 To: ietf-usefor@imc.org Subject: Re: References References: In-Reply-To: Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit X-Scanned-By: MIMEDefang 2.51 on 199.224.89.154 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Content-Transfer-Encoding: 7bit > Either we say "RFC1036 was right" regarding the References header and > followups and propogate that, with all the implications, into the new > standard, or we say "RFC1036 was wrong" and propogate that, with all the > implications, into the new standard. This half of one, half of the other > attitude is ridiculous. If RFC1036 is right, then Rererences is a > MUST/MUST NOT header, and followup is a valuable concept because we can > identify them at the recipient's end. If RFC1036 is wrong, then References > is a MAY/SHOULD/SOMETIMES header and followup is meaningless because we > cannot identify them. > > Me too. John has been pointing out problems with this all along, much past the point of getting ignored. So I want to be on record that I tend to think this paragraph sums it up. There is no reason USEFOR should get away with the result of something like a bad Abbot and Costello routine. From owner-ietf-usefor@mail.imc.org Fri Apr 22 16:41:58 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id QAA00932 for ; Fri, 22 Apr 2005 16:41:58 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3MKeJmu096898 for ; Fri, 22 Apr 2005 13:40:19 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3MKeJnl096897 for ietf-usefor-skb; Fri, 22 Apr 2005 13:40:19 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from a.mail.peak.org (a.mail.peak.org [69.59.192.41]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3MKeItM096864 for ; Fri, 22 Apr 2005 13:40:18 -0700 (PDT) (envelope-from stanley@peak.org) Received: from a.shell.peak.org ([69.59.192.81]) by a.mail.peak.org (8.12.10/8.12.8) with ESMTP id j3MKeDfO044114 (version=TLSv1/SSLv3 cipher=EDH-RSA-DES-CBC3-SHA bits=168 verify=NO) for ; Fri, 22 Apr 2005 13:40:13 -0700 (PDT) Date: Fri, 22 Apr 2005 13:40:13 -0700 (PDT) From: John Stanley X-X-Sender: stanley@a.shell.peak.org To: ietf-usefor@imc.org Subject: Re: References Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Spam-Score: 0 () X-Scanned-By: MIMEDefang 2.39 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: "Charles Lindsey" : >>Since References is now optional, yes, >No, they are NOT optional. Read Usepro. Read USEEFOR, where the header is defined, and then read RFC2822, which is what USEFOR refers to. That "SHOULD" in RFC2822 means it is not mandatory, it is just a "strong" recommendation. A recommendation is still just a recommendation. >Everything that comes out of a followup agent MUST have one. Not according to RFC2822, which is where the header is actually defined. >Please stop deliberately distorting what everyone else is saying. Coming from someone who deliberately and repeatedly attempts to put words in my mouth, your command is simply ludicrous. I'm not distorting anything; I'm reading the draft USEFOR standard and RFC2822. I suggest you try doing so before you toss about claims of distortion. >> something without References can be a followup. >No. Of course it can. RFC2822 says, and to avoid more condemnation from you that I am "distorting" something, I'll quote: references 0* 1 SHOULD occur in some replies - see 3.6.4 "Some replies". Doesn't even go as far as to say "SHOULD occur in all replies", just "some". And: Furthermore, reply messages SHOULD have "In-Reply-To:" and "References:" fields as appropriate, as described below. Just SHOULD. And: The "In-Reply-To:" field may be used to identify the message (or messages) to which the new message is a reply, while the "References:" field may be used to identify a "thread" of conversation. "May". Not a requirement. Furthermore, RFC2822 continues: If the parent message does not contain a "References:" field but does have an "In-Reply-To:" field containing a single message identifier, So, RFC2822 explicitely covers an example case of a "reply" (followup, in USENET terminology) that does NOT contain a References header. >No, its primary purpose is to facilitate threading, Threading is a user interface issue, and we do not deal with user interface issues. Furthermore, the "threading" you speak of is due to the followup nature of the articles it identifies as such. That makes its PRIMARY purpose to IDENTIFY followups (and the articles to which this is a followup); after that, the user agent can do what it wants with them when it displays them. >No. If you read USEPRO carefully, If you read RFC2822 carefully, which you must do if you want to understand USEFOR, you will see that References is OPTIONAL, and SHOULD appear in SOME replies. Not mandatory in anything. >you will see that threading (or other >means of presenting lists of articles for display) is an (optional) >feature of the protocol; It is ridiculous to argue that a header that implements an OPTIONAL "feature of the protocol" is in any way MANDATORY. But then, it's not a "feature of the protocol", it is just a user-interface issue, and that makes it even more ridiculous to argue that it is mandatory. Nothing at all cares about that header in the protocol; only the user agent that wants to do real threading cares, and it has to be able to deal with articles sans References headers already. It can deal with them; why can't you? >...they are entitled to assume that the headers they rely on for >the purpose (References, Subject, Date) have been properly constructed, Nobody but YOU is talking about the syntax of the headers. The issue at hand is a mandate for CONTAINING a header. Said mandate does not exist. It is an etirely different issue to say "IF you contain header X, it MUST be constructed this way" than "you MUST contain header X". The former is without question. We do it all the time. The latter requires a justification for interoperability on the EXISTANCE of the header, not just ability to decode it if it there. I'm quite happy if you want to say "the References header MUST contain ...". RFC2822 already deals with the question of when References headers are mandatory, and the answer is "they are not." >No it doesn't. It says they should be used for "Replies", which is not >quite the same thing. Close enough. All replies are followups. When RFC2822 says that the header SHOULD appear in replies, then whatever REPLIES it excludes from a mandate are also FOLLOWUPs that are excluded. >Well you are giving a good impression of not caring :-( . And even if you >don't care, then I still do. No, Charles, you do not. When you promote the concept "non-followups MAY contain References header", you exhibit a complete lack of caring about the ability to detect what is and is not a followup. It is only when the References header is in the current (RFC1036) MUST/MUST NOT version can you use it to identify a followup. MUST/MAY throws that ability away. >> and 2) the existing standard does, >Which is a good reason for keeping it so, One cannot "keep" what one has already let go. draft-usefor-03 removed the mandate. It's gone. If you want it back, justify it. "existing standard says" wasn't sufficient to keep it in usefor-03, so it's not sufficient to put it back. "Charles says" is even less convincing. >... and we should not change it >without a clear agreement in this WG to do so. Given your history of making changes to the drafts without any input from the WG at all, much less a "clear agreement", this sudden religion you have gotten for sticking to the process is, shall I say, questionable. And considering that we are discussing a change to the draft that was made with NO input from the WG in the first place ... >>Currently, you cannot construct a well-formed USENET message that is a >>followup without a References header. That is a direct effect of the >>language in RFC1036 that says "It is required for all follow-up messages". >And it's a direct effect of what USEPRO says. Baloney. USEPRO is a DRAFT standard. It has no effect on existing requirements, only on what might be required some day maybe. When I say "currently ...", USEPRO is irrelevant. IF USEFOR-03 ever makes it to standard, then USEFOR will tell me that I CAN construct a well-formed USENET message that is a followup without a References header, because RFC2822 says I can. POOF goes your caring about what is and is not a followup. >>We are not debating what is and is not a followup here. >Yes we are. Maybe you are, but the topic of THIS discussion is the lack of mandate for a References header, and that has NOTHING to do with what is or is not a followup, only on whether there is a mandate for using that header in any circumstance. RFC2822 says SHOULD. That is not a mandate. End of story. >And the opinion I am hearing (even from you now) is that it is >whatever a followup agent produces. Stop putting words in my mouth, Charles. I'm getting fucking tired of it. This is no longer just a difference of opinion, it's you pretending to speak for me, and you do NOT have the right to do that. Period. From owner-ietf-usefor@mail.imc.org Fri Apr 22 18:40:11 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id SAA11124 for ; Fri, 22 Apr 2005 18:40:11 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3MMdBaU019347 for ; Fri, 22 Apr 2005 15:39:11 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3MMdBfl019346 for ietf-usefor-skb; Fri, 22 Apr 2005 15:39:11 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from ciao.gmane.org (main.gmane.org [80.91.229.2]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3MMd9hm019333 for ; Fri, 22 Apr 2005 15:39:09 -0700 (PDT) (envelope-from usenet-format@gmane.org) Received: from list by ciao.gmane.org with local (Exim 4.43) id 1DP6iK-0001JV-JN for ietf-usefor@imc.org; Sat, 23 Apr 2005 00:33:32 +0200 Received: from 212.82.251.181 ([212.82.251.181]) by main.gmane.org with esmtp (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Sat, 23 Apr 2005 00:33:32 +0200 Received: from nobody by 212.82.251.181 with local (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Sat, 23 Apr 2005 00:33:32 +0200 X-Injected-Via-Gmane: http://gmane.org/ To: ietf-usefor@imc.org From: Frank Ellermann Subject: Obsolete vs. obsolescent (was: Fwd: http://www.ietf.org/internet-drafts/draft-ietf-usefor-usefor-03.txt) Date: Sat, 23 Apr 2005 00:33:26 +0200 Organization: Lines: 23 Message-ID: <42697BB6.6074@xyzzy.claranet.de> References: <200504111212.j3BCChM14516@clerew.man.ac.uk> <87psx1gnjc.fsf@windlord.stanford.edu> <87fyxvx1m6.fsf@windlord.stanford.edu> <87mzs2a1hu.fsf@windlord.stanford.edu> <87wtqxui6s.fsf@windlord.stanford.edu> <87d5so2bpk.fsf@windlord.stanford.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Complaints-To: usenet@sea.gmane.org X-Gmane-NNTP-Posting-Host: 212.82.251.181 X-Mailer: Mozilla 3.0 (OS/2; U) Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Content-Transfer-Encoding: 7bit Charles Lindsey wrote: > "Obsolete" means this is no longer an official Netnews > header. It would be labelled as "obsolete" in the IANA > Registy of headers. Bruce has created a draft with mail headers found in obsolete RfCs. Among others Fcc: found in RfC 724 obsoleted by RfC 733. For news we'd declare See-Also: and Also-Control: as obsolete, as soon as Henry publishes s-o-1036 for information (obsoleted by - I'm an optimist here - draft-ietf-usefor-usefor-04.tx) > "Obsolescent" means that is is still an official Netnews > header (just), but that its continued use is discouraged > (for reasons that are given), and that it is likely to be > declared fully obsolete in a future standard. Now as pessimist: There won't be any future standard, UseFor is the last news standard before its death. So Lines: will be obsolescent "forever" until the last news server goes jabber, or atom over soap based on beep, or whatever comes next. Bye. From owner-ietf-usefor@mail.imc.org Sun Apr 24 12:51:20 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id MAA02289 for ; Sun, 24 Apr 2005 12:51:20 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3OGoC0A000969 for ; Sun, 24 Apr 2005 09:50:12 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3OGoC4J000968 for ietf-usefor-skb; Sun, 24 Apr 2005 09:50:12 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from rufus.isode.com (rufus.isode.com [62.3.217.251]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3OGnjcu000937 for ; Sun, 24 Apr 2005 09:50:11 -0700 (PDT) (envelope-from alexey.melnikov-usefor@isode.com) Received: from [192.168.0.3] ([62.3.217.253]) by rufus.isode.com via TCP (internal) with ESMTPA; Sun, 24 Apr 2005 17:49:24 +0100 Message-ID: <426BB63D.3060506@isode.com> Date: Sun, 24 Apr 2005 16:07:41 +0100 From: Alexey Melnikov User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.7.2) Gecko/20040804 Netscape/7.2 (ax) X-Accept-Language: en-us, en MIME-Version: 1.0 To: Charles Lindsey CC: ietf-usefor@imc.org, Pete Resnick Subject: Re: Fixed (was: Broken Message-ID syntax) References: <4255153E.3080500@isode.com> <42594879.3B92@xyzzy.claranet.de> <425A8A63.268A@xyzzy.claranet.de> <425C607E.1B5A@xyzzy.claranet.de> <425D8031.D57@xyzzy.claranet.de> In-Reply-To: Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Content-Transfer-Encoding: 7bit Charles Lindsey wrote: >In <425D8031.D57@xyzzy.claranet.de> Frank Ellermann writes: > > >>Charles Lindsey wrote: >> >> >>>Yes, you have convinced me now. >>> >>> >>Great, we have a msg-id syntax. >> >> >>>there remains the issue of the naming of the syntax rules, >>> >>> >>We're talking about seven names. So far we agree on msg-id, >>and we disagree on names for LHS, RHS, and address-literal. >> >> >>Maybe we can solve it for the remaining three names without >>bothering Alexey, Henry, or Ken: >> >> > >Essentially, we either use the same names as RFC 2822 uses, saying that >"the following syntax rules replace the corresponding rules in RFC 2822" >(giving us id-left, id-right, no-fold-quote, no-fold literal, etc), or >else we depart from the RFC 2822 names entirely, except for the msg-id at >the top, in which case your names are as good as any. I don't see that >there is anything in-between. > >My preference is to retain the RFC 2822 naming, which accords with the >general principle that we do not change anything in RFC 2822 that we do >not need to change. > I concur with your preference. >If our Chair says that we can change these names, then >let it be so. Alexey? > > If WG members believe that better names should be used, I suggest they talk to Pete Resnick first. As far as I remember he has already agreed to publish 2822bis. Alexey From owner-ietf-usefor@mail.imc.org Sun Apr 24 12:51:27 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id MAA02307 for ; Sun, 24 Apr 2005 12:51:27 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3OGoC3Q000977 for ; Sun, 24 Apr 2005 09:50:12 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3OGoCSw000976 for ietf-usefor-skb; Sun, 24 Apr 2005 09:50:12 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from rufus.isode.com (rufus.isode.com [62.3.217.251]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3OGnjcv000937 for ; Sun, 24 Apr 2005 09:50:12 -0700 (PDT) (envelope-from alexey.melnikov-usefor@isode.com) Received: from [192.168.0.3] ([62.3.217.253]) by rufus.isode.com via TCP (internal) with ESMTPA; Sun, 24 Apr 2005 17:49:26 +0100 Message-ID: <426BB84C.8060309@isode.com> Date: Sun, 24 Apr 2005 16:16:28 +0100 From: Alexey Melnikov User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.7.2) Gecko/20040804 Netscape/7.2 (ax) X-Accept-Language: en-us, en MIME-Version: 1.0 To: Russ Allbery CC: ietf-usefor@imc.org Subject: Re: Differences between RFC 2822 and Usefor (fwd) References: <87ekd5t0ae.fsf@windlord.stanford.edu> In-Reply-To: <87ekd5t0ae.fsf@windlord.stanford.edu> Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Content-Transfer-Encoding: 7bit Russ Allbery wrote: >John Stanley writes: > > >>USEFOR-03 says it is defined as found in RFC2822, and is >>optional. Optional means "not mandatory". There is no MUST. >> >> > >I think it's obvious that the References header has to be optional in that >sense, since you can construct a well-formed Usenet message that doesn't >have a References header. > >I'll repeat my earlier suggestion here, which is to simply define >followups in such a way as to say that if something doesn't have a >References header, it's not a followup. That neatly avoids the whole >issue without needing to worry about MUSTs and SHOULDs. It becomes a >definitional issue. If it doesn't have References, it's not a followup. > > I fully agree. >>Like I already said, once you remove the ability to detect followups by >>making References optional, it doesn't matter when it may or may not be >>used. In fact, it may be used anytime (that's what "optional" means), >>even apparently when there is nothing at all for it to convey. >> >> >>By changing the requirement for References, we've solve a whole passle >>of things that should be be CHANGED BACK to the way they were before you >>decided to take it upon yourself to change them. Like "non-followups MAY >>contain" References headers, which, despite your repeated claims to the >>contrary, is a major change that nobody asked for. >> >> > >Well, to be fair, I think it's a reasonable way of dealing with the issue. >I'm not sure if that means I was asking for it, but I can see the logic >behind it. > >The way I look at this, there are two separate issues that we keep mixing >together: > > * What a news reader should do when constructing a posted reply to a > particular post. > > * How a news reader should display messages with References headers. > >The former is what we mean by a followup, and by definition it has to >build a References header. If it doesn't, it's doing something other than >posting a followup. > >The latter is a separate issue, a UI issue. It is expected that messages >connected by References will be shown in a thread, subject to various user >configuration and choices like whether changing the Subject should result >in a separate thread in the user's display (something that should be a >configuration option, not something the standard should be mandating >either way, as there are people who prefer it both ways and there is no >interoperability issue). > >There are various reasons why sometimes people like to connect messages >that are not posted replies to each other, and that do not, or only barely >do, fit into the notion of a followup. > >I don't think there's really anything else we can say that's more specific >than that. > > I agree. And my guess is that RFC 2822 says so little on the subject exactly for the same reasons. From owner-ietf-usefor@mail.imc.org Sun Apr 24 12:51:28 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id MAA02325 for ; Sun, 24 Apr 2005 12:51:28 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3OGoDWQ000986 for ; Sun, 24 Apr 2005 09:50:13 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3OGoDp2000985 for ietf-usefor-skb; Sun, 24 Apr 2005 09:50:13 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from rufus.isode.com (rufus.isode.com [62.3.217.251]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3OGnjcw000937 for ; Sun, 24 Apr 2005 09:50:12 -0700 (PDT) (envelope-from alexey.melnikov-usefor@isode.com) Received: from [192.168.0.3] ([62.3.217.253]) by rufus.isode.com via TCP (internal) with ESMTPA; Sun, 24 Apr 2005 17:49:29 +0100 Message-ID: <426BC455.2040108@isode.com> Date: Sun, 24 Apr 2005 17:07:49 +0100 From: Alexey Melnikov User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.7.2) Gecko/20040804 Netscape/7.2 (ax) X-Accept-Language: en-us, en MIME-Version: 1.0 To: Frank Ellermann CC: ietf-usefor@imc.org Subject: Re: Obsolete vs. obsolescent References: <200504111212.j3BCChM14516@clerew.man.ac.uk> <87psx1gnjc.fsf@windlord.stanford.edu> <87fyxvx1m6.fsf@windlord.stanford.edu> <87mzs2a1hu.fsf@windlord.stanford.edu> <87wtqxui6s.fsf@windlord.stanford.edu> <87d5so2bpk.fsf@windlord.stanford.edu> <42697BB6.6074@xyzzy.claranet.de> In-Reply-To: <42697BB6.6074@xyzzy.claranet.de> Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Content-Transfer-Encoding: 7bit Frank Ellermann wrote: >Charles Lindsey wrote: > > >>"Obsolete" means this is no longer an official Netnews >>header. It would be labelled as "obsolete" in the IANA >>Registy of headers. >> >> > >Bruce has created a draft with mail headers found in obsolete >RfCs. Among others Fcc: found in RfC 724 obsoleted by RfC 733. > >For news we'd declare See-Also: and Also-Control: as obsolete, >as soon as Henry publishes s-o-1036 for information (obsoleted >by - I'm an optimist here - draft-ietf-usefor-usefor-04.tx) > > In order to close this issue I suggest that the USEFOR document can include few sentences describing what do we mean when calling a particular header "obsolete". >>"Obsolescent" means that is is still an official Netnews >>header (just), but that its continued use is discouraged >>(for reasons that are given), and that it is likely to be >>declared fully obsolete in a future standard. >> >> > >Now as pessimist: There won't be any future standard, UseFor >is the last news standard before its death. So Lines: will be >obsolescent "forever" > I tend to share your pessimism, but maybe for a slightly different reason. > until the last news server goes jabber, >or atom over soap based on beep, or whatever comes next. Bye. > > From owner-ietf-usefor@mail.imc.org Mon Apr 25 12:15:45 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id MAA28750 for ; Mon, 25 Apr 2005 12:15:44 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3PGDsac038040 for ; Mon, 25 Apr 2005 09:13:54 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3PGDsKA038039 for ietf-usefor-skb; Mon, 25 Apr 2005 09:13:54 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from lon-mail-2.gradwell.net (lon-mail-2.gradwell.net [193.111.201.126]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3PGDrMO038024 for ; Mon, 25 Apr 2005 09:13:54 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from host81-144-65-120.midband.mdip.bt.net ([81.144.65.120]) by lon-mail-2.gradwell.net with esmtp (Gradwell gwh-smtpd 1.181) id 426d1740.e3c.3a for ietf-usefor@imc.org; Mon, 25 Apr 2005 17:13:52 +0100 (envelope-sender ) Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3PGCFg21653 for ietf-usefor@imc.org; Mon, 25 Apr 2005 17:12:15 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20719 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: References Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: Date: Mon, 25 Apr 2005 14:16:34 GMT Lines: 69 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In John Stanley writes: >"Charles Lindsey" : >Unfortunately, RFC2822 says only SHOULD regarding the References header >and followups, Please show me where the word "followup" is to be found in RFC 2822. If you use your newsreader to send an Email reply to the author of a News article, then indeed you (only) SHOULD include a References header (and In-Reply-To as well to do the job properly). If you use your newsreader to post a News followup to a News article (usually in the same newsgroups as the precursor) then RFC 2822 is silent on the issue (because the whole purpose of RFC 2822 is to tell you how to construct a proper Email. USEFOR incorparates, by reference, many concepts from RFC 2822, including many headers and their meanings (semantics). But nowhere does USEFOR equate the concepts of "Email Reply" with "Netnews Followup". Therefore they remain different concepts with different rules. I grant you that the concepts are sufficiently similar that, wherever we say that News Followups are different in some way from Email Replies (e.g. MUST rather than SHOULD), then we need to draw attention to this. Our Chair has ruled that the correct place to say such things is in USEPRO rather than USEFOR. >You are apparently incapable of differentiating between "poster intends a >followup in what he sends" and "reader detects a followup in what he >gets." The previous standards and drafts have said that the intent for >something to be a followup by a poster REQUIRES a References header to be >inserted (MUST), and otherwise that header is prohibited. That means that >the READER can then determine what the poster intends to be a followup by >looking for a References header. If a poster intends a followup, then a References header is REQUIRED. If the poster does not intend a follouwp, but nevertheless includes a References header (e.g. in a multi-part FAQ, or to stitch together a message/partial), then what is the effect? >And yes, Charles, when you say "non-followups MAY contain a References >header", you have removed that ability to DETECT at the reader's end of >the pipe what is and is not intended to be a followup. That is a change, >despite your repeated denials. Exactly. That has never been denied. You cannot tell from looking at the References header whether the poster intended to write a followup, or merely intended that it should be threaded like a followup (with all the other benefits accruing from a References header). So f*****g what? Why should the reader care? So long as the reader (or his agent) does the Right Thing (as intended by the poster in both cases), why should anybody get upset? -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 From owner-ietf-usefor@mail.imc.org Mon Apr 25 12:15:54 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id MAA28771 for ; Mon, 25 Apr 2005 12:15:54 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3PGDsAZ038032 for ; Mon, 25 Apr 2005 09:13:54 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3PGDs29038031 for ietf-usefor-skb; Mon, 25 Apr 2005 09:13:54 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from lon-mail-2.gradwell.net (lon-mail-2.gradwell.net [193.111.201.126]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3PGDrdc038023 for ; Mon, 25 Apr 2005 09:13:53 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from host81-144-65-120.midband.mdip.bt.net ([81.144.65.120]) by lon-mail-2.gradwell.net with esmtp (Gradwell gwh-smtpd 1.181) id 426d173f.e3c.39 for ietf-usefor@imc.org; Mon, 25 Apr 2005 17:13:51 +0100 (envelope-sender ) Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3PGCFi21649 for ietf-usefor@imc.org; Mon, 25 Apr 2005 17:12:15 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20718 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: References Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: <4266EF35.21A5@xyzzy.claranet.de> <42686E91.42E2@xyzzy.claranet.de> Date: Mon, 25 Apr 2005 13:47:06 GMT Lines: 48 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In <42686E91.42E2@xyzzy.claranet.de> Frank Ellermann writes: >Charles Lindsey wrote: >> How do we define the term "followup". >| A "followup" is an article containing a response to the >| contents of an earlier article (the followup's "precursor"). >Found in usefor-03. Or as Alternative-1 in USEPRO. >> or is it anything that has a References header. Which is Alternative-2 in USEPRO. I prefer Alternative-1, and I think most people are now of that view (modulo minor wording tweaks). >Not sure. If it's the RfC 2822 idea of a reply coming from a >mail2news gateway it probably is some "followup", but behind a >gateway you never know. Maybe it only had an In-Reply-To, the >gateway tried to fix it, but it was a broken pipermail archive >In-ReplyTo (the thread is then correct, but the "followup" is >shown at the wrong place in the thread). And lots of other ways it might have arisen. As Russ has said, referring to some of those cases as followups would be "artificial". > A "followup" is an article containing a response to the > contents of an earlier article (the followup's "precursor" > indicated in the "References" header field). Yes, I had already concluded that something like that would be helpful in the definition. Russ seems to agree. -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 From owner-ietf-usefor@mail.imc.org Mon Apr 25 15:01:23 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id PAA11773 for ; Mon, 25 Apr 2005 15:01:23 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3PJ03Ir063414 for ; Mon, 25 Apr 2005 12:00:03 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3PJ03PI063413 for ietf-usefor-skb; Mon, 25 Apr 2005 12:00:03 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from lon-mail-1.gradwell.net (lon-mail-1.gradwell.net [193.111.201.125]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3PJ01hR063403 for ; Mon, 25 Apr 2005 12:00:02 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from host81-144-77-174.midband.mdip.bt.net ([81.144.77.174]) by lon-mail-1.gradwell.net with esmtp (Gradwell gwh-smtpd 1.181) id 426d3e2f.2b99.6b for ietf-usefor@imc.org; Mon, 25 Apr 2005 19:59:59 +0100 (envelope-sender ) Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3PIv6522990 for ietf-usefor@imc.org; Mon, 25 Apr 2005 19:57:06 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20720 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: Fixed (was: Broken Message-ID syntax) Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: <4255153E.3080500@isode.com> <42594879.3B92@xyzzy.claranet.de> <425A8A63.268A@xyzzy.claranet.de> <425C607E.1B5A@xyzzy.claranet.de> <425D8031.D57@xyzzy.claranet.de> <426BB63D.3060506@isode.com> Date: Mon, 25 Apr 2005 16:17:30 GMT Lines: 131 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In <426BB63D.3060506@isode.com> Alexey Melnikov writes: >Charles Lindsey wrote: >>My preference is to retain the RFC 2822 naming, which accords with the >>general principle that we do not change anything in RFC 2822 that we do >>not need to change. >> >I concur with your preference. In that case, we are ready for some syntax. I have given below what I now think should be in the Message-ID section of USEFOR, including the necessary syntax changes and rewordings of the descriptive text. I have also taken the liberty of removing in favour of , which I think we agreed in some earlier discussions on this list (there will be a few consequential changes elsewhere in USEFOR). Also, I have suggested a final paragraph regarding use of domains in the , which closes a possible shortcoming in the wording of RFC 2822, as spotted by Frank. >> >If WG members believe that better names should be used, I suggest they >talk to Pete Resnick first. As far as I remember he has already agreed >to publish 2822bis. I don't think Pete has gone beyond agreeing in principle to 2822bis. Has he actually published any Internet Drafts? If so, I will be needing to comment upon them. Here follows my suggested Message-ID text: 3.1.3 Message-ID The Message-ID header contains a single unique message identifier. This document updates the construct from Section 3.6.4 of [RFC2822] so as to ensure that Internet Message Format Message-IDs are usable in widely deployed news software. The global uniqueness requirement for in [RFC2822] is to be understood as applying across all protocols using such message identifiers, and across both Email and Netnews in particular. A revised syntax for is given below, but the requirements and descriptive text from Section 3.6.4 of [RFC2822] still apply. | message-id = "Message-ID:" SP [FWS] msg-id [FWS] CRLF | | msg-id = "<" id-left "@" id-right ">" | ; maximum length is 250 octets id-left = dot-atom-text / no-fold-quote id-right = dot-atom-text / no-fold-literal | no-fold-quote = DQUOTE | ( "." *mqtext / | *mqtext "." / | *mqtext mqspecial *mqtext ) | DQUOTE | | mqtext = atext / "." / mqspecial | mqspecial = "(" / ")" / ; same as specials except "<" / ; "\" and DQUOTE quoted | "[" / "]" / ; "." doubled and ">" omitted ":" / ";" / "@" / "," / | ".." / "\\" / "\" DQUOTE no-fold-literal = "[" *( mdtext / "\[" / "\]" / "\\" ) "]" | mdtext = %d33-61 / ; The rest of the US-ASCII %d63-90 / ; characters not including %d94-126 ; ">", "[", "]", or "\" The msg-id-core MUST NOT be more than 250 octets in length. NOTE: The length restriction ensures that systems which accept message identifiers as a parameter when retrieving an article (e.g. [NNTP]) can rely on a bounded length. | Observe that includes the < and >. Observe also that in contrast to the corresponding header in | [RFC2822] | o the syntax does not allow comments within the Message-ID | header, | o it ensures that no string of characters is quoted if it was already | a (it must start or end with a ".", or contain at | least one mqspecial), | o it ensures that no single character is prefixed by a "\" in the | form of a unless strictly necessary, | o it excludes all control characters, and | o there is no possibility for ">" or WSP to occur inside a | , whether quoted or not. | | This is to simplify processing by relaying and serving agents, and to | ensure interoperability with existing implementations and compliance | with [NNTP]. Thus, whereas under [RFC2822] the following s | would be considered semantically equivalent, | | <"ab.cd"@example.com> | <"ab.\cd"@example.com> only the first of them is syntactically permitted by this standard, and hence a simple comparison of octets will always suffice to determine the identity of two s. | Also note that this updated ABNF applies wherever is used, | including the References header discussed in Section 3.2.1 and the | Supersedes header discussed in Section 3.2.5. | NOTE: It in RECOMMENDED in [RFC 2822] that, for ensuring global | uniqueness, the be some domain identifier within whose | scope the uniqueness of the can be guaranteed. When | following this recommendation, any or | used for the are to be interpreted as | s as described in section 3.4.1 of [RFC 2822]. -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 From owner-ietf-usefor@mail.imc.org Mon Apr 25 15:16:11 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id PAA13795 for ; Mon, 25 Apr 2005 15:16:10 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3PJFA0w065996 for ; Mon, 25 Apr 2005 12:15:10 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3PJFA05065995 for ietf-usefor-skb; Mon, 25 Apr 2005 12:15:10 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from a.mail.peak.org (a.mail.peak.org [69.59.192.41]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3PJF9u2065955 for ; Mon, 25 Apr 2005 12:15:09 -0700 (PDT) (envelope-from stanley@peak.org) Received: from a.shell.peak.org ([69.59.192.81]) by a.mail.peak.org (8.12.10/8.12.8) with ESMTP id j3PJF0fO064514 (version=TLSv1/SSLv3 cipher=EDH-RSA-DES-CBC3-SHA bits=168 verify=NO) for ; Mon, 25 Apr 2005 12:15:02 -0700 (PDT) Date: Mon, 25 Apr 2005 12:15:00 -0700 (PDT) From: John Stanley X-X-Sender: stanley@a.shell.peak.org To: ietf-usefor@imc.org Subject: Re: References Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Spam-Score: 0 () X-Scanned-By: MIMEDefang 2.39 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: "Charles Lindsey" : >Please show me where the word "followup" is to be found in RFC 2822. I won't bother doing so, because we've already covered RFC2822 using the term "reply" in the relevant section, and that all replies are followups, using any common-sense or even news-based definition of the word followup. Not all followups are necessarily replies, but every "reply" is "an article posted in response to another". Since RFC2822 clearly does NOT mandate References headers in all replies, then it just as clearly exempts some followups -- the same replies that have no mandate for its use are the same followups that have no mandate. You cannot say that all followups MUST have a References header without also saying all replies MUST, since the latter is a proper subset of the former. > USEFOR incorparates, by reference, many concepts from RFC 2822, Including, explicitely, the References header. Thus RFC2822 not only tells us how to construct the "proper Email [sic]" that you grant in your previous paragraph, but a "proper News" article, because USEFOR says that RFC2822 applies to news, too. >But nowhere does USEFOR >equate the concepts of "Email Reply" with "Netnews Followup". It doesn't need to. It is quite disengenuous of you to pretend that RFC2822 says "Email Reply" in the section about References headers, and to pretend that there has to be some equating of the two phrases for RFC2822 to apply. I've quoted the specific parts of RFC2822 that talk about "replies", and they do not limit themselves to email, and USEFOR-03 explicitely expands any limit of "Email Reply" that you may fictionalize for the References header and RFC2822 to "News reply", as well. And you know very well that the two concepts do not have to be equivalent for one to apply to the other. All replies are followups using standard common-sense and common-practice news definitions for the same. If you say, as RFC2822 does, and draft-usefor-03 brings into the news context, that not all replies require a References header, then you just as clearly say not all followups require a References header. This is now the third time in two days I've covered this ground for you, Charles. >I grant you that the concepts are sufficiently similar that, wherever we >say that News Followups are different in some way from Email Replies (e.g. >MUST rather than SHOULD), Except we DO NOT say this. We EXPLICITELY say that the RFC2822 definition of References applies (with a few small syntax differences) but we do NOT say that one of the differences is that References is a MUST. You can say that you grant that the concepts are similar, even though you have just argued that they are completely different and thus statements about one thing do not apply to the other, but you cannot grant that we say that there is a mandate, because draft-usefor-03 and RFC2822 clearly says otherwise. >Our Chair has ruled that the correct place to say such things is in USEPRO >rather than USEFOR. Then the Chair has effectively ruled that References is optional as defined in RFC2822 and that "followup" is a meaningless term, since nobody past the poster can identify what is and is not a followup using the optional definition of References. If the Chair thinks that we can say in one news standard that References is an optional header (by referring to the SHOULD appear and "may appear" definition in RFC2822) and then change our minds in mid-stream by saying that it is MANDATORY in another news standard, then he is demonstrating schizophrenia. Either it IS mandatory in followups or it is not. I'll be happy with whichever position we take: either RFC1036 had it right and References is MUST/MUST NOT and followups are thus identifiable; or RFC1036 has it wrong, References is SHOULD/MAY and followups are not identifiable and thus a meaningless term. But since it has already been changed, it now requires justification for bringing back the MUST/anything language, since RFC2119 clearly limits such language to serious matters, and we can no longer use the 'it was pre-existing' excuse. It's not existing anymore; can you justify bringing it back? I'll even suggest that if you cannot justify the return of a mandate (and do not just unilaterally reinsert it as is your wont to do with our work products), we can create a specific header to replace References, call it "Thread-Me-Next-To:" or "Display-Near:" or something similar that makes it clear this is a display issue and not a protocol one, and make that optional. That's fine. It's being honest then about the value of the information and not pretending it is some critical protocol interoperability thing we're trying to solve. And before you start playing the "show me where it says X" game with RFC1036 and MUST/MUST NOT, I'll point out that RFC1036 predates RFC2119 but it uses clear terms like "required" and "prohibited", which map into RFC2119 MUST/MUST NOT. But we've been through that before, too, so I don't expect this preemptive note to have any effect on your argument. >If a poster intends a followup, then a References header is REQUIRED. No. Refer to RFC2822, which says that References is only SHOULD and "may", and even talks about messages that are clearly replies (contains In-Reply-To) but has no References. >> That is a change, despite your repeated denials. >Exactly. That has never been denied. You've repeatedly denied it. You keep saying we can use your proposed wording because it makes no difference from what was there originally, regarding MUST/MUST NOT and followups. There's a difference from how things are now. There's a change. >You cannot tell from looking at the >References header whether the poster intended to write a followup, or >merely intended that it should be threaded like a followup (with all the >other benefits accruing from a References header). Let's put back some context to this, shall we? If you put back the MUST/MUST NOT dichotomy for the References header, which has been removed from draft-usefor-03 now (but still exists in news today due to RFC1036), then you can tell precisely if the poster intended a followup or not because a followup MUST have the header and a non-followup MUST NOT. (The third option is the poster is stupid/incompetent/ignorant and cannot properly form a news article, but let's stick with properly formattted news articles and not try to guess at the meaning of broken ones, ok?) It is only when you water down the References header as it has now been that you cannot tell what was intended. An article with a References header (under draft-usefor-03) might be a followup or it might not. An article without a References header might not be a followup, or it might be. You have no way of knowing. You can only guess. Technical standards are broken if they involve guessing about things that are supposed to be important, especially when they could easily be written to remove the guessing. >So f*****g what? Why should the reader care? If it is not important to the reader, then the References header absolutely does not merit any RFC2119 language regarding its use, much less MUST. It is not an interoperability issue in any case, and if the information that it conveys is "so fucking what?" for the reader, then lets make it obsolete and be done with it. I'm fine with that. References is obsolete. The term "followup" is removed from the drafts. Let's git-r-done. When can we expect you to produce the next version with these changes we've agreed to? From owner-ietf-usefor@mail.imc.org Tue Apr 26 10:06:07 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id KAA10453 for ; Tue, 26 Apr 2005 10:06:07 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3QE3EH2082286 for ; Tue, 26 Apr 2005 07:03:14 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3QE3Ecc082285 for ietf-usefor-skb; Tue, 26 Apr 2005 07:03:14 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from eagle.oceana.com (eagle.oceana.com [208.17.123.12]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3QE3DCm082276 for ; Tue, 26 Apr 2005 07:03:13 -0700 (PDT) (envelope-from ken@oceana.com) Received: from [192.168.10.26] (KEN.oceana.com [192.168.10.26]) by eagle.oceana.com (8.13.2/8.13.2) with ESMTP id j3QE36Gc008111 for ; Tue, 26 Apr 2005 10:03:06 -0400 (EDT) Message-ID: <426E49EB.3010709@oceana.com> Date: Tue, 26 Apr 2005 10:02:19 -0400 From: Ken Murchison Organization: Oceana Matrix Ltd. User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.7) Gecko/20040514 X-Accept-Language: en,pdf MIME-Version: 1.0 To: ietf-usefor@imc.org Subject: Re: Fwd: http://www.ietf.org/internet-drafts/draft-ietf-usefor-usefor-03.txt References: <200504111212.j3BCChM14516@clerew.man.ac.uk> In-Reply-To: <200504111212.j3BCChM14516@clerew.man.ac.uk> Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit X-Spam-Score: 0 (BAYES_00) X-Scanned-By: MIMEDefang 2.39 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Content-Transfer-Encoding: 7bit Charles Lindsey wrote: > In local.usefor you write: > > >>------- Forwarded message ------- >>From: "Dr John Stockton" >>To: "Charles. H. Lindsey" >>Subject: >>http://www.ietf.org/internet-drafts/draft-ietf-usefor-usefor-03.txt >>Date: Sun, 10 Apr 2005 19:44:55 +0100 > > > >>"1.1 Basic Concepts >> "Netnews" is a set of protocols for generating, storing and >> retrieving news "articles" (which are a subset of Email messages)" > > >>They are surely not a subset of E-mail messages. They might be >>described as similar. But E-mail is now commonly HTML & News should not >>be; I'd omit the parenthetic analogy. > > > Yes, it might be better to say "(whose format is a subset of that for > Email messages)". Ken? Yes, I'll make this change. -- Kenneth Murchison Oceana Matrix Ltd. Software Engineer 21 Princeton Place 716-662-8973 x26 Orchard Park, NY 14127 --PGP Public Key-- http://www.oceana.com/~ken/ksm.pgp From owner-ietf-usefor@mail.imc.org Tue Apr 26 10:10:18 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id KAA11060 for ; Tue, 26 Apr 2005 10:10:17 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3QE90HC083430 for ; Tue, 26 Apr 2005 07:09:00 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3QE905H083429 for ietf-usefor-skb; Tue, 26 Apr 2005 07:09:00 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from eagle.oceana.com (eagle.oceana.com [208.17.123.12]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3QE8xII083408 for ; Tue, 26 Apr 2005 07:09:00 -0700 (PDT) (envelope-from ken@oceana.com) Received: from [192.168.10.26] (KEN.oceana.com [192.168.10.26]) by eagle.oceana.com (8.13.2/8.13.2) with ESMTP id j3QE8r4j008207; Tue, 26 Apr 2005 10:08:53 -0400 (EDT) Message-ID: <426E4B45.7010205@oceana.com> Date: Tue, 26 Apr 2005 10:08:05 -0400 From: Ken Murchison Organization: Oceana Matrix Ltd. User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.7) Gecko/20040514 X-Accept-Language: en,pdf MIME-Version: 1.0 To: Alexey Melnikov CC: ietf-usefor@imc.org Subject: Re: Obsolete vs. obsolescent References: <200504111212.j3BCChM14516@clerew.man.ac.uk> <87psx1gnjc.fsf@windlord.stanford.edu> <87fyxvx1m6.fsf@windlord.stanford.edu> <87mzs2a1hu.fsf@windlord.stanford.edu> <87wtqxui6s.fsf@windlord.stanford.edu> <87d5so2bpk.fsf@windlord.stanford.edu> <42697BB6.6074@xyzzy.claranet.de> <426BC455.2040108@isode.com> In-Reply-To: <426BC455.2040108@isode.com> Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit X-Spam-Score: 0 (BAYES_00) X-Scanned-By: MIMEDefang 2.39 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Content-Transfer-Encoding: 7bit Alexey Melnikov wrote: >> > In order to close this issue I suggest that the USEFOR document can > include few sentences describing what do we mean when calling a > particular header "obsolete". Care to craft some text and tell me where you'd like to see it? -- Kenneth Murchison Oceana Matrix Ltd. Software Engineer 21 Princeton Place 716-662-8973 x26 Orchard Park, NY 14127 --PGP Public Key-- http://www.oceana.com/~ken/ksm.pgp From owner-ietf-usefor@mail.imc.org Tue Apr 26 10:14:17 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id KAA11536 for ; Tue, 26 Apr 2005 10:14:17 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3QECeWq083990 for ; Tue, 26 Apr 2005 07:12:40 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3QECexc083989 for ietf-usefor-skb; Tue, 26 Apr 2005 07:12:40 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from eagle.oceana.com (eagle.oceana.com [208.17.123.12]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3QECdim083972 for ; Tue, 26 Apr 2005 07:12:40 -0700 (PDT) (envelope-from ken@oceana.com) Received: from [192.168.10.26] (KEN.oceana.com [192.168.10.26]) by eagle.oceana.com (8.13.2/8.13.2) with ESMTP id j3QECXs1008268 for ; Tue, 26 Apr 2005 10:12:33 -0400 (EDT) Message-ID: <426E4C22.1040808@oceana.com> Date: Tue, 26 Apr 2005 10:11:46 -0400 From: Ken Murchison Organization: Oceana Matrix Ltd. User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.7) Gecko/20040514 X-Accept-Language: en,pdf MIME-Version: 1.0 To: ietf-usefor@imc.org Subject: Re: Differences between RFC 2822 and Usefor References: In-Reply-To: Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit X-Spam-Score: 0 (BAYES_00) X-Scanned-By: MIMEDefang 2.39 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Content-Transfer-Encoding: 7bit John Stanley wrote: > > "Charles Lindsey" : > > >>And why does the Usefor draft contain the words > > >> None of the headers appearing in this section is required to appear >> in every article but some of them are required in certain types of >> article, such as followups. > > > Because someone (an editor, perhaps?) overlooked this statement when > removing the mandate for References? Remove the "such as" clause and the > problem is fixed. Do it and lets move on. Problem solved. Done. -- Kenneth Murchison Oceana Matrix Ltd. Software Engineer 21 Princeton Place 716-662-8973 x26 Orchard Park, NY 14127 --PGP Public Key-- http://www.oceana.com/~ken/ksm.pgp From owner-ietf-usefor@mail.imc.org Tue Apr 26 10:19:05 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id KAA12132 for ; Tue, 26 Apr 2005 10:19:05 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3QEHG3v084678 for ; Tue, 26 Apr 2005 07:17:16 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3QEHGlD084677 for ietf-usefor-skb; Tue, 26 Apr 2005 07:17:16 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from eagle.oceana.com (eagle.oceana.com [208.17.123.12]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3QEHFgk084662 for ; Tue, 26 Apr 2005 07:17:15 -0700 (PDT) (envelope-from ken@oceana.com) Received: from [192.168.10.26] (KEN.oceana.com [192.168.10.26]) by eagle.oceana.com (8.13.2/8.13.2) with ESMTP id j3QEH8Le008349 for ; Tue, 26 Apr 2005 10:17:08 -0400 (EDT) Message-ID: <426E4D35.6080409@oceana.com> Date: Tue, 26 Apr 2005 10:16:21 -0400 From: Ken Murchison Organization: Oceana Matrix Ltd. User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.7) Gecko/20040514 X-Accept-Language: en,pdf MIME-Version: 1.0 To: ietf-usefor@imc.org Subject: Re: Differences between RFC 2822 and Usefor (fwd) References: In-Reply-To: Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit X-Spam-Score: 0 (BAYES_00) X-Scanned-By: MIMEDefang 2.39 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Content-Transfer-Encoding: 7bit Charles Lindsey wrote: > In John Stanley writes: > > >>"Charles Lindsey" : > > >>>The "MUST" in question arises from the long held consensus in this Working >>>Group that the References header in News is NOT an "optional extra". > > >>Not according to draft-usefor-03. According to draft-usefor-03, >>References is defined to be the same as in RFC2822, with the addition of a >>few bits that are not mandates for using it. This consensus you claim >>exists has a very odd way of being expressed in our products. Almost as if >>it doesn't really exist. > > >>The fact that I am the only one who even commented on it being gone is a >>pretty good sign that nobody cares that it is gone. > > > No, I knew that it wasn't there, and there are a lot of other things that > should be in the Usefor draft and that have not made it there yet. But I > do not want to be seen arguing with my co-editor in public if it can be > avoided. Especially on issues that do not appear to be the subject of > controversy. > > > >>>That clear requirement has been in all our drafts up to article-13. > > >>It is no longer a requirement at all in our draft. > > > Indeed, but a draft is only a draft. There has been no WG decision to > change the requirement for a References header in followups. Our Chair has > taken the view that Usefor is not the place to say it, but that Usepro IS > the proper place (and indeed there is already a corresponding MUST in > Usepro, though it may need a little more amplification). I do not agree > with the Chair on this one, but if nobody else speaks up and if his > ruling stands, then Usepro is where it will be said (which also, > incidentally, removes most of the cause of the disagreement there has been > about how to word it all). I am in agreement with the Chair that this discussion belongs in USEPRO (I thought we had discussed this offline, but perhaps not). -- Kenneth Murchison Oceana Matrix Ltd. Software Engineer 21 Princeton Place 716-662-8973 x26 Orchard Park, NY 14127 --PGP Public Key-- http://www.oceana.com/~ken/ksm.pgp From owner-ietf-usefor@mail.imc.org Tue Apr 26 10:27:01 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id KAA12802 for ; Tue, 26 Apr 2005 10:27:01 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3QEPYje085824 for ; Tue, 26 Apr 2005 07:25:34 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3QEPYVo085823 for ietf-usefor-skb; Tue, 26 Apr 2005 07:25:34 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from eagle.oceana.com (eagle.oceana.com [208.17.123.12]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3QEPXJo085813 for ; Tue, 26 Apr 2005 07:25:34 -0700 (PDT) (envelope-from ken@oceana.com) Received: from [192.168.10.26] (KEN.oceana.com [192.168.10.26]) by eagle.oceana.com (8.13.2/8.13.2) with ESMTP id j3QEPRw1008512 for ; Tue, 26 Apr 2005 10:25:27 -0400 (EDT) Message-ID: <426E4F27.2020607@oceana.com> Date: Tue, 26 Apr 2005 10:24:39 -0400 From: Ken Murchison Organization: Oceana Matrix Ltd. User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.7) Gecko/20040514 X-Accept-Language: en,pdf MIME-Version: 1.0 To: ietf-usefor@imc.org Subject: Re: References References: <4266EF35.21A5@xyzzy.claranet.de> <87mzrtt1fh.fsf@windlord.stanford.edu> In-Reply-To: Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit X-Spam-Score: 0 (BAYES_00) X-Scanned-By: MIMEDefang 2.39 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Content-Transfer-Encoding: 7bit Charles Lindsey wrote: > In <87mzrtt1fh.fsf@windlord.stanford.edu> Russ Allbery writes: > > >>Sounds right to me. > > >>If it doesn't have References, it's not a follow-up, but in practice >>people use References for things other than follow-ups, and I don't see >>any reason to ban those uses. So the question is whether we want to >>artificially define all those things as follow-ups as well, or not worry >>about it and just say that References is required for follow-ups but may >>also be used for other things. > > > Exactly. Those other uses for the References header deserve a brief > mention somewhere, probably in USEFOR, but it would indeed be somewhat > artificial to define those cases as "followups", and I personally do not > want to do that. I disagree. It is my opinion (and I believe our Chair's) that USEFOR only document the syntax of an article. The use or non-use of any particular "optional" header such as References belongs in either USEPRO or USEAGE. -- Kenneth Murchison Oceana Matrix Ltd. Software Engineer 21 Princeton Place 716-662-8973 x26 Orchard Park, NY 14127 --PGP Public Key-- http://www.oceana.com/~ken/ksm.pgp From owner-ietf-usefor@mail.imc.org Tue Apr 26 10:30:16 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id KAA12971 for ; Tue, 26 Apr 2005 10:30:15 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3QESG9c086177 for ; Tue, 26 Apr 2005 07:28:16 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3QESGqX086176 for ietf-usefor-skb; Tue, 26 Apr 2005 07:28:16 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from eagle.oceana.com (eagle.oceana.com [208.17.123.12]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3QESFDT086157 for ; Tue, 26 Apr 2005 07:28:16 -0700 (PDT) (envelope-from ken@oceana.com) Received: from [192.168.10.26] (KEN.oceana.com [192.168.10.26]) by eagle.oceana.com (8.13.2/8.13.2) with ESMTP id j3QES9ZJ008552 for ; Tue, 26 Apr 2005 10:28:09 -0400 (EDT) Message-ID: <426E4FC9.7020903@oceana.com> Date: Tue, 26 Apr 2005 10:27:21 -0400 From: Ken Murchison Organization: Oceana Matrix Ltd. User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.7) Gecko/20040514 X-Accept-Language: en,pdf MIME-Version: 1.0 To: ietf-usefor@imc.org Subject: Re: References References: <4266EF35.21A5@xyzzy.claranet.de> <42686E91.42E2@xyzzy.claranet.de> In-Reply-To: Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit X-Spam-Score: 0 (BAYES_00) X-Scanned-By: MIMEDefang 2.39 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Content-Transfer-Encoding: 7bit Charles Lindsey wrote: > In <42686E91.42E2@xyzzy.claranet.de> Frank Ellermann writes: > >> A "followup" is an article containing a response to the >> contents of an earlier article (the followup's "precursor" >> indicated in the "References" header field). > > > Yes, I had already concluded that something like that would be helpful in > the definition. Russ seems to agree. Done. -- Kenneth Murchison Oceana Matrix Ltd. Software Engineer 21 Princeton Place 716-662-8973 x26 Orchard Park, NY 14127 --PGP Public Key-- http://www.oceana.com/~ken/ksm.pgp From owner-ietf-usefor@mail.imc.org Tue Apr 26 11:03:37 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id LAA15618 for ; Tue, 26 Apr 2005 11:03:37 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3QF1gil091742 for ; Tue, 26 Apr 2005 08:01:42 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3QF1gmq091741 for ietf-usefor-skb; Tue, 26 Apr 2005 08:01:42 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from eagle.oceana.com (eagle.oceana.com [208.17.123.12]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3QF1fu2091697 for ; Tue, 26 Apr 2005 08:01:42 -0700 (PDT) (envelope-from ken@oceana.com) Received: from [192.168.10.26] (KEN.oceana.com [192.168.10.26]) by eagle.oceana.com (8.13.2/8.13.2) with ESMTP id j3QF1YoV009273 for ; Tue, 26 Apr 2005 11:01:34 -0400 (EDT) Message-ID: <426E579F.1060809@oceana.com> Date: Tue, 26 Apr 2005 11:00:47 -0400 From: Ken Murchison Organization: Oceana Matrix Ltd. User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.7) Gecko/20040514 X-Accept-Language: en,pdf MIME-Version: 1.0 To: ietf-usefor@imc.org Subject: Re: References References: In-Reply-To: Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit X-Spam-Score: 0 (BAYES_00) X-Scanned-By: MIMEDefang 2.39 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Content-Transfer-Encoding: 7bit John Stanley wrote: > > "Charles Lindsey" : > > >>Frank Ellermann writes: > > >>>No References => no follow-up. > > >>Which indeed follows from the fact that USEPRO says that a followup agent >>MUST incorporate a References header (and do it correctly, of course). > > > Unfortunately, RFC2822 says only SHOULD regarding the References header > and followups, and USEFOR says that the References header is defined as > found in RFC2822 (and for the pedants who cannot identify the context of > this discussion, I'll add "other than a few syntax changes that don't > change the requirements for usage, which is the context of this discussion > here.) So References in a followup is only a SHOULD, and if USEPRO says > otherwise, it contradicts both RFC2822 and USEFOR. While one of those we > can modify for good reason, the other is the result of our modifications, > and we did NOT modify the References header to make it mandatory in any > use. I think you're taking the fact that neither RFC2822 nor USEFOR mandate References too literally. USEFOR only documents what is necessary to make an article syntactically correct, not what makes it useful in the Netnews world. In this regard, References is optional. A "followup" article which doesn't have a References header is still syntactically correct. This doesn't mean that its recommended or useful. USEPRO documents how to construct an article ("followup" or otherwise) which will play nice in the Netnews world. In this regard, USEPRO can and should mandate that References be present in "followup" articles. USEPRO (and possibly USEAGE) will dictate when the optional headers documented in USEFOR should and should not be used. I don't see a contradiction here, but maybe I'm wrong. -- Kenneth Murchison Oceana Matrix Ltd. Software Engineer 21 Princeton Place 716-662-8973 x26 Orchard Park, NY 14127 --PGP Public Key-- http://www.oceana.com/~ken/ksm.pgp From owner-ietf-usefor@mail.imc.org Tue Apr 26 11:17:21 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id LAA17227 for ; Tue, 26 Apr 2005 11:17:21 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3QFFodk094300 for ; Tue, 26 Apr 2005 08:15:50 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3QFFoE7094299 for ietf-usefor-skb; Tue, 26 Apr 2005 08:15:50 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from eagle.oceana.com (eagle.oceana.com [208.17.123.12]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3QFFnSQ094282 for ; Tue, 26 Apr 2005 08:15:49 -0700 (PDT) (envelope-from ken@oceana.com) Received: from [192.168.10.26] (KEN.oceana.com [192.168.10.26]) by eagle.oceana.com (8.13.2/8.13.2) with ESMTP id j3QFFgDe009529 for ; Tue, 26 Apr 2005 11:15:42 -0400 (EDT) Message-ID: <426E5AEF.3040407@oceana.com> Date: Tue, 26 Apr 2005 11:14:55 -0400 From: Ken Murchison Organization: Oceana Matrix Ltd. User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.7) Gecko/20040514 X-Accept-Language: en,pdf MIME-Version: 1.0 To: ietf-usefor@imc.org Subject: Re: References References: In-Reply-To: Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit X-Spam-Score: 0 (BAYES_00) X-Scanned-By: MIMEDefang 2.39 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Content-Transfer-Encoding: 7bit John Stanley wrote: > > "Charles Lindsey" : > > >>Our Chair has ruled that the correct place to say such things is in USEPRO >>rather than USEFOR. > > > Then the Chair has effectively ruled that References is optional as > defined in RFC2822 and that "followup" is a meaningless term, since nobody > past the poster can identify what is and is not a followup using the > optional definition of References. If the Chair thinks that we can say in > one news standard that References is an optional header (by referring to > the SHOULD appear and "may appear" definition in RFC2822) and then change > our minds in mid-stream by saying that it is MANDATORY in another news > standard, then he is demonstrating schizophrenia. Either it IS mandatory > in followups or it is not. It is mandatory as stated in USEPRO. I'll reiterate what I stated elsewhere in the thread. USEFOR documents the proper syntax of the set of headers typically used in news articles. USEPRO documents how to properly construct an article from this set per the Netnews "protocol" or "etiquette" -- Kenneth Murchison Oceana Matrix Ltd. Software Engineer 21 Princeton Place 716-662-8973 x26 Orchard Park, NY 14127 --PGP Public Key-- http://www.oceana.com/~ken/ksm.pgp From owner-ietf-usefor@mail.imc.org Tue Apr 26 12:17:50 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id MAA22472 for ; Tue, 26 Apr 2005 12:17:50 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3QGFA7t005548 for ; Tue, 26 Apr 2005 09:15:10 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3QGFA7d005547 for ietf-usefor-skb; Tue, 26 Apr 2005 09:15:10 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from lon-mail-4.gradwell.net (lon-mail-4.gradwell.net [193.111.201.130]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3QGF8rS005537 for ; Tue, 26 Apr 2005 09:15:09 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from host81-144-72-114.midband.mdip.bt.net ([81.144.72.114]) by lon-mail-4.gradwell.net with esmtp (Gradwell gwh-smtpd 1.182) id 426e690a.6ead.47 for ietf-usefor@imc.org; Tue, 26 Apr 2005 17:15:06 +0100 (envelope-sender ) Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3QGCGl00858 for ietf-usefor@imc.org; Tue, 26 Apr 2005 17:12:16 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20722 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Suggested References texts Message-ID: X-Newsreader: NN version 6.5.2 (NOV) Date: Tue, 26 Apr 2005 15:12:10 GMT Lines: 116 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: I think we are now sufficiently agreed about the References header (except perhaps for John) to write some text. Here is what I suggest. Note that, in addition to the USEFOR text written by Ken, I have added some preliminary wording to outline the purpose of the References header, and some semantics text to explain the "meaning" of the list of message-ids. I have also fixed the syntax error (the CFWS between the s is obligatory), and have continued the liberty of using rather than . There is also a tentative NOTE at the end regarding the use of the References header in message/partial. This is a place where what RFC 2046 suggests is woefully inadequate, and that needs to be pointed out somewhere. But maybe this is not the right place. Share and Enjoy! Definitions (for inclusion in USEFOR or USEPRO as appropriate). A "followup" is an article containing a response to the contents of an earlier article. It will always include a References header pointing to that earlier article and any other "precursors". An article is a "precursor" of some later article which is a followup to it, or which is otherwise intended to be grouped with it for purposes of display (e.g. as a part of a multipart posting such as a FAQ). A "followup agent" is a combination of reading agent and posting agent that aids in the preparation and posting of a followup. Within USEFOR: 3.2.1 References | The References header is used in followups, and other articles with | related precursors (e.g. multipart FAQs and the fragments of a | message/partial), to facilitate the display or retrieval, by reading | and other agents, of threads of related articles. It is the same as | that specified in Section 3.6.4 of [RFC2822] with the added | restrictions detailed in Section 2.2 and those listed below: o The updated construct defined in Section 3.1.3 MUST be used. o Message identifiers MUST be separated with CFWS. o Comments in CFWS between message identifiers can cause interoperability problems, so comments SHOULD NOT be generated, but MUST be accepted. | references = "References:" SP [CFWS] 1*( msg-id [CFWS] ) | CRLF | The list is composed of message identifiers of precursors of the | current article, sorted so that no article precedes any of its own | precursors. It SHOULD include both the earliest and the immediate | precursors on the current article, even if some of the intermediate | ones are omitted. A given message identifier MUST NOT appear more | than once. [That actually defines a partial list. Even though USEPRO makes no provsision for followups to several articles, people are bound to try to do it. That is a minimal wording which reduces to the usual linear list in the case of followups to a single precursor.] | | The process of generating a References header by a followup agent is | set out in [USEPRO]. | | NOTE: The suggestion in [RFC 2046] that each fragment of a | message/partial should contain a References header referring only | to the immediatelyt preceding fragment does not fulfil the | recommendation above. It would therefore be better for each | fragment to refer to all of the preceding ones. Within USEPRO: 7.6. Duties of a Followup Agent ....... 4. If the precursor did not have a References header (F-3.2.1), the content of the followup's References header MUST be inherited from that of the Message-ID header of the precursor. A followup to an article which already had a References header MUST have a References header comprising the precursor's References header (subject to trimming as described below) followed by CFWS and the Message-ID header content of the precursor. | NOTE: The two "MUST"s above ensure compliance with the | definition of the term "followup". They are to be contrasted | with the weaker recomendation using "SHOULD" applied, in [RFC | 2822], to the generation of "replies" in email. Moreover, in | Netnews, there is no expectation of any In-Reply-To header in a | followup. If the resulting References header is excessively long, it MAY be trimmed, but the first and the last two message identifiers MUST NOT be removed. -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 From owner-ietf-usefor@mail.imc.org Tue Apr 26 12:40:10 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id MAA24814 for ; Tue, 26 Apr 2005 12:40:09 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3QGcB4t010478 for ; Tue, 26 Apr 2005 09:38:11 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3QGcB5N010477 for ietf-usefor-skb; Tue, 26 Apr 2005 09:38:11 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from soy.epix.net (soy.epix.net [199.224.64.64]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3QGcAaX010470 for ; Tue, 26 Apr 2005 09:38:10 -0700 (PDT) (envelope-from mibsoft@epix.net) Received: from [192.168.2.11] (hrbg-216-108-206-228-pppoe.dsl.hrbg.epix.net [216.108.206.228]) by soy.epix.net (8.12.10/2004120601/PL) with ESMTP id j3QGbrgE004420; Tue, 26 Apr 2005 12:38:04 -0400 (EDT) Message-ID: <426E6E66.2070109@epix.net> Date: Tue, 26 Apr 2005 12:37:58 -0400 From: "Forrest J. Cavalier III" User-Agent: Mozilla Thunderbird 0.7 (Windows/20040616) X-Accept-Language: en-us, en MIME-Version: 1.0 To: chl@clerew.man.ac.uk CC: ietf-usefor@imc.org Subject: Re: Suggested References texts References: In-Reply-To: Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit X-Scanned-By: MIMEDefang 2.51 on 199.224.89.153 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Content-Transfer-Encoding: 7bit Charles Lindsey wrote: > I think we are now sufficiently agreed about the References header (except > perhaps for John) to write some text. Point of order. The chair must declare consensus. There are problems with what you wrote, and I think that is precisely because you misunderstand the consensus..... > > Here is what I suggest. Note that, in addition to the USEFOR text written > by Ken, I have added some preliminary wording to outline the purpose of > the References header, and some semantics text to explain the "meaning" > of the list of message-ids. If USEFOR is going to be syntax only, then there is no place for "purpose of the References" header that can be vanilla enough to be in USEFOR. So, the following is wrong: > Within USEFOR: > > 3.2.1 References > > | The References header is used in followups, and other articles with > | related precursors (e.g. multipart FAQs and the fragments of a > | message/partial), to facilitate the display or retrieval, by reading > | and other agents, of threads of related articles. It is the same as > | that specified in Section 3.6.4 of [RFC2822] with the added > | restrictions detailed in Section 2.2 and those listed below: This isn't acceptable wording at all, especially not for USEFOR. A news client must be able to determine "followup" by looking at an article, or the term is meaningless and should not appear in any standards document. From owner-ietf-usefor@mail.imc.org Tue Apr 26 12:50:26 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id MAA25554 for ; Tue, 26 Apr 2005 12:50:25 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3QGnOgP012487 for ; Tue, 26 Apr 2005 09:49:24 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3QGnOJk012486 for ietf-usefor-skb; Tue, 26 Apr 2005 09:49:24 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from a.mail.peak.org (a.mail.peak.org [69.59.192.41]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3QGnN1J012463 for ; Tue, 26 Apr 2005 09:49:23 -0700 (PDT) (envelope-from stanley@peak.org) Received: from a.shell.peak.org ([69.59.192.81]) by a.mail.peak.org (8.12.10/8.12.8) with ESMTP id j3QGnFfO080426 (version=TLSv1/SSLv3 cipher=EDH-RSA-DES-CBC3-SHA bits=168 verify=NO) for ; Tue, 26 Apr 2005 09:49:17 -0700 (PDT) Date: Tue, 26 Apr 2005 09:49:15 -0700 (PDT) From: John Stanley X-X-Sender: stanley@a.shell.peak.org To: ietf-usefor@imc.org Subject: Re: Differences between RFC 2822 and Usefor (fwd) Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Spam-Score: 0 () X-Scanned-By: MIMEDefang 2.39 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: I said: >The fact that I am the only one who even commented on it being gone is a >pretty good sign that nobody cares that it is gone. Ken Murchison responds: >No, I knew that it wasn't there, What are you saying "no" to? You did not comment on it being gone, at least not in any public forum I know of designed for this discussion. Don't tell me I'm wrong when clearly I am not. The fact that you knew it wasn't there and did not comment is even stronger evidence that it not being there is ok by you. If you cared, and knew, you could have commented. In fact, since your name is on the draft as an editor, you could have put it back had you thought it necessary; the fact you had access to the document and did not act to change it is an even stronger indication that it is ok by you. And that it came OUT under your name is proof that it is ok with you. >and there are a lot of other things that >should be in the Usefor draft and that have not made it there yet. This group has been working on this stuff for YEARS. "Not made it there, yet"? If it hasn't "made it there, yet", it's way too late. How many "last calls" has this group been through, already? >I disagree. It is my opinion (and I believe our Chair's) that USEFOR only >document the syntax of an article. The use or non-use of any particular >"optional" header such as References belongs in either USEPRO or USEAGE. If a header is MANDATORY in an article, then it is part of the syntax of that article. >I think you're taking the fact that neither RFC2822 nor USEFOR mandate >References too literally. Excuse me? RFC2822 uses specific terms that say that the References header is NOT mandatory, and even gives an example of how to deal with a reply that does NOT contain a References header, and I'm being "too literal"? This is a technical standard. Either it says what it means or it is broken. >USEFOR only documents what is necessary to make an article syntactically >correct, not what makes it useful in the Netnews world. If a followup REQUIRES a References header, then a References header is REQUIRED to make that article "syntactically correct". Screw "useful", "useful" doesn't get you RFC2119 mandates. >In this regard, References is optional. Yep. According to RFC2822 and draft-usefor-03, References is optional. In any article, even replies, and since replies are followups in our terminology, optional even in followups. >A "followup" article which doesn't have a References header is still >syntactically correct. Not if you think it MUST contain a References header. If you think it MUST contain a References header, a "followup that doesn't have a References header" is broken. Unfortunately, RFC2822 tells us how to process replies (one kind of followup) that don't contain a References header, so trying to pretend there is a mandate is just silly. >USEPRO documents how to construct an article >("followup" or otherwise) which will play nice in the Netnews world. Can we drop this stupid anthropomorphism? Articles don't "play", they are either correct or they are not. Transport agents either transport them or drop them. Reading agents either display them or do not. Articles without References headers exist and they are transported and displayed by the news system every day. They are not mandatory, and our draft now makes that clear. I'm fine with that. I'm not fine with this pretense that we can somehow sneak a mandate back into the standard after explicitely removing it. >In this regard, USEPRO can and should mandate that References be present >in "followup" articles. Show me in RFC2119 where it uses the words "play nice" in regards to MUST/MUST NOT language. In fact, from what you've said about the References header, I see that RFC2119 says that we must not use RFC2119 imperatives, since there is no requirement for interoperability in anything you've said. "Play nice" and "useful" are not reasons for RFC2119 language. >USEPRO (and possibly USEAGE) will dictate when the optional headers >documented in USEFOR should and should not be used. >I don't see a contradiction here, but maybe I'm wrong. Since you conveniently reworded the problem into one of "should and should not" instead of the actual "MUST and MUST NOT", you can easily say you see no contradiction and then pretend that this lack of contradiction applies to the real issue at hand. Nice trick. Didn't work this time. You cannot say "this is optional" and "this is mandatory" at the same time. Pick one, stick with it, and accept the implications from your choice. I've covered the implications, and "mandatory option" is not one of them. From owner-ietf-usefor@mail.imc.org Tue Apr 26 13:44:27 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id NAA29599 for ; Tue, 26 Apr 2005 13:44:26 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3QHgsiG020130 for ; Tue, 26 Apr 2005 10:42:54 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3QHgsu5020129 for ietf-usefor-skb; Tue, 26 Apr 2005 10:42:54 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from eagle.oceana.com (eagle.oceana.com [208.17.123.12]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3QHgsnM020107 for ; Tue, 26 Apr 2005 10:42:54 -0700 (PDT) (envelope-from ken@oceana.com) Received: from [192.168.10.26] (KEN.oceana.com [192.168.10.26]) by eagle.oceana.com (8.13.2/8.13.2) with ESMTP id j3QHgkJW012123 for ; Tue, 26 Apr 2005 13:42:46 -0400 (EDT) Message-ID: <426E7D66.40008@oceana.com> Date: Tue, 26 Apr 2005 13:41:58 -0400 From: Ken Murchison Organization: Oceana Matrix Ltd. User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.7) Gecko/20040514 X-Accept-Language: en,pdf MIME-Version: 1.0 To: ietf-usefor@imc.org Subject: Re: Differences between RFC 2822 and Usefor (fwd) References: In-Reply-To: Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit X-Spam-Score: 0 (BAYES_00) X-Scanned-By: MIMEDefang 2.39 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Content-Transfer-Encoding: 7bit John Stanley wrote: > > I said: > > >>The fact that I am the only one who even commented on it being gone is a >>pretty good sign that nobody cares that it is gone. > > > Ken Murchison responds: > > >>No, I knew that it wasn't there, You have misquoted me here. The above was written by Charles. > > What are you saying "no" to? You did not comment on it being gone, at > least not in any public forum I know of designed for this discussion. > Don't tell me I'm wrong when clearly I am not. The fact that you knew it > wasn't there and did not comment is even stronger evidence that it not > being there is ok by you. If you cared, and knew, you could have > commented. > > In fact, since your name is on the draft as an editor, you could have put > it back had you thought it necessary; the fact you had access to the > document and did not act to change it is an even stronger indication that > it is ok by you. And that it came OUT under your name is proof that it is > ok with you. > > >>and there are a lot of other things that >>should be in the Usefor draft and that have not made it there yet. Again a misquote. If you're going to quote text, please attribute it to the correct party. > Since you conveniently reworded the problem into one of "should and should > not" instead of the actual "MUST and MUST NOT", you can easily say you see > no contradiction and then pretend that this lack of contradiction applies > to the real issue at hand. Nice trick. Didn't work this time. > > You cannot say "this is optional" and "this is mandatory" at the same > time. Pick one, stick with it, and accept the implications from your > choice. I've covered the implications, and "mandatory option" is not one > of them. Thank you for your pleasant and constructive tone. Is your argument that References shouldn't be required for followups (regardless of what USEFOR/USEPRO currently says) or that the current text is incorrect/ambiguous/contradictory/shitty? If the former, please state so. If the latter, please suggest some text for one or more of the documents that you feel is technically sound. -- Kenneth Murchison Oceana Matrix Ltd. Software Engineer 21 Princeton Place 716-662-8973 x26 Orchard Park, NY 14127 --PGP Public Key-- http://www.oceana.com/~ken/ksm.pgp From owner-ietf-usefor@mail.imc.org Tue Apr 26 15:11:34 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id PAA07015 for ; Tue, 26 Apr 2005 15:11:34 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3QJ8Q2M036280 for ; Tue, 26 Apr 2005 12:08:26 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3QJ8Q7V036279 for ietf-usefor-skb; Tue, 26 Apr 2005 12:08:26 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtp2.Stanford.EDU (smtp2.Stanford.EDU [171.67.16.125]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3QJ8PAN036272 for ; Tue, 26 Apr 2005 12:08:25 -0700 (PDT) (envelope-from rra@stanford.edu) Received: from windlord.stanford.edu (windlord.Stanford.EDU [171.64.19.147]) by smtp2.Stanford.EDU (8.12.11/8.12.11) with SMTP id j3QJ8ODB018791 for ; Tue, 26 Apr 2005 12:08:24 -0700 Received: (qmail 23475 invoked by uid 1000); 26 Apr 2005 19:08:24 -0000 To: ietf-usefor@imc.org Subject: Re: Suggested References texts In-Reply-To: <426E6E66.2070109@epix.net> (Forrest J. Cavalier, III's message of "Tue, 26 Apr 2005 12:37:58 -0400") References: <426E6E66.2070109@epix.net> From: Russ Allbery Organization: The Eyrie Date: Tue, 26 Apr 2005 12:08:24 -0700 Message-ID: <87ekcxnyon.fsf@windlord.stanford.edu> User-Agent: Gnus/5.1007 (Gnus v5.10.7) XEmacs/21.4 (Jumbo Shrimp, linux) MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Forrest J Cavalier writes: > This isn't acceptable wording at all, especially not for USEFOR. A news > client must be able to determine "followup" by looking at an article, or > the term is meaningless and should not appear in any standards document. Well, John's right about this. If this is a requirement, then we have to define everything that contains a References header as a followup, whether somewhat artificially or not, since there isn't any other way to detect a followup. I'm not sure where this requirement comes from, though. It seems to me that certain concepts that cannot be detected in software are still useful to talk about in a standard when they affect what software does in response to common user interface commands (like "post followup"). While I can't tell in software if a random article on the wire is a followup or is some other case where References is useful, I *can* tell if a particular news client does the right thing by looking at it, which to me makes it standards territory. -- Russ Allbery (rra@stanford.edu) From owner-ietf-usefor@mail.imc.org Tue Apr 26 15:13:21 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id PAA07312 for ; Tue, 26 Apr 2005 15:13:20 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3QJC8vJ036852 for ; Tue, 26 Apr 2005 12:12:08 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3QJC8vN036851 for ietf-usefor-skb; Tue, 26 Apr 2005 12:12:08 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from a.mail.peak.org (a.mail.peak.org [69.59.192.41]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3QJC3ni036815 for ; Tue, 26 Apr 2005 12:12:07 -0700 (PDT) (envelope-from stanley@peak.org) Received: from a.shell.peak.org ([69.59.192.81]) by a.mail.peak.org (8.12.10/8.12.8) with ESMTP id j3QJBtZ1074485 (version=TLSv1/SSLv3 cipher=EDH-RSA-DES-CBC3-SHA bits=168 verify=NO) for ; Tue, 26 Apr 2005 12:11:58 -0700 (PDT) Date: Tue, 26 Apr 2005 12:11:54 -0700 (PDT) From: John Stanley X-X-Sender: stanley@a.shell.peak.org To: ietf-usefor@imc.org Subject: Suggested References texts Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Spam-Score: 0 () X-Scanned-By: MIMEDefang 2.39 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: "Charles Lindsey" : > A "followup" is an article containing a response to the contents of > an earlier article. It will always include a References header > pointing to that earlier article and any other "precursors". USEFOR and RFC2822 contradict this. Since we've already decided to follow RFC2822 in this matter, your text is wrong. RFC2822 is explicit in providing guidance for processing of replies to replies that have no References header; clearly a reply sans References header is allowed by RFC2822. > 4. If the precursor did not have a References header (F-3.2.1), the > content of the followup's References header MUST be inherited from > that of the Message-ID header of the precursor. A followup to an > article which already had a References header MUST have a > References header comprising the precursor's References header > (subject to trimming as described below) followed by CFWS and the > Message-ID header content of the precursor. RFC2822 contradicts this, and we have already decided that RFC2822 is the definition we are using. Further, these mandates have not been justified wrt RFC2119, which prohibits their use except for good cause. > NOTE: The two "MUST"s above ensure compliance with the > definition of the term "followup". This is absolutely ridiculous. RFC2119 does not allow mandates to be used just to "ensure compliance with a definition". There has to be an interoperatbility issue, and no such issue has been shown. This "MUST" is specious and violates existing standards, and contradicts our own draft regarding article format. > with the weaker recomendation using "SHOULD" applied, in [RFC > 2822], to the generation of "replies" in email. RFC2822 is not limited to "email", and it is dishonest to try to pretend that it is, especially when we defer to RFC2822 for news headers. >I think we are now sufficiently agreed about the References header (except >perhaps for John) to write some text. It is dishonest to ignore people who disagree with you, Charles. From owner-ietf-usefor@mail.imc.org Tue Apr 26 15:30:02 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id PAA09167 for ; Tue, 26 Apr 2005 15:30:01 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3QJT3he039778 for ; Tue, 26 Apr 2005 12:29:03 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3QJT3IV039777 for ietf-usefor-skb; Tue, 26 Apr 2005 12:29:03 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from b.mail.peak.org (b.mail.peak.org [69.59.192.42]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3QJT2se039759 for ; Tue, 26 Apr 2005 12:29:02 -0700 (PDT) (envelope-from stanley@peak.org) Received: from a.shell.peak.org ([69.59.192.81]) by b.mail.peak.org (8.12.10/8.12.8) with ESMTP id j3QJSptO024401 (version=TLSv1/SSLv3 cipher=EDH-RSA-DES-CBC3-SHA bits=168 verify=NO) for ; Tue, 26 Apr 2005 12:28:57 -0700 (PDT) Date: Tue, 26 Apr 2005 12:28:51 -0700 (PDT) From: John Stanley X-X-Sender: stanley@a.shell.peak.org To: ietf-usefor@imc.org Subject: Re: References Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Spam-Score: 0 () X-Scanned-By: MIMEDefang 2.39 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Ken Murchison : > It is mandatory as stated in USEPRO. It cannot be mandatory in USEPRO when it is already defined to be optional in both USEFOR and RFC2822. > I'll reiterate what I stated elsewhere in the thread. > USEFOR documents the proper syntax of the set of headers typically used in > news articles. That is not what you said. You said: > It is my opinion (and I believe our Chair's) that USEFOR only document the > syntax of an article. The "syntax of an article" includes what headers it is required to contain. The "syntax of a header" defines the contents of a header; the syntax of an article defines the contents of an article. USEFOR is USENET Format -- which is clearly not just the headers but the contents of articles that are valid USENET articles. USEFOR does not just tell us the correct syntax of the From header, it tells us that the From header is MANDATORY. It does the same for several other headers. Thus, it is clear that USEFOR is not just "syntax of headers", it includes "syntax of articles". In that job of "syntax of articles", USEFOR tells us that References is optional. References is optional. USEFOR and USEPRO are supposed to complement each other, not contradict. Our Chair has said USEFOR is right not to talk about a mandate, although he's said that because he thinks it can be slipped in after the fact in USEPRO. I agree USEFOR is right not to talk about a mandate -- for a different reason. If you want a mandate, do it honestly and in conformance with RFC2119. Justify it and you can have it. So far, you haven't. "Easier" and "play nice" don't even come close. From owner-ietf-usefor@mail.imc.org Tue Apr 26 16:15:50 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id QAA13597 for ; Tue, 26 Apr 2005 16:15:48 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3QKENtm047613 for ; Tue, 26 Apr 2005 13:14:23 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3QKENRI047612 for ietf-usefor-skb; Tue, 26 Apr 2005 13:14:23 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from rufus.isode.com (rufus.isode.com [62.3.217.251]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3QKELBR047589 for ; Tue, 26 Apr 2005 13:14:22 -0700 (PDT) (envelope-from alexey.melnikov-usefor@isode.com) Received: from [192.168.0.3] ([62.3.217.253]) by rufus.isode.com via TCP (internal) with ESMTPA; Tue, 26 Apr 2005 21:11:51 +0100 Message-ID: <426EA086.6020400@isode.com> Date: Tue, 26 Apr 2005 21:11:50 +0100 From: Alexey Melnikov User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.7.2) Gecko/20040804 Netscape/7.2 (ax) X-Accept-Language: en-us, en MIME-Version: 1.0 To: John Stanley CC: ietf-usefor@imc.org Subject: Re: References References: In-Reply-To: Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Content-Transfer-Encoding: 7bit John Stanley wrote: >Ken Murchison : > > >>It is mandatory as stated in USEPRO. >> >> >It cannot be mandatory in USEPRO when it is already defined to be optional >in both USEFOR and RFC2822. > Surely it can, because USEFOR equally applies to *all* types of USEFOR articles and USEPRO may add additional restrictions for particular types of articles. >>I'll reiterate what I stated elsewhere in the thread. >> >> >>USEFOR documents the proper syntax of the set of headers typically used in >>news articles. >> >> >That is not what you said. You said: > > >>It is my opinion (and I believe our Chair's) that USEFOR only document the >>syntax of an article. >> >> > >The "syntax of an article" includes what headers it is required to >contain. The "syntax of a header" defines the contents of a header; the >syntax of an article defines the contents of an article. USEFOR is USENET >Format -- which is clearly not just the headers but the contents of >articles that are valid USENET articles. > >USEFOR does not just tell us the correct syntax of the From header, it >tells us that the From header is MANDATORY. It does the same for several >other headers. Thus, it is clear that USEFOR is not just "syntax of >headers", it includes "syntax of articles". > >In that job of "syntax of articles", USEFOR tells us that References is >optional. References is optional. > > See my comment above. From owner-ietf-usefor@mail.imc.org Tue Apr 26 16:30:45 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id QAA14572 for ; Tue, 26 Apr 2005 16:30:41 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3QKSHQK049791 for ; Tue, 26 Apr 2005 13:28:17 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3QKSHwR049790 for ietf-usefor-skb; Tue, 26 Apr 2005 13:28:17 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from a.mail.peak.org (a.mail.peak.org [69.59.192.41]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3QKSGwt049763 for ; Tue, 26 Apr 2005 13:28:16 -0700 (PDT) (envelope-from stanley@peak.org) Received: from a.shell.peak.org ([69.59.192.81]) by a.mail.peak.org (8.12.10/8.12.8) with ESMTP id j3QKS8Z1023583 (version=TLSv1/SSLv3 cipher=EDH-RSA-DES-CBC3-SHA bits=168 verify=NO) for ; Tue, 26 Apr 2005 13:28:10 -0700 (PDT) Date: Tue, 26 Apr 2005 13:28:08 -0700 (PDT) From: John Stanley X-X-Sender: stanley@a.shell.peak.org To: ietf-usefor@imc.org Subject: Re: Differences between RFC 2822 and Usefor (fwd) Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Spam-Score: 0 () X-Scanned-By: MIMEDefang 2.39 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Ken Murchison : >You have misquoted me here. The above was written by Charles. You are correct. My mistake, I'm sorry. I'm used to the standard methods of quoting material (using '>') and not this blue-line crap that doesn't cut-and-paste. I'm also used to people eliding things they are NOT responding to and not including large volumes of text just to add one line at the bottom. > Thank you for your pleasant and constructive tone. You are welcome. I'm sorry you are unhappy that you got caught trying to reword the problem to make it go away. "Should/should not" is not the issue and pretending it is is pretty insulting to the person you tried pulling that on. Excuse me for pointing out that I caught you. >Is your argument that References shouldn't be required for followups >(regardless of what USEFOR/USEPRO currently says) or that the current text >is incorrect/ambiguous/contradictory/shitty? My argument is exactly as I stated: you cannot say "this is optional" and "this is mandatory" at the same time. This is a problem; at least, I think most people view contradictions in technical standards as a problem. Perhaps you do not. The solution is as I stated: pick one and stick with it, including all the implications that go with it. >If the former, please state so. If the latter, please suggest some text >for one or more of the documents that you feel is technically sound. Before we can write text we must make the choice. To do otherwise is a complete waste of time. Writing text and THEN justifying it is backwards, and it is why we are spending so much time now. The mandate has been removed from the draft and people are busy trying to pretend it is still there or should simply be reinstated without justification. And yet, there must have been a reason it was removed. We have two (sane) options: 1. RFC1036 was correct. Being able to identify "followups" is so important that the header that allows this merits RFC2119 mandates. The implication from this is that References MUST appear in followups and MUST NOT appear elsewhere, since any other combination of RFC2119 language would not accomplish the goal of allowing identification of followups. A further implication of this choice is that we MUST document the difference from RFC2822 where the header and article format are defined. That's in USEFOR. Another implication is that we have to define followups. That's not really that hard; I've provided a proposed definition that meets current practice. 2. RFC1036 was incorrect. Being able to identify "followups" is not important enough to merit RFC2119 language and the header that allows this is thus optional. Since the only method of identifying a followup has been made optional, the concept of followup is thus not worth defining since nobody can detect them when they appear anyway. At least, not detect them with any certainty. (Do you have a mole on your left shoulder? Your messages don't contain any way of telling; fortunately, I don't care, although someone else might.) Once we have chosen the path we wish to take, we can look at how to accomplish it. Unfortunately for position 1, we've already removed the mandate. USEFOR defers to RFC2822, RFC2822 says "SHOULD" and "may contain". Putting a mandate BACK after deciding to remove it should, in a correctly operating system, require justification. We were getting by with "historical reasons" when it was a mandate copied from RFC1036; if that wasn't sufficient to keep it in place, then it won't be enough to put it back now that it is gone. I've seen no interoperability issues; do you have something to report? So, looking at RFC2119, it appears the only remaining option is 2. USEFOR already makes the header optional. The best that USEPRO can do is say "IF you include the header, it must contain X". Most of the text Charles just proposed meets that criterion. Remove the parts that refer to the specious mandate. How do we deal with "duties of a followup agent"? Well, RFC2822 talks about "replies". That's a pretty common english word, why is it unavailable to us? USEPRO can simply talk about what headers are available for a reply and what the headers that DO get used in the reply contain. E.g., "If a References header is included, it MUST contain ...". That's fine. "You MUST include" this optional header is not. And when you talk about headers like Newsgroups and Path, you can say "This MUST be present and MUST contain..." because USEFOR already says they are mandatory. Now, I've said it all again. Anyone up for a third pass? And, if you are keeping score on such things, your demand that I repeat myself or provide text, with the implications of that demand, is pretty insulting, too. Other people can read it the first time I send it; you ought to be able to, too. From owner-ietf-usefor@mail.imc.org Wed Apr 27 03:15:00 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id DAA08707 for ; Wed, 27 Apr 2005 03:14:59 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3R7DsBj095446 for ; Wed, 27 Apr 2005 00:13:54 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3R7DsdU095445 for ietf-usefor-skb; Wed, 27 Apr 2005 00:13:54 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from ciao.gmane.org (main.gmane.org [80.91.229.2]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3R7DpKk095422 for ; Wed, 27 Apr 2005 00:13:52 -0700 (PDT) (envelope-from usenet-format@gmane.org) Received: from list by ciao.gmane.org with local (Exim 4.43) id 1DQgdd-0004JE-D1 for ietf-usefor@imc.org; Wed, 27 Apr 2005 09:07:13 +0200 Received: from 212.82.251.220 ([212.82.251.220]) by main.gmane.org with esmtp (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Wed, 27 Apr 2005 09:07:13 +0200 Received: from nobody by 212.82.251.220 with local (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Wed, 27 Apr 2005 09:07:13 +0200 X-Injected-Via-Gmane: http://gmane.org/ To: ietf-usefor@imc.org From: Frank Ellermann Subject: Re: References Date: Wed, 27 Apr 2005 09:11:16 +0200 Organization: Lines: 20 Message-ID: <426F3B14.54C1@xyzzy.claranet.de> References: <426E579F.1060809@oceana.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Complaints-To: usenet@sea.gmane.org X-Gmane-NNTP-Posting-Host: 212.82.251.220 X-Mailer: Mozilla 3.0 (OS/2; U) Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Content-Transfer-Encoding: 7bit Ken Murchison wrote: > I don't see a contradiction here There's no contradiction. We have followup => references, and after you've modified the definition as indicated in... <426E4FC9.7020903@oceana.com> | Done. ...this issue is settled for Usefor-04 and the future RfC, as far as I'm concerned. One WG member apparently wants an equivalences in the form of followup <=> references and followup <=> 2822-reply, but that's obviously wrong, and the proposed alternative of no relation at all between followup and references is also wrong. Bye, Frank From owner-ietf-usefor@mail.imc.org Wed Apr 27 03:57:49 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id DAA11434 for ; Wed, 27 Apr 2005 03:57:48 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3R7uikl015581 for ; Wed, 27 Apr 2005 00:56:44 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3R7uiog015580 for ietf-usefor-skb; Wed, 27 Apr 2005 00:56:44 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from ciao.gmane.org (main.gmane.org [80.91.229.2]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3R7ugGJ015568 for ; Wed, 27 Apr 2005 00:56:43 -0700 (PDT) (envelope-from usenet-format@gmane.org) Received: from list by ciao.gmane.org with local (Exim 4.43) id 1DQhJT-0000Y4-RO for ietf-usefor@imc.org; Wed, 27 Apr 2005 09:50:27 +0200 Received: from 212.82.251.220 ([212.82.251.220]) by main.gmane.org with esmtp (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Wed, 27 Apr 2005 09:50:27 +0200 Received: from nobody by 212.82.251.220 with local (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Wed, 27 Apr 2005 09:50:27 +0200 X-Injected-Via-Gmane: http://gmane.org/ To: ietf-usefor@imc.org From: Frank Ellermann Subject: Re: Suggested References texts Date: Wed, 27 Apr 2005 09:50:20 +0200 Organization: Lines: 65 Message-ID: <426F443C.6561@xyzzy.claranet.de> References: Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Complaints-To: usenet@sea.gmane.org X-Gmane-NNTP-Posting-Host: 212.82.251.220 X-Mailer: Mozilla 3.0 (OS/2; U) Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Content-Transfer-Encoding: 7bit Charles Lindsey wrote: > I have also fixed the syntax error (the CFWS between the > s is obligatory), and have continued the liberty of > using rather than . Good (but see below). > what RFC 2046 suggests is woefully inadequate > A "followup" is an article containing a response to the > contents of an earlier article. It will always include a > References header pointing to that earlier article and any > other "precursors". Strong NAK to any change of <426E4FC9.7020903@oceana.com> We know that References do _not_ always include 'any other "precursors"', in fact there's a complete chapter explaining how to trim the References. > An article is a "precursor" of some later article which is a > followup to it Strong NAK to any change of <426E4FC9.7020903@oceana.com> > or which is otherwise intended to be grouped with it for > purposes of display (e.g. as a part of a multipart posting > such as a FAQ). Strong NAK, the different parts of a multipart FAQ might be related by References, but this relation is not necessarily that of a "precursor". A "precursor" is simply the opposite of a "followup", so if A is the precursor of B, then B is a followup of A. There's no case where A is the precursor of B, but B is no followup of A, that would be madness. > Within USEFOR: Strong NAK, after <426E4FC9.7020903@oceana.com> this issue is settled in Usefor, we only have to fix the syntax. >| references = "References:" SP [CFWS] 1*(msg-id [CFWS]) CRLF Strong NAK to anything below this line. And the syntax is not correct, we wanted at least one WSP to separate the msg-id. references = "References:" SP msg-id-list CRLF msg-id-list = [CFWS] msg-id *( CFWS msg-id ) [CFWS] That's the simple variant where a comment is enough to separate the Message-IDs as in (c) but adjacent Message-IDs as in aren't allowed. [Usepro] > If the resulting References header is excessively long, it > MAY be trimmed, but the first and the last two message > identifiers MUST NOT be removed. I'd go for a SHOULD here, because some news servers and tools really hate it if the References are too long, I've seen cases where the References were truncated Bye, Frank From owner-ietf-usefor@mail.imc.org Wed Apr 27 04:41:49 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id EAA14692 for ; Wed, 27 Apr 2005 04:41:48 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3R8eZTh032765 for ; Wed, 27 Apr 2005 01:40:35 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3R8eZLJ032764 for ietf-usefor-skb; Wed, 27 Apr 2005 01:40:35 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from ciao.gmane.org (main.gmane.org [80.91.229.2]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3R8eXoZ032744 for ; Wed, 27 Apr 2005 01:40:33 -0700 (PDT) (envelope-from usenet-format@gmane.org) Received: from list by ciao.gmane.org with local (Exim 4.43) id 1DQi0K-0005dk-HY for ietf-usefor@imc.org; Wed, 27 Apr 2005 10:34:44 +0200 Received: from 212.82.251.220 ([212.82.251.220]) by main.gmane.org with esmtp (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Wed, 27 Apr 2005 10:34:44 +0200 Received: from nobody by 212.82.251.220 with local (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Wed, 27 Apr 2005 10:34:44 +0200 X-Injected-Via-Gmane: http://gmane.org/ To: ietf-usefor@imc.org From: Frank Ellermann Subject: Re: Obsolete vs. obsolescent Date: Wed, 27 Apr 2005 10:38:20 +0200 Organization: Lines: 46 Message-ID: <426F4F7C.7385@xyzzy.claranet.de> References: <200504111212.j3BCChM14516@clerew.man.ac.uk> <87psx1gnjc.fsf@windlord.stanford.edu> <87fyxvx1m6.fsf@windlord.stanford.edu> <87mzs2a1hu.fsf@windlord.stanford.edu> <87wtqxui6s.fsf@windlord.stanford.edu> <87d5so2bpk.fsf@windlord.stanford.edu> <42697BB6.6074@xyzzy.claranet.de> <426BC455.2040108@isode.com> <426E4B45.7010205@oceana.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Complaints-To: usenet@sea.gmane.org X-Gmane-NNTP-Posting-Host: 212.82.251.220 X-Mailer: Mozilla 3.0 (OS/2; U) Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Content-Transfer-Encoding: 7bit Ken Murchison wrote: > Care to craft some text and tell me where you'd like to see > it? IMHO all that should be done is to update this statement: | As a result, this header is to be regarded as obsolete, and | it will likely be removed entirely in a future version of | this standard. As a result, this header is to be regarded as obsolescent. Servers and UAs MUST accept but SHOULD ignore it, and SHOULD NOT generate it. For the obsolete headers in 3.3 you already have a definition: | Articles containing these headers MUST NOT be generated. | Persons writing new agents SHOULD ignore any former meanings | these headers. That should be "of these headers", but the sentence makes no sense, the handling depends on the header field. Probably new agents MUST ignore the former meaning of the obsolete header fields. But for Also-Control I'd opt for the same procedure as for a Subject: cmsg without a corresponding Control: Oops, "new" problem: Subject: cmsg is not mentioned anywhere in usefor-03, that's a serious bug. It's mentioned in usepro-03, but it only says that this is now obsolete. That's of course FUBAR, it's a severe security violation. We cannot jump from RfC 1036 to a completely different procedure at date X. AFAIK RfCs must not have a flag day. Even if it's only in theory (which it's not, Subject: cmsg has serious difficulties to propagate, the last time I checked it Google ignored a Subject: Re: cmsg, and that's not the only oddity with a Subject: cmsg) Maybe publishing Henry's s-o-1036 first could formally save this situation, otherwise Subject: cmsg is IMO a showstopper. Bye, Frank From owner-ietf-usefor@mail.imc.org Wed Apr 27 06:44:21 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id GAA24088 for ; Wed, 27 Apr 2005 06:44:20 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RAh80x085887 for ; Wed, 27 Apr 2005 03:43:08 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3RAh8op085886 for ietf-usefor-skb; Wed, 27 Apr 2005 03:43:08 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from ciao.gmane.org (main.gmane.org [80.91.229.2]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RAh50T085859 for ; Wed, 27 Apr 2005 03:43:06 -0700 (PDT) (envelope-from usenet-format@gmane.org) Received: from list by ciao.gmane.org with local (Exim 4.43) id 1DQju9-0002J9-Ve for ietf-usefor@imc.org; Wed, 27 Apr 2005 12:36:29 +0200 Received: from 212.82.251.220 ([212.82.251.220]) by main.gmane.org with esmtp (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Wed, 27 Apr 2005 12:36:29 +0200 Received: from nobody by 212.82.251.220 with local (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Wed, 27 Apr 2005 12:36:29 +0200 X-Injected-Via-Gmane: http://gmane.org/ To: ietf-usefor@imc.org From: Frank Ellermann Subject: Re: Fixed Date: Wed, 27 Apr 2005 12:34:48 +0200 Organization: Lines: 44 Message-ID: <426F6AC8.5045@xyzzy.claranet.de> References: <4255153E.3080500@isode.com> <42594879.3B92@xyzzy.claranet.de> <425A8A63.268A@xyzzy.claranet.de> <425C607E.1B5A@xyzzy.claranet.de> <425D8031.D57@xyzzy.claranet.de> <426BB63D.3060506@isode.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Complaints-To: usenet@sea.gmane.org X-Gmane-NNTP-Posting-Host: 212.82.251.220 X-Mailer: Mozilla 3.0 (OS/2; U) Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Content-Transfer-Encoding: 7bit Alexey Melnikov wrote: >> the general principle that we do not change anything in RFC >> 2822 that we do not need to change. > I concur with your preference. I don't. The RfC 2822 concept of Message-IDs is "wrong", and the RfC 1036 concept is better. Of course at least I believe that the RfC 2822 concept is in fact still the same. But as far as I can tell it at least Charles thinks that it's not. I'm not sure what he believes what an "id-right" is. > I suggest they talk to Pete Resnick first. As far as I > remember he has already agreed to publish 2822bis. I don't recall any discussion about an 2822bis in the past 12 months on the 822-list, but I could post the two proposals for the _equivalent_ msg-id syntax there (and the reason for this controversy, "id-right" vs. "id-domain"). I've added an "id-fication" of my proposal below, bye, Frank ~~~ begin ~~~ msg-id = "<" id-local "@" id-domain ">" id-local = dot-atom-text / ( DQUOTE id-quote DQUOTE ) id-quote = ( "." [id-text] ) / ( [id-text] "." ) / ( [id-text] id-special [id-text] ) id-text = 1*( atext / "." / id-special ) id-special = "(" / ")" / "," / ; all specials, minus ">", "[" / "]" / "@" / ; minus DQUOTE, minus "\", ":" / ";" / "<" / ; minus single ".", plus: ".." / "\\" / ( "\" DQUOTE ) id-domain = dot-atom-text / ("[" id-literal "]") id-literal = 1*( %d33-61 / ; printable ASCII minus %d63-90 / ; ">", "[", "\", "]" %d94-126 / ; plus "\[", "\\, "\]" "\[" / "\\" / "\]" ) ~~~ end ~~~ From owner-ietf-usefor@mail.imc.org Wed Apr 27 06:45:05 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id GAA24231 for ; Wed, 27 Apr 2005 06:45:05 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RAi8kx086346 for ; Wed, 27 Apr 2005 03:44:09 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3RAi8Ra086344 for ietf-usefor-skb; Wed, 27 Apr 2005 03:44:08 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from anchor-post-35.mail.demon.net (anchor-post-35.mail.demon.net [194.217.242.85]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RAi78D086331 for ; Wed, 27 Apr 2005 03:44:08 -0700 (PDT) (envelope-from richard@highwayman.com) Received: from gti.noc.demon.net ([195.11.55.101] helo=happyday.al.cl.cam.ac.uk) by anchor-post-35.mail.demon.net with esmtp (Exim 4.42) id 1DQjyd-000EcH-Hq for ietf-usefor@imc.org; Wed, 27 Apr 2005 10:41:07 +0000 Message-ID: Date: Wed, 27 Apr 2005 11:43:33 +0100 To: ietf-usefor@imc.org From: Richard Clayton Subject: Re: Differences between RFC 2822 and Usefor (fwd) References: In-Reply-To: MIME-Version: 1.0 X-Mailer: Turnpike Integrated Version 5.02 M Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 In message , John Stanley writes >We have two (sane) options: I come at this as a newsreading client designer, author and implementer >1. RFC1036 was correct. Being able to identify "followups" is so important >that the header that allows this merits RFC2119 mandates. the readers of Usenet believe in things like threads and followups; they like the ability to split a series of rambling conversations into disparate piles of articles and understand each of those piles one at a time... of course we normally call the piles "threads" but a thread is always an artificial construct because it cannot be reliably constructed without a complete set of articles to hand and Usenet never guarantees that you have this -- so "piles" is rather closer to reality :) there are flaws with a divide and conquer approach to newsgroup reading, not least if you check a newsgroup for new material faster than new articles flow in -- but in general it has been found to work; and we can see the concept being copied into other media such as web-based forums. viz: it may be flawed, but it's the best UI model we currently have However, there seems to be an argument being presented that when you look deeply at the specifications the notion of followups and threads is mainly an illusion fostered by the reading tools and one can lawfully miss out all the headers that allow the illusion to be constructed. But even if the illusion is built on sand (and my experience tells me how many heuristics you need to stabilise the sand), it still has considerable practical value >2. RFC1036 was incorrect. Being able to identify "followups" is not >important enough to merit RFC2119 language and the header that allows this >is thus optional. if there is no standardised way of making the piles of articles in a consistent manner then as time goes on and posting software diverges from traditional working (and there is no document to guide them otherwise) those heuristics I mentioned above will have less and less to work with. ie: the piles will be different in different software. This will lead to considerable confusion amongst the readers of the newsgroup. Some will believe that a particular article belongs in one pile, some in another. Some will miss the article altogether because they have used "kill" features of their software to exclude some of the piles as uninteresting. They will express this confusion when they write further articles. Most know not to say "the thread above this one" to express article relationships -- but they often say "earlier in this thread" to say where to find other material... So it seems to me that this is a genuine inter-operability issue -- not at the protocol levels : the articles get to the far end just fine and the servers will serve them perfectly well. The issue occurs at the level at which the readers view the newsgroups. For that reason I think that a standardised way of expressing which pile an article belongs in is of very significant value. Yes at present some software is a bit flaky, but that doesn't mean we should fail to say what should be done. >Unfortunately for position 1, we've already removed the mandate. sounds like a cock-up :( so best fix it then! ASAP!! >Other people can read it the first time I send it; you >ought to be able to, too. ah.... perhaps I could find it earlier in the "thread" ? :) - -- richard @ highwayman . com "Nothing seems the same Still you never see the change from day to day And no-one notices the customs slip away" -----BEGIN PGP SIGNATURE----- Version: PGPsdk version 1.7.1 iQA/AwUBQm9s1ZoAxkTY1oPiEQKVqwCeNf1Jz07FuhsHwNLToDuA/9n1+K0AniMv tuxgFCozhLxFHv8oL/P/Lmz3 =2oPP -----END PGP SIGNATURE----- From owner-ietf-usefor@mail.imc.org Wed Apr 27 07:16:22 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id HAA26358 for ; Wed, 27 Apr 2005 07:16:22 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RBEUog098132 for ; Wed, 27 Apr 2005 04:14:30 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3RBEUKR098131 for ietf-usefor-skb; Wed, 27 Apr 2005 04:14:30 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from lentil.epix.net (lentil.epix.net [199.224.64.67]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RBET1h098097 for ; Wed, 27 Apr 2005 04:14:29 -0700 (PDT) (envelope-from mibsoft@epix.net) Received: from [192.168.2.11] (hrbg-199-224-121-129-pppoe.dsl.hrbg.epix.net [199.224.121.129]) by lentil.epix.net (8.12.10/2004120601/PL) with ESMTP id j3RBEPl8015331 for ; Wed, 27 Apr 2005 07:14:26 -0400 (EDT) Message-ID: <426F7411.2010903@epix.net> Date: Wed, 27 Apr 2005 07:14:25 -0400 From: "Forrest J. Cavalier III" User-Agent: Mozilla Thunderbird 0.7 (Windows/20040616) X-Accept-Language: en-us, en MIME-Version: 1.0 To: ietf-usefor@imc.org Subject: Re: References References: <426E579F.1060809@oceana.com> <426F3B14.54C1@xyzzy.claranet.de> In-Reply-To: <426F3B14.54C1@xyzzy.claranet.de> Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit X-Scanned-By: MIMEDefang 2.51 on 199.224.89.153 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Content-Transfer-Encoding: 7bit Frank Ellermann wrote: > One WG member apparently wants an equivalences in the form > of followup <=> references and followup <=> 2822-reply, but > that's obviously wrong, and the proposed alternative of no > relation at all between followup and references is also wrong. Did anyone ever advocate that 2822-reply => followup? It isn't one WG member that wants this: followup <=> references I think it is a requirement that a news client be able to determine "followup" by inspecting the article headers. If it is not a requirement, then the term "followup" is meaningless and should not appear in any document. In support of this, I point that the References header is for the use of the READER, not the POSTER. The poster already knows how all the articles are related. Also, if not mandating anything required by the READER for interoperability, how can a document merit anything stronger than a "MAY add a References header"? (And if it is this weak, what's the point?) Frank, you allowed that followup => references. If you don't hold that references => followup, then does the standard allow an article to reference more than one thread? I think that allowing References to appear in an article which is not a reply to a single article (which I thought was the term "followup"), leads to undecidable actions about threading. What is the READER agent supposed to do in that situation? Is there a term for articles that can have references but are not followups? If not, why the special term "followup"? What does the apperance of References header mean, from a READER perspective? Sure it is POSSIBLE to define a standard which describes unambiguous construction of an article, but not unambiguous interpretation of that article. But I think that's ludicrous. Is there any other IETF standard that does this? From owner-ietf-usefor@mail.imc.org Wed Apr 27 07:19:40 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id HAA26613 for ; Wed, 27 Apr 2005 07:19:40 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RBIQmj000179 for ; Wed, 27 Apr 2005 04:18:26 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3RBIQNx000178 for ietf-usefor-skb; Wed, 27 Apr 2005 04:18:26 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from gundel.de.clara.net (gundel.de.clara.net [212.82.225.86]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RBIOh4000156 for ; Wed, 27 Apr 2005 04:18:25 -0700 (PDT) (envelope-from nobody@xyzzy.claranet.de) Received: from [212.82.251.220] (helo=xyzzy) by gundel.de.clara.net with smtp (Exim 4.30; FreeBSD) id 1DQklm-000PmB-Rf for ietf-usefor@imc.org; Wed, 27 Apr 2005 13:31:55 +0200 To: ietf-usefor@imc.org Message-ID: <426F6107.666C@xyzzy.claranet.de> Date: Wed, 27 Apr 2005 11:53:11 +0200 From: Frank Ellermann Organization: X-Mailer: Mozilla 3.0 (OS/2; U) MIME-Version: 1.0 Subject: Again broken (was: Fixed) References: <4255153E.3080500@isode.com> <42594879.3B92@xyzzy.claranet.de> <425A8A63.268A@xyzzy.claranet.de> <425C607E.1B5A@xyzzy.claranet.de> <425D8031.D57@xyzzy.claranet.de> <426BB63D.3060506@isode.com> Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Charles Lindsey wrote: ~~~ begin quoted ABNF ~~~ message-id = "Message-ID:" SP [FWS] msg-id [FWS] CRLF msg-id = "<" id-left "@" id-right ">" ; maximum length is 250 octets id-left = dot-atom-text / no-fold-quote id-right = dot-atom-text / no-fold-literal no-fold-quote = DQUOTE ( "." *mqtext / *mqtext "." / *mqtext mqspecial *mqtext ) DQUOTE mqtext = atext / "." / mqspecial mqspecial = "(" / ")" / ; same as specials except "<" / ; "\" and DQUOTE quoted "[" / "]" / ; "." doubled and ">" omitted ":" / ";" / "@" / "," / ".." / "\\" / "\" DQUOTE no-fold-literal = "[" *( mdtext / "\[" / "\]" / "\\" ) "]" mdtext = %d33-61 / ; The rest of the US-ASCII %d63-90 / ; characters not including %d94-126 ; ">", "[", "]", or "\" ~~~ end of quoted sytax ~~~ Is that as you want it ? I'm not sure about Alexey's external expert for 2822bis, so I think the most simple solution to get an expert review for the two proposals is to post them on the rfc-822 mailing list. Further proposed text: > The msg-id-core MUST NOT be more than 250 octets in length. s/msg-id-core/msg-id/g >| Also note that this updated ABNF applies wherever is used, >| including the References header discussed in Section 3.2.1 and the >| Supersedes header discussed in Section 3.2.5. IMHO that's obvious and needs no "note", excl. References: the two other cases (Supersedes: and Control: cancel) don't exist in mail. BTW, in Control: you have the same syntax problem as in the References:, you separate the arguments by an _optional_ [FWS]: Control: cancel Control: newgroupthis.is.new Control: rmgroupthis.is.old You need a mandatory FWS in 3.2.4. "value" is apparently not defined in usefor-03 or 2822. "token" is apparently also not defined in usefor-03 or 2822. Did anybody try an ABNF checker on usefor-03 ? (I didn't) >| NOTE: It in RECOMMENDED in [RFC 2822] that, for ensuring >| global uniqueness, the be some domain identifier It's unnecessary to quote that RECOMMENDATION, least of all the "some domain identifier". Out of its original very convoluted RfC 2822 context this quote is in fact _worse_ than RfC 2822. RfC 1036 or s-o-1036 is _much_ better than this sad parody of a Message-ID. Usenet is better off without this idea, it is harmful, it should never become an RfC. IIRC we had a "pseudo-consensus" (= you and me) about domain- literals in addresses, Message-IDs, and Path-identifiers, i.e. say something like STD 11 without shouting: | Note: THE USE OF DOMAIN-LITERALS IS STRONGLY DISCOURAGED. Apparently that didn't make it yet into usefor-03 or usepro-03, it's roughly equivalent to a SHOULD NOT mentioning the "domain" in all "addr-spec" found in any From: / Sender: / Reply-To: / Approved:, the "id-right" (your 2822-terminology) of a msg-id, the "path-identity", do we need it anywhere else ? Bye, Frank From owner-ietf-usefor@mail.imc.org Wed Apr 27 12:16:24 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id MAA29064 for ; Wed, 27 Apr 2005 12:16:24 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RGFJqY051793 for ; Wed, 27 Apr 2005 09:15:19 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3RGFJv7051792 for ietf-usefor-skb; Wed, 27 Apr 2005 09:15:19 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from lon-mail-3.gradwell.net (lon-mail-3.gradwell.net [193.111.201.127]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RGFIJn051775 for ; Wed, 27 Apr 2005 09:15:19 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from host81-144-72-27.midband.mdip.bt.net ([81.144.72.27]) by lon-mail-3.gradwell.net with esmtp (Gradwell gwh-smtpd 1.181) id 426fba95.12d8a.217 for ietf-usefor@imc.org; Wed, 27 Apr 2005 17:15:17 +0100 (envelope-sender ) Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3RGCGK09003 for ietf-usefor@imc.org; Wed, 27 Apr 2005 17:12:16 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20739 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: Suggested References texts Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: <426E6E66.2070109@epix.net> Date: Wed, 27 Apr 2005 12:04:07 GMT Lines: 125 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In <426E6E66.2070109@epix.net> "Forrest J. Cavalier III" writes: >Charles Lindsey wrote: >> I think we are now sufficiently agreed about the References header (except >> perhaps for John) to write some text. >Point of order. The chair must declare consensus. Since when has the Chair's permission been required before a member of this WG can suggest some text? >If USEFOR is going to be syntax only, then there is no place for "purpose of >the References" header that can be vanilla enough to be in USEFOR. There are two issues here. A. Should USEFOR including brief explanatory wording for each header to say what its purpose is? Remember that USEFOR is semantics as well as syntax. B. If so, then what should be said when introducing the References header? Dealing with A first, RFC 2822 regularly provides such explanatory wording. The "Subject:" field ... contains a short string identifying the topic of the message. These three fields are intended to have only human-readable content ... The originator fields indicate the mailbox(es) of the source of the message. The "From:" field specifies the author(s) of the message, ... And for the References field it says: The "In-Reply-To:" and "References:" fields are used when creating a reply to a message. Note: Some implementations parse the "References:" field to display the "thread of the discussion". ... And USEFOR also contains similar descriptive material (though often it simply inherits what RFC 2822 had already said): The Path header indicates the route taken by an article since its injection into the Netnews system. The Injection-Date header contains ... Its purpose is to prevent the reinjection into the news stream of "stale" articles which have already expired by the time they arrive at some relaying or serving agent. So there is plenty of precedent for incorporating such material. So now to issue B: >> | The References header is used in followups, and other articles with >> | related precursors (e.g. multipart FAQs and the fragments of a >> | message/partial), to facilitate the display or retrieval, by reading >> | and other agents, of threads of related articles. ... >This isn't acceptable wording at all, especially not for USEFOR. A >news client must be able to determine "followup" by looking at an article, >or the term is meaningless and should not appear in any standards document. Let us recap the history here. It is accepted, I believe, (using Frank Ellermann's notation) followup -> references header no references header -> not a followup and at one time we had references header -> followup But then, some months back, it was pointed out that some writers of multipart FAQs regularly used References headers to join the later parts to the earlier ones, and RFC 2046 suggests the use of References headers in message/partial. And there were other examples too. Did we want to forbid these practices, and would they actually disappear even if we did? Far from forbidding them, the view expressed (without dissent, so far as I recall) was that these practices were useful, and should perhaps even be encouraged. Essentially, anyone incorporating a References header for such reasons was in effect inviting reading agents (those that did threading anyway) to present the various parts of the FAQ together and in the correct order (irrespective of their order of arrival, differences in Subject, or order of their Date headers). Inevitably, this changes the information that a reading agent can deduce when it sees a References header. It can no longer deduce "this is a followup". It can only deduce "the poster wants you to treat this like a followup" (by threading it, or by letting you retrieve the precursors by clicking on them, or whatever other bells and whistles you provide). So you can no longer assert "references header -> followup". The question is 'Does this matter?'. I don't see why it should. Why is is necessary for a reading agent to be able to distinguish between (conventional) followups and multipart FAQs (note that humans can still distinguish them easily)? In any case, if this practice continues (as it surely will), there is nothing that can be done about it, since the References header is the only tool available. Now there have been suggestions to "clean it up" by artificially redefining the term "followup" to include any article that has a References header. But that is all "smoke and mirrors" - it makes it appear that you can detect "followups", but in reality it makes no difference at all as to what reading agents can or cannot actually do. Yes, if the WG wants that revised definition, it can be done, but I do not see the point since it would make no technical difference to what actually happens. -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 From owner-ietf-usefor@mail.imc.org Wed Apr 27 13:00:56 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id NAA03054 for ; Wed, 27 Apr 2005 13:00:55 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RGxMmm055889 for ; Wed, 27 Apr 2005 09:59:22 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3RGxMAj055888 for ietf-usefor-skb; Wed, 27 Apr 2005 09:59:22 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from a.mail.peak.org (a.mail.peak.org [69.59.192.41]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RGxL0x055874 for ; Wed, 27 Apr 2005 09:59:21 -0700 (PDT) (envelope-from stanley@peak.org) Received: from a.shell.peak.org ([69.59.192.81]) by a.mail.peak.org (8.12.10/8.12.8) with ESMTP id j3RGxFZ1044836 (version=TLSv1/SSLv3 cipher=EDH-RSA-DES-CBC3-SHA bits=168 verify=NO) for ; Wed, 27 Apr 2005 09:59:16 -0700 (PDT) Date: Wed, 27 Apr 2005 09:59:15 -0700 (PDT) From: John Stanley X-X-Sender: stanley@a.shell.peak.org To: ietf-usefor@imc.org Subject: Re: Suggested References texts Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Spam-Score: 0 () X-Scanned-By: MIMEDefang 2.39 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Russ Allbery : >If this is a requirement, then we have to >define everything that contains a References header as a followup, You have it backwards. We define a followup by the poster's intent; we DETECT followups with the References header. There is a difference. Your definition leaves us with a circle: a followup contains a References header, and a References header must be used in a followup. Ok, what's a followup? From owner-ietf-usefor@mail.imc.org Wed Apr 27 13:02:37 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id MAA29065 for ; Wed, 27 Apr 2005 12:16:24 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RGFJTA051782 for ; Wed, 27 Apr 2005 09:15:19 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3RGFJYd051781 for ietf-usefor-skb; Wed, 27 Apr 2005 09:15:19 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from lon-mail-3.gradwell.net (lon-mail-3.gradwell.net [193.111.201.127]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RGFHeX051773 for ; Wed, 27 Apr 2005 09:15:18 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from host81-144-72-27.midband.mdip.bt.net ([81.144.72.27]) by lon-mail-3.gradwell.net with esmtp (Gradwell gwh-smtpd 1.181) id 426fba94.12d8a.216 for ietf-usefor@imc.org; Wed, 27 Apr 2005 17:15:16 +0100 (envelope-sender ) Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3RGCHU09009 for ietf-usefor@imc.org; Wed, 27 Apr 2005 17:12:17 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20740 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: References Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: <4266EF35.21A5@xyzzy.claranet.de> <87mzrtt1fh.fsf@windlord.stanford.edu> <426E4F27.2020607@oceana.com> Date: Wed, 27 Apr 2005 13:04:25 GMT Lines: 31 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In <426E4F27.2020607@oceana.com> Ken Murchison writes: >I disagree. It is my opinion (and I believe our Chair's) that USEFOR >only document the syntax of an article. The use or non-use of any >particular "optional" header such as References belongs in either USEPRO > or USEAGE. It also documents "semantics" - that is material that explains the "meaning" of the header contents, and other explanatory/introductory material, as in RFC 2822. See my reply to Forrest. For sure, USEPRO tells you how to construct a "followup", but if there is to be mention of "other" uses of References, then there is no obvious place in USEPRO to mention them (unless you can suggest one). The essential semantics of the References header, of course, is that it contains a list of "precursors", and the special significance of the order in which they are listed. But that in turn relies on a careful definition of "precursor" (which is another possible place to point out that they apply to more than "followups"). -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 From owner-ietf-usefor@mail.imc.org Wed Apr 27 13:13:41 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id NAA03923 for ; Wed, 27 Apr 2005 13:13:40 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RHCLDU056981 for ; Wed, 27 Apr 2005 10:12:21 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3RHCLIw056980 for ietf-usefor-skb; Wed, 27 Apr 2005 10:12:21 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from b.mail.peak.org (b.mail.peak.org [69.59.192.42]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RHCK6f056961 for ; Wed, 27 Apr 2005 10:12:20 -0700 (PDT) (envelope-from stanley@peak.org) Received: from a.shell.peak.org ([69.59.192.81]) by b.mail.peak.org (8.12.10/8.12.8) with ESMTP id j3RHCEtO001331 (version=TLSv1/SSLv3 cipher=EDH-RSA-DES-CBC3-SHA bits=168 verify=NO) for ; Wed, 27 Apr 2005 10:12:15 -0700 (PDT) Date: Wed, 27 Apr 2005 10:12:14 -0700 (PDT) From: John Stanley X-X-Sender: stanley@a.shell.peak.org To: ietf-usefor@imc.org Subject: Re: References Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Spam-Score: 0 () X-Scanned-By: MIMEDefang 2.39 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: I said: > It cannot be mandatory in USEPRO when it is already defined to be > optional in both USEFOR and RFC2822. Alexey Melnikov : >Surely it can, because USEFOR equally applies to *all* types of USEFOR >articles and USEPRO may add additional restrictions for particular types >of articles. While USEFOR applies to all types of articles, the section of USEFOR being discussed applies to the References header, and the section of RFC2822 that USEFOR defers to talks specifically about "replies". USEFOR says "see RFC2822", and RFC2822 says References is optional in replies. In addition, since USEFOR applies to ALL articles, a definition therein that makes a header optional makes it optional for ALL articles. >See my comment above. Stop being insulting, as if I would ignore your "comment above" until I saw your command to "see it" here. Frank Ellermann : >One WG member apparently wants an equivalences in the form >of followup <=> references and followup <=> 2822-reply, but >that's obviously wrong, Yes, that's obviously wrong. Can you tell us who it is who has said this? >and the proposed alternative of no >relation at all between followup and references is also wrong. It is not wrong if that is how the standards make it, and that is the direction that we are heading by making References optional. From owner-ietf-usefor@mail.imc.org Wed Apr 27 13:24:58 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id NAA04789 for ; Wed, 27 Apr 2005 13:24:58 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RHNhMe057860 for ; Wed, 27 Apr 2005 10:23:43 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3RHNhAj057859 for ietf-usefor-skb; Wed, 27 Apr 2005 10:23:43 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from soy.epix.net (soy.epix.net [199.224.64.64]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RHNfoS057853 for ; Wed, 27 Apr 2005 10:23:42 -0700 (PDT) (envelope-from mibsoft@epix.net) Received: from [192.168.2.11] (hrbg-199-224-121-129-pppoe.dsl.hrbg.epix.net [199.224.121.129]) by soy.epix.net (8.12.10/2004120601/PL) with ESMTP id j3RHNZgE028084; Wed, 27 Apr 2005 13:23:36 -0400 (EDT) Message-ID: <426FCA98.3070307@epix.net> Date: Wed, 27 Apr 2005 13:23:36 -0400 From: "Forrest J. Cavalier III" User-Agent: Mozilla Thunderbird 0.7 (Windows/20040616) X-Accept-Language: en-us, en MIME-Version: 1.0 To: Charles Lindsey CC: ietf-usefor@imc.org Subject: Re: Suggested References texts References: <426E6E66.2070109@epix.net> In-Reply-To: Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit X-Scanned-By: MIMEDefang 2.51 on 199.224.89.133 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Content-Transfer-Encoding: 7bit Charles Lindsey wrote: > Inevitably, this changes the information that a reading agent can deduce > when it sees a References header. It can no longer deduce "this is a > followup". It can only deduce "the poster wants you to treat this like a > followup" (by threading it, or by letting you retrieve the precursors by > clicking on them, or whatever other bells and whistles you provide). Do you agree we should be concentrating on how the reading agent can process articles? I think you do. And the text you suggested is flawed because it declares nothing of use for reading agents. (Here is the text:) >> | The References header is used in followups, and other articles with >> | related precursors (e.g. multipart FAQs and the fragments of a >> | message/partial), to facilitate the display or retrieval, by reading >> | and other agents, of threads of related articles. ... What does "facilitate" mean? What does "other articles" mean? Since "e.g." is not "i.e.", what other kinds of articles can have References? What stops someone from using References for other kinds of facilitation of display or retrieval? (Style sheets!) That text is useless. It doesn't even guide posters, much less readers. If people really want to allow FAQs and multi-part to get threaded, I wonder if the USEFOR text should be the following single sentence, (which cleverly does not mention followup or FAQ or multi-part.) A References header indicates the previously posted context in which the article was created. (That "the" means that you can only have at most one context per article.) I would then say to leave "followup" out of USEFOR, (and maybe out of USEPRO too, appearing in USEAGE, but I am not sure how all the documents are supposed to share duties.) From owner-ietf-usefor@mail.imc.org Wed Apr 27 13:36:05 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id NAA05392 for ; Wed, 27 Apr 2005 13:36:04 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RHZ8tU058515 for ; Wed, 27 Apr 2005 10:35:08 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3RHZ8gF058514 for ietf-usefor-skb; Wed, 27 Apr 2005 10:35:08 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from b.mail.peak.org (b.mail.peak.org [69.59.192.42]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RHZ7pK058504 for ; Wed, 27 Apr 2005 10:35:08 -0700 (PDT) (envelope-from stanley@peak.org) Received: from a.shell.peak.org ([69.59.192.81]) by b.mail.peak.org (8.12.10/8.12.8) with ESMTP id j3RHZ2wU012976 (version=TLSv1/SSLv3 cipher=EDH-RSA-DES-CBC3-SHA bits=168 verify=NO) for ; Wed, 27 Apr 2005 10:35:02 -0700 (PDT) Date: Wed, 27 Apr 2005 10:35:02 -0700 (PDT) From: John Stanley X-X-Sender: stanley@a.shell.peak.org To: ietf-usefor@imc.org Subject: Re: Differences between RFC 2822 and Usefor (fwd) Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Spam-Score: 0 () X-Scanned-By: MIMEDefang 2.39 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Richard Clayton : >the readers of Usenet believe in things like threads and followups; they >like the ability to split a series of rambling conversations into >disparate piles of articles and understand each of those piles one at a >time... If you read USENET for any length of time, you will find that readers (at least posters) of USENET believe in all kinds of oddball things. Does their belief in something elevate that thing into an interoperability issue, or even something that we need to deal with? >However, there seems to be an argument being presented that when you >look deeply at the specifications the notion of followups and threads is >mainly an illusion fostered by the reading tools and one can lawfully >miss out all the headers that allow the illusion to be constructed. I'm trying to understand what you are saying here, but it is couched in such odd language that I'm not certain. "Lawfully miss out?" If I get what you are saying, then yes, of course we can refuse to mandate a header that allows illusions to be propogated. In fact, propogating illusions is not the job of a technical standard. However, if you are trying to argue that "followup" is an important concept, and detecting such is an important action, then ok, maybe it is. Does it reach the level of importance that it is an interoperability issue? Will newsreaders break if they get an article without a References header? Does it truly merit a MUST, or is it only worth a SHOULD? >So it seems to me that this is a genuine inter-operability issue -- And yet, not a single thing you mentioned is interoperability. In fact, right now, news readers MUST deal with articles that do not have a References header. An article without a References header CANNOT be an interoperability issue, since the system would be collapsing today. >The issue occurs at the level at which the readers view the newsgroups. One newsreader dealing with one article for one reader is hardly an interoperability issue. What you've just described is a display issue, and we do not deal with display issues. Readers are free to display articles as they see fit. >For that reason I think >that a standardised way of expressing which pile an article belongs in >is of very significant value. That may be true. Since we are not undefining the References header, we are not removing the 'standardized way of expressing' anything. It's just been made optional. >Yes at present some software is a bit >flaky, but that doesn't mean we should fail to say what should be done. We already say what should be done. What change do you want made? >sounds like a cock-up :( Sounds like a decision and the natural consequences of that decision. >so best fix it then! ASAP!! Best determine that it is broken before demanding it be fixed. This group has been working for a long time; have you read the history of this discussion before telling us something we have done is broken? >ah.... perhaps I could find it earlier in the "thread" ? It isn't hard. It doesn't even require a References header, which is, by the way, optional in mail, and requires knowledge of the precursor message ids in any case. When the precursor message id is not known, it is improper to fabricate one. From owner-ietf-usefor@mail.imc.org Wed Apr 27 13:43:56 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id NAA05960 for ; Wed, 27 Apr 2005 13:43:56 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RHgQW1059003 for ; Wed, 27 Apr 2005 10:42:26 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3RHgQFq059002 for ietf-usefor-skb; Wed, 27 Apr 2005 10:42:26 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from a.mail.peak.org (a.mail.peak.org [69.59.192.41]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RHgP8R058984 for ; Wed, 27 Apr 2005 10:42:25 -0700 (PDT) (envelope-from stanley@peak.org) Received: from a.shell.peak.org ([69.59.192.81]) by a.mail.peak.org (8.12.10/8.12.8) with ESMTP id j3RHgJgU067732 (version=TLSv1/SSLv3 cipher=EDH-RSA-DES-CBC3-SHA bits=168 verify=NO) for ; Wed, 27 Apr 2005 10:42:19 -0700 (PDT) Date: Wed, 27 Apr 2005 10:42:19 -0700 (PDT) From: John Stanley X-X-Sender: stanley@a.shell.peak.org To: ietf-usefor@imc.org Subject: Re: References Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Spam-Score: 0 () X-Scanned-By: MIMEDefang 2.39 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: "Forrest J. Cavalier III" : >Did anyone ever advocate that 2822-reply => followup? Yes. It is pretty clear that all replies are followups. Look at the definitions we have come up with for "followup", and each one has included the set of articles that are "a response to another". Sometimes the definition also includes other relationships, but "response to another" has always been included. And that is a "reply". From owner-ietf-usefor@mail.imc.org Wed Apr 27 13:48:22 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id NAA06281 for ; Wed, 27 Apr 2005 13:48:22 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RHlfL6059544 for ; Wed, 27 Apr 2005 10:47:41 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3RHlfDP059543 for ietf-usefor-skb; Wed, 27 Apr 2005 10:47:41 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from anchor-post-36.mail.demon.net (anchor-post-36.mail.demon.net [194.217.242.86]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RHleEK059537 for ; Wed, 27 Apr 2005 10:47:40 -0700 (PDT) (envelope-from richard@highwayman.com) Received: from gti.noc.demon.net ([195.11.55.101] helo=happyday.al.cl.cam.ac.uk) by anchor-post-36.mail.demon.net with esmtp (Exim 4.42) id 1DQqdP-000Kz9-Ji for ietf-usefor@imc.org; Wed, 27 Apr 2005 17:47:39 +0000 Message-ID: <2PvOFUX9$8bCFA94@highwayman.com> Date: Wed, 27 Apr 2005 18:46:37 +0100 To: ietf-usefor@imc.org From: Richard Clayton Subject: Re: Differences between RFC 2822 and Usefor (fwd) References: In-Reply-To: MIME-Version: 1.0 X-Mailer: Turnpike Integrated Version 5.02 M Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 In message , John Stanley writes >However, if you are trying to argue that "followup" is an important >concept, and detecting such is an important action, then ok, maybe it is. >Does it reach the level of importance that it is an interoperability >issue? in my view yes >Will newsreaders break if they get an article without a References >header? one article they will cope with -- and their heuristics may mean that they function as if the header had been present -- but if no articles have References then they will fail to function as their users currently expect. I call that broken, and so will they. >Does it truly merit a MUST, or is it only worth a SHOULD? it should be a MUST >>So it seems to me that this is a genuine inter-operability issue -- > >And yet, not a single thing you mentioned is interoperability. I suspect that's because you're thinking about servers and not about the people who read Usenet and the software that they use for that. I gave some examples of how that causes problems earlier >We already say what should be done. What change do you want made? I believe that References: should be compulsory, and that all the other clues we currently have like "Re:", the unmangled subject, a later datestamp, and "In article xxx so-and-so said" can be made optional (or in the interests of tidyness abandoned altogether) >>sounds like a cock-up :( > >Sounds like a decision and the natural consequences of that decision. > >>so best fix it then! ASAP!! > >Best determine that it is broken before demanding it be fixed. This group >has been working for a long time; have you read the history of this >discussion before telling us something we have done is broken? oh yes, I've read a lot ... all the way down the years :( ... however, now it's clear that there's the prospect of (a) the effort finishing (one way or another) and (b) posters being permitted to break "threading" (and if there is no language to compel them, then they will) then it is time to speak up - -- richard Richard Clayton They that can give up essential liberty to obtain a little temporary safety deserve neither liberty nor safety. Benjamin Franklin -----BEGIN PGP SIGNATURE----- Version: PGPsdk version 1.7.1 iQA/AwUBQm/P/ZoAxkTY1oPiEQKq8ACgkMoFxn0uNpSEhbUFVTzVnhHVjbcAnjyZ v40IGNzGWVFSpus4Jt8cEJjL =JMS3 -----END PGP SIGNATURE----- From owner-ietf-usefor@mail.imc.org Wed Apr 27 13:51:20 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id NAA06535 for ; Wed, 27 Apr 2005 13:51:20 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RHoTxC059708 for ; Wed, 27 Apr 2005 10:50:29 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3RHoTe6059707 for ietf-usefor-skb; Wed, 27 Apr 2005 10:50:29 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from mail1.panix.com (mail1.panix.com [166.84.1.72]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RHoS3g059694 for ; Wed, 27 Apr 2005 10:50:28 -0700 (PDT) (envelope-from sethb@panix.com) Received: from panix5.panix.com (panix5.panix.com [166.84.1.5]) by mail1.panix.com (Postfix) with ESMTP id CA10558ACC for ; Wed, 27 Apr 2005 13:50:27 -0400 (EDT) Received: (from sethb@localhost) by panix5.panix.com (8.11.6p3/8.8.8/PanixN1.1) id j3RHoR604675; Wed, 27 Apr 2005 13:50:27 -0400 (EDT) Date: Wed, 27 Apr 2005 13:50:27 -0400 (EDT) Message-Id: <200504271750.j3RHoR604675@panix5.panix.com> From: Seth Breidbart To: ietf-usefor@imc.org In-reply-to: (message from John Stanley on Wed, 27 Apr 2005 10:35:02 -0700 (PDT)) Subject: Re: Differences between RFC 2822 and Usefor (fwd) References: Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: John Stanley wrote: > However, if you are trying to argue that "followup" is an important > concept, and detecting such is an important action, then ok, maybe > it is. Does it reach the level of importance that it is an > interoperability issue? Will newsreaders break if they get an > article without a References header? Does it truly merit a MUST, or > is it only worth a SHOULD? Define "break". If you mean "crash and burn and let the magic smoke out of the computer" then no. If you mean "Do the wrong thing (from the viewpoints of both the poster and reader)" then yes. I'd prefer a MUST but would settle for a SHOULD. > And yet, not a single thing you mentioned is interoperability. In > fact, right now, news readers MUST deal with articles that do not > have a References header. An article without a References header > CANNOT be an interoperability issue, since the system would be > collapsing today. If the system continues to act, but sometimes acts wrongly, it isn't collapsing. But whatever is making it act wrongly _is_ an interoperability issue. If a random set of Subject header content was changed to "3" the system wouldn't collapse, either; so why is specifying that some headers MUST NOT be changed by transit servers an interoperability issue? Answer: That's doing _the wrong thing_. Seth From owner-ietf-usefor@mail.imc.org Wed Apr 27 13:59:01 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id NAA06958 for ; Wed, 27 Apr 2005 13:59:00 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RHvsqN060105 for ; Wed, 27 Apr 2005 10:57:54 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3RHvsUC060104 for ietf-usefor-skb; Wed, 27 Apr 2005 10:57:54 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtp3.Stanford.EDU (smtp3.Stanford.EDU [171.67.16.138]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RHvrKA060098 for ; Wed, 27 Apr 2005 10:57:54 -0700 (PDT) (envelope-from rra@stanford.edu) Received: from windlord.stanford.edu (windlord.Stanford.EDU [171.64.19.147]) by smtp3.Stanford.EDU (8.12.11/8.12.11) with SMTP id j3RHvpWe023780 for ; Wed, 27 Apr 2005 10:57:51 -0700 Received: (qmail 9748 invoked by uid 1000); 27 Apr 2005 17:57:51 -0000 To: ietf-usefor@imc.org Subject: Re: Obsolete vs. obsolescent In-Reply-To: <426F4F7C.7385@xyzzy.claranet.de> (Frank Ellermann's message of "Wed, 27 Apr 2005 10:38:20 +0200") References: <200504111212.j3BCChM14516@clerew.man.ac.uk> <87psx1gnjc.fsf@windlord.stanford.edu> <87fyxvx1m6.fsf@windlord.stanford.edu> <87mzs2a1hu.fsf@windlord.stanford.edu> <87wtqxui6s.fsf@windlord.stanford.edu> <87d5so2bpk.fsf@windlord.stanford.edu> <42697BB6.6074@xyzzy.claranet.de> <426BC455.2040108@isode.com> <426E4B45.7010205@oceana.com> <426F4F7C.7385@xyzzy.claranet.de> From: Russ Allbery Organization: The Eyrie Date: Wed, 27 Apr 2005 10:57:51 -0700 Message-ID: <87k6moksps.fsf@windlord.stanford.edu> User-Agent: Gnus/5.1007 (Gnus v5.10.7) XEmacs/21.4 (Jumbo Shrimp, linux) MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Frank Ellermann writes: > IMHO all that should be done is to update this statement: > | As a result, this header is to be regarded as obsolete, and > | it will likely be removed entirely in a future version of > | this standard. > As a result, this header is to be regarded as obsolescent. > Servers and UAs MUST accept but SHOULD ignore it, and SHOULD > NOT generate it. I dislike saying that servers MUST accept a header, since servers are not required to accept *any* article. I would also leave in the bit about removing it in a future version of the standard. > Subject: cmsg is not mentioned anywhere in usefor-03, that's a > serious bug. It's mentioned in usepro-03, but it only says > that this is now obsolete. That's of course FUBAR, it's a > severe security violation. We cannot jump from RfC 1036 to a > completely different procedure at date X. AFAIK RfCs must not > have a flag day. I think USEFOR does need an explicit mention that subjects starting with cmsg should not be considered control messages, to document the change from RFC 1036. -- Russ Allbery (rra@stanford.edu) From owner-ietf-usefor@mail.imc.org Wed Apr 27 14:02:02 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id OAA07323 for ; Wed, 27 Apr 2005 14:02:01 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RI0fww060281 for ; Wed, 27 Apr 2005 11:00:41 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3RI0fZD060280 for ietf-usefor-skb; Wed, 27 Apr 2005 11:00:41 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from a.mail.peak.org (a.mail.peak.org [69.59.192.41]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RI0eZI060268 for ; Wed, 27 Apr 2005 11:00:40 -0700 (PDT) (envelope-from stanley@peak.org) Received: from a.shell.peak.org ([69.59.192.81]) by a.mail.peak.org (8.12.10/8.12.8) with ESMTP id j3RI0XgU077455 (version=TLSv1/SSLv3 cipher=EDH-RSA-DES-CBC3-SHA bits=168 verify=NO) for ; Wed, 27 Apr 2005 11:00:35 -0700 (PDT) Date: Wed, 27 Apr 2005 11:00:33 -0700 (PDT) From: John Stanley X-X-Sender: stanley@a.shell.peak.org To: ietf-usefor@imc.org Subject: Re: Suggested References texts Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Spam-Score: 0 () X-Scanned-By: MIMEDefang 2.39 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: "Charles Lindsey" : >Let us recap the history here. It is accepted, I believe, (using Frank >Ellermann's notation) Frank's notation is confusing. Does '=>' mean "is a member of", "implies", or something else? Simple english works so much better here, I suggest it be used. > followup -> references header A followup requires a References header? (Yes under RFC1036, no under RFC2822). A followup contains a References header? A followup is defined by a References header? What? > no references header -> not a followup Ahhh, perhaps '=>' means "means"? (Is '->' the same as '=>'?) So, no References means an article is not a followup? Well, yes, according to RFC1036, not according to USEFOR draft 03 and RFC2822. >and at one time we had > references header -> followup And then you changed it. >But then, some months back, it was pointed out that some writers of multipart >FAQs regularly used References headers to join the later parts to the >earlier ones, and RFC 2046 suggests the use of References headers in >message/partial. And there were other examples too. None of which required the change you made. >Did we want to forbid these practices, and would they actually disappear >even if we did? So the choice was to evicerate the References header instead of expand the definition of followup trivially. >Inevitably, this changes the information that a reading agent can deduce >when it sees a References header. Wrong. Making References legal in non-followups changed the information that could be deduced. "Is this a followup" cannot be answered using your changes. It could before you made them. >It can no longer deduce "this is a followup". I keep telling you this, and yet, you keep telling me that this is not a change. >So you can no longer assert "references header -> followup". The question >is 'Does this matter?'. I don't see why it should. That is your value judgement, not a fact. In fact, many people do think it matters, and many people want that ability. Richard Clayton was just here arguing that it is important, and that he even considers it an interoperability issue. Do you understand the difference between saying "I don't think it is important and thus shouldn't be possible" and "I don't think it is important enough to merit RFC2119 mandates?" But now that it is gone, we need to unify on one position. Either it is or it isn't important. One document written as if it isn't and one written as if it is is just ridiculous. Pick one and stick with it for more than a week. From owner-ietf-usefor@mail.imc.org Wed Apr 27 14:08:06 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id OAA07904 for ; Wed, 27 Apr 2005 14:08:06 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RI4aJs061123 for ; Wed, 27 Apr 2005 11:04:36 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3RI4arC061122 for ietf-usefor-skb; Wed, 27 Apr 2005 11:04:36 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtp1.Stanford.EDU (smtp1.Stanford.EDU [171.67.16.123]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RI4ZhF061114 for ; Wed, 27 Apr 2005 11:04:35 -0700 (PDT) (envelope-from rra@stanford.edu) Received: from windlord.stanford.edu (windlord.Stanford.EDU [171.64.19.147]) by smtp1.Stanford.EDU (8.12.11/8.12.11) with SMTP id j3RI4YtV004185 for ; Wed, 27 Apr 2005 11:04:35 -0700 Received: (qmail 10674 invoked by uid 1000); 27 Apr 2005 18:04:34 -0000 To: ietf-usefor@imc.org Subject: Re: References In-Reply-To: (John Stanley's message of "Wed, 27 Apr 2005 10:42:19 -0700 (PDT)") References: From: Russ Allbery Organization: The Eyrie Date: Wed, 27 Apr 2005 11:04:34 -0700 Message-ID: <87fyxcksel.fsf@windlord.stanford.edu> User-Agent: Gnus/5.1007 (Gnus v5.10.7) XEmacs/21.4 (Jumbo Shrimp, linux) MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: John Stanley writes: > "Forrest J. Cavalier III" : >> Did anyone ever advocate that 2822-reply => followup? > Yes. It is pretty clear that all replies are followups. What about, er, replies? In the trn sense. An e-mail reply to a Usenet message. Those are what RFC 2822 more obviously applies to; followups in the Usenet sense have additional restrictions that RFC 2822 has no comment on and aren't really something RFC 2822 was trying to deal with. This thread is convincing me that it's a bit confusing to defer to RFC 2822 entirely for References when RFC 2822 only talks about replies and clearly is intended to focus on mail messages. Usenet followups are a bit of a different beast with historically stronger restrictions on the presence of the References header that e-mail is only now starting to adopt. I still feel like this could be dealt with definitionally, but followups in the Usenet sense have References headers, period. It's not just a SHOULD. And I don't feel like that's a divergence from RFC 2822, since RFC 2822 doesn't deal with Usenet followups and the semantics of the References header haven't changed. -- Russ Allbery (rra@stanford.edu) From owner-ietf-usefor@mail.imc.org Wed Apr 27 14:09:30 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id OAA07948 for ; Wed, 27 Apr 2005 14:09:29 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RI8q2g062183 for ; Wed, 27 Apr 2005 11:08:52 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3RI8qtJ062182 for ietf-usefor-skb; Wed, 27 Apr 2005 11:08:52 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtp1.Stanford.EDU (smtp1.Stanford.EDU [171.67.16.123]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RI8q9I062175 for ; Wed, 27 Apr 2005 11:08:52 -0700 (PDT) (envelope-from rra@stanford.edu) Received: from windlord.stanford.edu (windlord.Stanford.EDU [171.64.19.147]) by smtp1.Stanford.EDU (8.12.11/8.12.11) with SMTP id j3RI8ne4005898 for ; Wed, 27 Apr 2005 11:08:49 -0700 Received: (qmail 10732 invoked by uid 1000); 27 Apr 2005 18:08:49 -0000 To: ietf-usefor@imc.org Subject: Re: Suggested References texts In-Reply-To: (John Stanley's message of "Wed, 27 Apr 2005 09:59:15 -0700 (PDT)") References: From: Russ Allbery Organization: The Eyrie Date: Wed, 27 Apr 2005 11:08:49 -0700 Message-ID: <87br80ks7i.fsf@windlord.stanford.edu> User-Agent: Gnus/5.1007 (Gnus v5.10.7) XEmacs/21.4 (Jumbo Shrimp, linux) MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: John Stanley writes: > Russ Allbery : >> If this is a requirement, then we have to define everything that >> contains a References header as a followup, > You have it backwards. We define a followup by the poster's intent; we > DETECT followups with the References header. There is a difference. Your > definition leaves us with a circle: a followup contains a References > header, and a References header must be used in a followup. Ok, what's a > followup? Either we define followups by poster intent and some articles with Reference headers aren't followups, or we define followups by the presence of a Reference header and we can't make a strong statement about intent. We can't have it both ways; Reference headers are used in practice for things other than followups, and I don't expect that practice to change (nor do I think we really *want* it to change). -- Russ Allbery (rra@stanford.edu) From owner-ietf-usefor@mail.imc.org Wed Apr 27 14:44:41 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id OAA10957 for ; Wed, 27 Apr 2005 14:44:40 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RIhEc8067300 for ; Wed, 27 Apr 2005 11:43:14 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3RIhEEY067299 for ietf-usefor-skb; Wed, 27 Apr 2005 11:43:14 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from eagle.oceana.com (eagle.oceana.com [208.17.123.12]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RIhDu9067275 for ; Wed, 27 Apr 2005 11:43:13 -0700 (PDT) (envelope-from ken@oceana.com) Received: from [192.168.10.26] (KEN.oceana.com [192.168.10.26]) by eagle.oceana.com (8.13.2/8.13.2) with ESMTP id j3RIh6wi005248; Wed, 27 Apr 2005 14:43:06 -0400 (EDT) Message-ID: <426FDD09.5030102@oceana.com> Date: Wed, 27 Apr 2005 14:42:17 -0400 From: Ken Murchison Organization: Oceana Matrix Ltd. User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.7) Gecko/20040514 X-Accept-Language: en,pdf MIME-Version: 1.0 To: ietf-usefor@imc.org CC: Alexey Melnikov Subject: Re: References References: <87fyxcksel.fsf@windlord.stanford.edu> In-Reply-To: <87fyxcksel.fsf@windlord.stanford.edu> Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit X-Spam-Score: 0 (BAYES_00) X-Scanned-By: MIMEDefang 2.39 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Content-Transfer-Encoding: 7bit Russ Allbery wrote: > This thread is convincing me that it's a bit confusing to defer to RFC > 2822 entirely for References when RFC 2822 only talks about replies and > clearly is intended to focus on mail messages. Usenet followups are a bit > of a different beast with historically stronger restrictions on the > presence of the References header that e-mail is only now starting to > adopt. I still feel like this could be dealt with definitionally, but > followups in the Usenet sense have References headers, period. It's not > just a SHOULD. And I don't feel like that's a divergence from RFC 2822, > since RFC 2822 doesn't deal with Usenet followups and the semantics of the > References header haven't changed. I'm leaning towards the same conclusion. Deferring to RFC 2822 in draft-03 is looking like bad idea. As Charles stated, USEFOR discusses semantics elsewhere so we can and probably should discuss the use of the References header in a more Usenet-centric way. Does our Chair agree? -- Kenneth Murchison Oceana Matrix Ltd. Software Engineer 21 Princeton Place 716-662-8973 x26 Orchard Park, NY 14127 --PGP Public Key-- http://www.oceana.com/~ken/ksm.pgp From owner-ietf-usefor@mail.imc.org Wed Apr 27 15:21:40 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id PAA15257 for ; Wed, 27 Apr 2005 15:21:39 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RJKGYN072811 for ; Wed, 27 Apr 2005 12:20:16 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3RJKGww072810 for ietf-usefor-skb; Wed, 27 Apr 2005 12:20:16 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from rufus.isode.com (rufus.isode.com [62.3.217.251]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RJKFi0072804 for ; Wed, 27 Apr 2005 12:20:15 -0700 (PDT) (envelope-from alexey.melnikov-usefor@isode.com) Received: from [172.16.2.180] (shiny.isode.com [62.3.217.250]) by rufus.isode.com via TCP (internal) with ESMTPA; Wed, 27 Apr 2005 20:20:11 +0100 Message-ID: <426FE5EB.5030302@isode.com> Date: Wed, 27 Apr 2005 20:20:11 +0100 From: Alexey Melnikov User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.7.2) Gecko/20040804 Netscape/7.2 (ax) X-Accept-Language: en-us, en To: Ken Murchison CC: ietf-usefor@imc.org Subject: Re: References References: <87fyxcksel.fsf@windlord.stanford.edu> <426FDD09.5030102@oceana.com> In-Reply-To: <426FDD09.5030102@oceana.com> MIME-version: 1.0 Content-type: text/plain; charset="ISO-8859-1"; format="flowed" Content-transfer-encoding: 7bit Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Content-Transfer-Encoding: 7bit Ken Murchison wrote: > Russ Allbery wrote: > >> This thread is convincing me that it's a bit confusing to defer to RFC >> 2822 entirely for References when RFC 2822 only talks about replies and >> clearly is intended to focus on mail messages. Usenet followups are >> a bit >> of a different beast with historically stronger restrictions on the >> presence of the References header that e-mail is only now starting to >> adopt. I still feel like this could be dealt with definitionally, but >> followups in the Usenet sense have References headers, period. It's not >> just a SHOULD. And I don't feel like that's a divergence from RFC 2822, >> since RFC 2822 doesn't deal with Usenet followups and the semantics >> of the >> References header haven't changed. > > I'm leaning towards the same conclusion. Deferring to RFC 2822 in > draft-03 is looking like bad idea. As Charles stated, USEFOR > discusses semantics elsewhere so we can and probably should discuss > the use of the References header in a more Usenet-centric way. > > Does our Chair agree? I do, but the devil is in details: what exactly should be said and where. From owner-ietf-usefor@mail.imc.org Wed Apr 27 15:30:58 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id PAA16097 for ; Wed, 27 Apr 2005 15:30:58 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RJU8mR074222 for ; Wed, 27 Apr 2005 12:30:08 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3RJU8Fa074221 for ietf-usefor-skb; Wed, 27 Apr 2005 12:30:08 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from rufus.isode.com (rufus.isode.com [62.3.217.251]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RJU792074207 for ; Wed, 27 Apr 2005 12:30:07 -0700 (PDT) (envelope-from alexey.melnikov-usefor@isode.com) Received: from [172.16.2.180] (shiny.isode.com [62.3.217.250]) by rufus.isode.com via TCP (internal) with ESMTPA; Wed, 27 Apr 2005 20:30:06 +0100 Message-ID: <426FE83E.3040503@isode.com> Date: Wed, 27 Apr 2005 20:30:06 +0100 From: Alexey Melnikov User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.7.2) Gecko/20040804 Netscape/7.2 (ax) X-Accept-Language: en-us, en To: ietf-usefor@imc.org Subject: Re: References References: <87fyxcksel.fsf@windlord.stanford.edu> <426FDD09.5030102@oceana.com> <426FE5EB.5030302@isode.com> In-Reply-To: <426FE5EB.5030302@isode.com> MIME-version: 1.0 Content-type: text/plain; charset="ISO-8859-1"; format="flowed" Content-transfer-encoding: 7bit Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Content-Transfer-Encoding: 7bit Alexey Melnikov wrote: > Ken Murchison wrote: > >> Russ Allbery wrote: >> >>> This thread is convincing me that it's a bit confusing to defer to RFC >>> 2822 entirely for References when RFC 2822 only talks about replies and >>> clearly is intended to focus on mail messages. Usenet followups are >>> a bit >>> of a different beast with historically stronger restrictions on the >>> presence of the References header that e-mail is only now starting to >>> adopt. I still feel like this could be dealt with definitionally, but >>> followups in the Usenet sense have References headers, period. It's >>> not >>> just a SHOULD. And I don't feel like that's a divergence from RFC >>> 2822, >>> since RFC 2822 doesn't deal with Usenet followups and the semantics >>> of the >>> References header haven't changed. >> >> I'm leaning towards the same conclusion. Deferring to RFC 2822 in >> draft-03 is looking like bad idea. As Charles stated, USEFOR >> discusses semantics elsewhere so we can and probably should discuss >> the use of the References header in a more Usenet-centric way. >> >> Does our Chair agree? > > I do, but the devil is in details: what exactly should be said and where. The USEFOR document can say that certain types of messages MAY require the References header. But I just think this should be obvious. From owner-ietf-usefor@mail.imc.org Wed Apr 27 17:19:04 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id RAA07698 for ; Wed, 27 Apr 2005 17:19:03 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RLHTX5089026 for ; Wed, 27 Apr 2005 14:17:29 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3RLHTZT089025 for ietf-usefor-skb; Wed, 27 Apr 2005 14:17:29 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from b.mail.peak.org (b.mail.peak.org [69.59.192.42]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RLHSGU089007 for ; Wed, 27 Apr 2005 14:17:29 -0700 (PDT) (envelope-from stanley@peak.org) Received: from a.shell.peak.org ([69.59.192.81]) by b.mail.peak.org (8.12.10/8.12.8) with ESMTP id j3RLHMwU009175 (version=TLSv1/SSLv3 cipher=EDH-RSA-DES-CBC3-SHA bits=168 verify=NO) for ; Wed, 27 Apr 2005 14:17:23 -0700 (PDT) Date: Wed, 27 Apr 2005 14:17:21 -0700 (PDT) From: John Stanley X-X-Sender: stanley@a.shell.peak.org To: ietf-usefor@imc.org Subject: Re: Differences between RFC 2822 and Usefor (fwd) Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Spam-Score: 0 () X-Scanned-By: MIMEDefang 2.39 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Richard Clayton : > one article they will cope with -- Many articles they will cope with, I hope. Many articles have no References header. >-- but if no articles >have References then they will fail to function as their users currently >expect. I call that broken, and so will they. Once again we are stuck with the sometimes unusual expectations of USENET users. And how did we get to "no articles have References headers?" Are you imagining some prohibition on the use of References headers? >I believe that References: should be compulsory, RFC2119 defines when MUST is appropriate. This is not such a time. >however, >now it's clear that there's the prospect of (a) the effort finishing >(one way or another) and (b) posters being permitted to break >"threading" (and if there is no language to compel them, then they will) >then it is time to speak up Posters are always permitted to "break threading". They will always be permitted to "break threading". Seth Breidbart : >Define "break". If you mean "crash and burn and let the magic smoke >out of the computer" then no. Well, let's see. "Break" as in "fail to process articles", ok? No, since we are talking about interoperability, that would have to be "fail to accept articles or to pass them to other news system software." I read news today. I found articles (gasp!) without References headers. My newsreader did not break. It simply did not care. >If you mean "Do the wrong thing (from >the viewpoints of both the poster and reader)" then yes. Since this is a display issue, what is "the wrong thing" and how do you know when it happens? "Wrong" is completely subjective. Subjective is not how one writes standards. In this specific case, some people think that "right" is "sort by subject content". Some think that "sort by precursor" is "right" and "sort by subject" is wrong. Who is "right" and who is "wrong"? >If a random set of Subject header content was changed to "3" the >system wouldn't collapse, either; so why is specifying that some >headers MUST NOT be changed by transit servers an interoperability >issue? Because transport agents are DEFINED not to change the message they are transporing, and the header you are referring to is a USER controlled header, not a news system controlled one. >Answer: That's doing _the wrong thing_. That's your opinion. I could imagine a system where the Subject header IS changed in transport that is quite correct. In fact, some email spam systems do exactly that. From owner-ietf-usefor@mail.imc.org Wed Apr 27 17:38:12 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id RAA10394 for ; Wed, 27 Apr 2005 17:38:12 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RLbWu6092074 for ; Wed, 27 Apr 2005 14:37:32 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3RLbWBo092073 for ietf-usefor-skb; Wed, 27 Apr 2005 14:37:32 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from a.mail.peak.org (a.mail.peak.org [69.59.192.41]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RLbVux092055 for ; Wed, 27 Apr 2005 14:37:31 -0700 (PDT) (envelope-from stanley@peak.org) Received: from a.shell.peak.org ([69.59.192.81]) by a.mail.peak.org (8.12.10/8.12.8) with ESMTP id j3RLbPgU086379 (version=TLSv1/SSLv3 cipher=EDH-RSA-DES-CBC3-SHA bits=168 verify=NO) for ; Wed, 27 Apr 2005 14:37:25 -0700 (PDT) Date: Wed, 27 Apr 2005 14:37:25 -0700 (PDT) From: John Stanley X-X-Sender: stanley@a.shell.peak.org To: ietf-usefor@imc.org Subject: Re: References Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Spam-Score: 0 () X-Scanned-By: MIMEDefang 2.39 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Russ Allbery : Me: Yes. It is pretty clear that all replies are followups. >What about, er, replies? In the trn sense. An e-mail reply to a Usenet >message. Those are what RFC 2822 more obviously applies to; RFC2822 does not say "email replies". Since USEFOR defers to RFC2822, we have chosen to expand any "email" limitation of RFC2822 into USENET. If we are so pedantic that we say "RFC2822 is EMAIL only" even in the face of explicit references to RFC2822 by draft news standards, then exactly what headers DOES news have and how are they defined? Many of them are defined as "as found in RFC2822". But since those are EMAIL headers, then they don't apply to news. Sigh. >This thread is convincing me that it's a bit confusing to defer to RFC >2822 entirely for References when RFC 2822 only talks about replies and >clearly is intended to focus on mail messages. Usenet followups are a bit >of a different beast with historically stronger restrictions on the >presence of the References header that e-mail is only now starting to >adopt. "Historically stronger" is not found in RFC2119 as a reason for mandates, and "historically stronger" is apparently not strong enough to have resulted in the mandates being continued into draft-03 of USEFOR. >I still feel like this could be dealt with definitionally, but >followups in the Usenet sense have References headers, period. It's not >just a SHOULD. Currently, it is just a SHOULD. If we choose to return it to MUST, we can do so, but what do we use to justify that change, and then, do we have the nerve to do it right or do we continue to contradict ourselves? >Either we define followups by poster intent and some articles with >Reference headers aren't followups, The latter does not follow from the former. We can easily (and have, in the past) prohibit References headers in non-followups. > or we define followups by the presence >of a Reference header and we can't make a strong statement about intent. So I can change an article from being a followup to not being a followup simply by removing one header? It seems odd that I can change the origination of an article by simply removing a header when his article reaches me halfway around the world from where he posted it. >We can't have it both ways; Reference headers are used in practice for >things other than followups, and I don't expect that practice to change >(nor do I think we really *want* it to change). We can solve this problem by simply expanding trivially the definition of followup. We don't have to dick with the mandates for when a References header is or is not to appear in an article. But "we" chose the latter course, to the point that they are now optional and followups are indetectable. Ken Murchison : >...USEFOR discusses semantics elsewhere so we can and probably should >discuss the use of the References header in a more Usenet-centric way. And exactly what shall our excuse be for ignoring RFC2119 after removing mandates regarding a header whose mandates have always been a point of debate? From owner-ietf-usefor@mail.imc.org Wed Apr 27 19:37:18 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id TAA19813 for ; Wed, 27 Apr 2005 19:37:17 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RNZD7n003222 for ; Wed, 27 Apr 2005 16:35:13 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3RNZDi0003221 for ietf-usefor-skb; Wed, 27 Apr 2005 16:35:13 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from b.mail.peak.org (b.mail.peak.org [69.59.192.42]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RNZCoZ003213 for ; Wed, 27 Apr 2005 16:35:13 -0700 (PDT) (envelope-from stanley@peak.org) Received: from a.shell.peak.org ([69.59.192.81]) by b.mail.peak.org (8.12.10/8.12.8) with ESMTP id j3RNZ6wU068498 (version=TLSv1/SSLv3 cipher=EDH-RSA-DES-CBC3-SHA bits=168 verify=NO) for ; Wed, 27 Apr 2005 16:35:07 -0700 (PDT) Date: Wed, 27 Apr 2005 16:35:06 -0700 (PDT) From: John Stanley X-X-Sender: stanley@a.shell.peak.org To: ietf-usefor@imc.org Subject: Re: References Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Spam-Score: 0 () X-Scanned-By: MIMEDefang 2.39 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Alexey Melnikov : >The USEFOR document can say that certain types of messages MAY require the >References header. But I just think this should be obvious. It is interesting to note that just this morning in the office we had a discussion about the use of the terms "obvious" and "clearly", and that when someone says something is either "obvious" or "clearly", it usually isn't. In this case, USEFOR defers to RFC2822, which clearly and obviously says that References is optional -- specifically, it uses the RFC2119 term "SHOULD", and in some places only "may", and further talks about how to process replies that do not contain a References header. It is hard to imagine that this is a requirement. Saying that USEFOR "obviously" says that some articles require a References header is, well, contrary to what USEFOR actually says. That's fine, it has been changed and we can live with that, as long as the change is propogated to all relevant documents, which includes USEPRO. We could change it back, but so far only a very small handful of people have been arguing that RFC2119 language is justified, and even then they haven't been providing any examples of interoperability issues, just things like "ensure that definitions are obeyed" and they think it should say "MUST". From owner-ietf-usefor@mail.imc.org Thu Apr 28 07:15:46 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id HAA04239 for ; Thu, 28 Apr 2005 07:15:45 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3SBEO4H076547 for ; Thu, 28 Apr 2005 04:14:24 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3SBEMaR076546 for ietf-usefor-skb; Thu, 28 Apr 2005 04:14:22 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from ciao.gmane.org (main.gmane.org [80.91.229.2]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3SBEKc4076512 for ; Thu, 28 Apr 2005 04:14:21 -0700 (PDT) (envelope-from usenet-format@gmane.org) Received: from list by ciao.gmane.org with local (Exim 4.43) id 1DR6sb-0003D4-Si for ietf-usefor@imc.org; Thu, 28 Apr 2005 13:08:25 +0200 Received: from c-134-88-108.hh.dial.de.ignite.net ([62.134.88.108]) by main.gmane.org with esmtp (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Thu, 28 Apr 2005 13:08:25 +0200 Received: from nobody by c-134-88-108.hh.dial.de.ignite.net with local (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Thu, 28 Apr 2005 13:08:25 +0200 X-Injected-Via-Gmane: http://gmane.org/ To: ietf-usefor@imc.org From: Frank Ellermann Subject: Re: References Date: Thu, 28 Apr 2005 13:12:15 +0200 Organization: Lines: 17 Message-ID: <4270C50F.1637@xyzzy.claranet.de> References: <87fyxcksel.fsf@windlord.stanford.edu> <426FDD09.5030102@oceana.com> <426FE5EB.5030302@isode.com> <426FE83E.3040503@isode.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Complaints-To: usenet@sea.gmane.org X-Gmane-NNTP-Posting-Host: c-134-88-108.hh.dial.de.ignite.net X-Mailer: Mozilla 3.0 (OS/2; U) Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Content-Transfer-Encoding: 7bit Alexey Melnikov wrote: > I just think this should be obvious. Sure. It was discussed here for the major part of 2004. News is not the same as mail and 2822. There's a note in s-o-1036 explaining why some References should be a See-Also, but we know that See-Also didn't fly. So we have References used for followups as defined in Usepro, and we have References as defined in STD 11 4.6, and unfortunately that's not always identical. What else is new ? Is Usefor supposed to document common practice or wishful thinking based on 2822 ? Bye, Frank From owner-ietf-usefor@mail.imc.org Thu Apr 28 08:09:36 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id IAA08553 for ; Thu, 28 Apr 2005 08:09:36 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3SC8Kv0098885 for ; Thu, 28 Apr 2005 05:08:20 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3SC8K1M098884 for ietf-usefor-skb; Thu, 28 Apr 2005 05:08:20 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from ciao.gmane.org (main.gmane.org [80.91.229.2]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3SC8IpU098867 for ; Thu, 28 Apr 2005 05:08:19 -0700 (PDT) (envelope-from usenet-format@gmane.org) Received: from list by ciao.gmane.org with local (Exim 4.43) id 1DR7hx-0000fZ-8B for ietf-usefor@imc.org; Thu, 28 Apr 2005 14:01:29 +0200 Received: from c-134-88-108.hh.dial.de.ignite.net ([62.134.88.108]) by main.gmane.org with esmtp (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Thu, 28 Apr 2005 14:01:29 +0200 Received: from nobody by c-134-88-108.hh.dial.de.ignite.net with local (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Thu, 28 Apr 2005 14:01:29 +0200 X-Injected-Via-Gmane: http://gmane.org/ To: ietf-usefor@imc.org From: Frank Ellermann Subject: Re: References Date: Thu, 28 Apr 2005 14:04:33 +0200 Organization: Lines: 54 Message-ID: <4270D151.776C@xyzzy.claranet.de> References: <426E579F.1060809@oceana.com> <426F3B14.54C1@xyzzy.claranet.de> <426F7411.2010903@epix.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Complaints-To: usenet@sea.gmane.org X-Gmane-NNTP-Posting-Host: c-134-88-108.hh.dial.de.ignite.net X-Mailer: Mozilla 3.0 (OS/2; U) Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Content-Transfer-Encoding: 7bit Forrest J. Cavalier III wrote: > I think it is a requirement that a news client be able to > determine "followup" by inspecting the article headers. Maybe you like to join Bruce and some others including me in our time travel excursion to fix some RfCs, you could be the advisor to Dave Crocker when he fixes STD 11 4.6.3. We could do radical things like publishing s-o-1036 as RfC. Seriously, this "requirement" is wishful thinking, you find it as See-Also in s-o-1036. All we can do today is to say how a followup is supposed to be created and how References are trimmed. We cannot guarantee that all References were built this way. The very best we can do is to be silent about more creative ways to generate References. > If it is not a requirement, then the term "followup" is > meaningless and should not appear in any document. "Followup" has a perfect meaning defined by Followup-To (or Newsgroups). > What is the READER agent supposed to do in that situation? It's supposed to expect the worst, but to work for the 99.9% ordinary followups. In the same way as it did for the last decade. BTW, I've tested the effect of an article referencing itself on this list some moths ago. Let's say that I won't test this again. > Is there a term for articles that can have references but > are not followups? How about "literal Reference" ? "STD 11 4.6.3 Reference" ? "Multi-2822 quasi-In-Reply-To" ? Or good old "See-Also" ? > why the special term "followup"? Maybe ask Henry or Bruce, it must be very old, "Followup-To" is very old. For me nothing before s-o-1036 is real, all I knew about Usenet before s-o-1036 was "Emily Postnews", and the fascinating differences from FidoNet, because I used to hang out in a gateways NG ("echo" from my POV) for some time. That's also the reason why I consider attempts to devalue Usenet Message-IDs to some as blasphemy. Bye, Frank From owner-ietf-usefor@mail.imc.org Thu Apr 28 09:05:59 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id JAA12203 for ; Thu, 28 Apr 2005 09:05:59 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3SD2iKk021418 for ; Thu, 28 Apr 2005 06:02:44 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3SD2ini021417 for ietf-usefor-skb; Thu, 28 Apr 2005 06:02:44 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from ciao.gmane.org (main.gmane.org [80.91.229.2]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3SD2gx6021392 for ; Thu, 28 Apr 2005 06:02:43 -0700 (PDT) (envelope-from usenet-format@gmane.org) Received: from list by ciao.gmane.org with local (Exim 4.43) id 1DR8YZ-0007Ye-4r for ietf-usefor@imc.org; Thu, 28 Apr 2005 14:55:51 +0200 Received: from c-134-88-108.hh.dial.de.ignite.net ([62.134.88.108]) by main.gmane.org with esmtp (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Thu, 28 Apr 2005 14:55:51 +0200 Received: from nobody by c-134-88-108.hh.dial.de.ignite.net with local (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Thu, 28 Apr 2005 14:55:51 +0200 X-Injected-Via-Gmane: http://gmane.org/ To: ietf-usefor@imc.org From: Frank Ellermann Subject: Re: Obsolete vs. obsolescent Date: Thu, 28 Apr 2005 14:59:39 +0200 Organization: Lines: 34 Message-ID: <4270DE3B.4BCC@xyzzy.claranet.de> References: <200504111212.j3BCChM14516@clerew.man.ac.uk> <87psx1gnjc.fsf@windlord.stanford.edu> <87fyxvx1m6.fsf@windlord.stanford.edu> <87mzs2a1hu.fsf@windlord.stanford.edu> <87wtqxui6s.fsf@windlord.stanford.edu> <87d5so2bpk.fsf@windlord.stanford.edu> <42697BB6.6074@xyzzy.claranet.de> <426BC455.2040108@isode.com> <426E4B45.7010205@oceana.com> <426F4F7C.7385@xyzzy.claranet.de> <87k6moksps.fsf@windlord.stanford.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Complaints-To: usenet@sea.gmane.org X-Gmane-NNTP-Posting-Host: c-134-88-108.hh.dial.de.ignite.net X-Mailer: Mozilla 3.0 (OS/2; U) Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Content-Transfer-Encoding: 7bit Russ Allbery wrote: > I dislike saying that servers MUST accept a header Yes, we have some seriouss case of "Deja-Vu" on this list, I'm not immune, sorry. What I really wanted to say, there's a difference between the "Lines:" and other obsolete header fields in 3.3. "Lines:" is deprecated _now_ by the future Usefor RfC. It was optional and unreliable in s-o-1036. Some other headers mentioned in 3.3 are something I've never seen or heard of: Relay-Version, Posting-Version, Date-Received. s-o-1036 says "thoroughly obsolete" about these three header fields listed in its appendix A.3, that was 1994. Finally there are some headers introduced by s-o-1036 that didn't make it (unlike Supersedes): See-Also, Article-Names, Article-Replacing, and Also-Control. Strange, my copy of s-o-1036 (Jan 17 1994) says "Article-Replacing", Usefor-03 says "Article-Updates". Nothing of this is critical, but Also-Control could be a special case like Subject: cmsg cancel. It would be really easier to build on s-o-1036, pro forma. Even if the future Usefor RfC is published only one second after s-o-1036 it could eliminate some weird logical problems. Bye, Frank From owner-ietf-usefor@mail.imc.org Thu Apr 28 12:15:45 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id MAA00943 for ; Thu, 28 Apr 2005 12:15:44 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3SGEfqo058318 for ; Thu, 28 Apr 2005 09:14:41 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3SGEfpR058317 for ietf-usefor-skb; Thu, 28 Apr 2005 09:14:41 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from lon-mail-1.gradwell.net (lon-mail-1.gradwell.net [193.111.201.125]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3SGEdsm058301 for ; Thu, 28 Apr 2005 09:14:40 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from host81-144-74-47.midband.mdip.bt.net ([81.144.74.47]) by lon-mail-1.gradwell.net with esmtp (Gradwell gwh-smtpd 1.181) id 42710bee.165c1.5d for ietf-usefor@imc.org; Thu, 28 Apr 2005 17:14:38 +0100 (envelope-sender ) Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3SGCMJ16999 for ietf-usefor@imc.org; Thu, 28 Apr 2005 17:12:22 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20769 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: Suggested References texts Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: <426E6E66.2070109@epix.net> <426FCA98.3070307@epix.net> Date: Thu, 28 Apr 2005 13:51:26 GMT Lines: 68 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In <426FCA98.3070307@epix.net> "Forrest J. Cavalier III" writes: >Charles Lindsey wrote: > > Inevitably, this changes the information that a reading agent can deduce > > when it sees a References header. It can no longer deduce "this is a > > followup". It can only deduce "the poster wants you to treat this like a > > followup" (by threading it, or by letting you retrieve the precursors by > > clicking on them, or whatever other bells and whistles you provide). >Do you agree we should be concentrating on how the reading agent can >process articles? I think you do. And the text you suggested is flawed >because it declares nothing of use for reading agents. Then what else can be written? If References headers are to be allowed for "pseudo-followups" (and there seems to be agreement on that), then what the reading agent can or cannot deduce is fixed by the structure that we have allowed. Playing around with the definitions cannot alter that. > >> | The References header is used in followups, and other articles with > >> | related precursors (e.g. multipart FAQs and the fragments of a > >> | message/partial), to facilitate the display or retrieval, by reading > >> | and other agents, of threads of related articles. ... >What does "facilitate" mean? What does "other articles" mean? Since "e.g." >is not "i.e.", what other kinds of articles can have References? What stops >someone from using References for other kinds of facilitation of display or >retrieval? (Style sheets!) "Facilitate" means that they can use the information (whose meaning is defined by the semantics further down) in any way they choose in order to provide threading, retrieval, etc. In fact, there are well-known algorithms for threading that work pretty well, but implementors are free to experiment with supposedly better ones, and to some extent different users prefer different presentations. Yes, why not "style sheets"? The important thing is that they need to rely on the References header being present, and being constructed according to the given semantics. >That text is useless. It doesn't even guide posters, much less readers. No, the guidance comes in the semantics paragraph further down, which is supposed to tell them what they can rely on. >If people really want to allow FAQs and multi-part to get threaded, >I wonder if the USEFOR text should be the following single sentence, (which cleverly >does not mention followup or FAQ or multi-part.) > A References header indicates the previously posted context in which > the article was created. Yes, alternative texts to the one I suggested can be considered. >(That "the" means that you can only have at most one context per article.) But people are going to try and followup to more than one article, whether we like it or not. Someday there may even be an agreed way to do it. Hence my semantics was as minimal and future proof as I could make it. -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 From owner-ietf-usefor@mail.imc.org Thu Apr 28 12:15:46 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id MAA00944 for ; Thu, 28 Apr 2005 12:15:44 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3SGEbXJ058285 for ; Thu, 28 Apr 2005 09:14:37 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3SGEbYb058284 for ietf-usefor-skb; Thu, 28 Apr 2005 09:14:37 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from lon-mail-1.gradwell.net (lon-mail-1.gradwell.net [193.111.201.125]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3SGEagN058261 for ; Thu, 28 Apr 2005 09:14:37 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from host81-144-74-47.midband.mdip.bt.net ([81.144.74.47]) by lon-mail-1.gradwell.net with esmtp (Gradwell gwh-smtpd 1.181) id 42710beb.165c1.5a for ietf-usefor@imc.org; Thu, 28 Apr 2005 17:14:35 +0100 (envelope-sender ) Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3SGCKQ16989 for ietf-usefor@imc.org; Thu, 28 Apr 2005 17:12:20 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20767 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: News headers (was References) Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: Date: Thu, 28 Apr 2005 12:06:46 GMT Lines: 58 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In John Stanley writes: >RFC2822 does not say "email replies". Since USEFOR defers to RFC2822, we >have chosen to expand any "email" limitation of RFC2822 into USENET. If we >are so pedantic that we say "RFC2822 is EMAIL only" even in the face of >explicit references to RFC2822 by draft news standards, then exactly what >headers DOES news have and how are they defined? Many of them are defined >as "as found in RFC2822". But since those are EMAIL headers, then they >don't apply to news. Sigh. You do have a point there regarding which headers are officially part of "News". RFC 2822 defines "Optional fields", so that the following header is perfectly legal in any email: Foobar: bar foo However, the clear intent of that field was to allow other standards to define further email headers, such as Content-type: text/plain and have them automatically be acceptable to RFC 2822. It was also intended to allow any header beginning with "X-". But those intents were not clearly stated, which is a pity IMO. The situation was somewhat restored by RFC 3864, which establishes an IANA registry for email (and also news and HTTP) headers. What I would like to see in USEFOR is a clear statement that the headers it explicitly mentions (From, Subject, Newsgroups, etc) are "official Netnews headers" and that ones it does not mention are "passengers" which are generally to be ignored by news software, but which should nevertheless be passed on untouched by transports. Not that I am proposing to define "passenger" as a technical term for the final document. Examples of such "passengers" are the Received and In-Reply-To headers, which are not, and never have been, a part of Netnews, but which do show up occasionally, especially in articles gatewayed from email. Indeed, I would go further, and say that any header appearing in a News article SHOULD either be one blessed by the RFC 3864 registry, or else should begin with "X-". In fact, the old draft-13 contained wording of this nature, but it seems not to have made its way into USEFOR yet. -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 From owner-ietf-usefor@mail.imc.org Thu Apr 28 12:15:47 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id MAA00981 for ; Thu, 28 Apr 2005 12:15:47 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3SGEZ4E058258 for ; Thu, 28 Apr 2005 09:14:35 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3SGEZIi058257 for ietf-usefor-skb; Thu, 28 Apr 2005 09:14:35 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from lon-mail-1.gradwell.net (lon-mail-1.gradwell.net [193.111.201.125]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3SGEYko058235 for ; Thu, 28 Apr 2005 09:14:35 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from host81-144-74-47.midband.mdip.bt.net ([81.144.74.47]) by lon-mail-1.gradwell.net with esmtp (Gradwell gwh-smtpd 1.181) id 42710be9.165c1.58 for ietf-usefor@imc.org; Thu, 28 Apr 2005 17:14:33 +0100 (envelope-sender ) Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3SGCS817033 for ietf-usefor@imc.org; Thu, 28 Apr 2005 17:12:28 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20772 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: Fixed Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: <4255153E.3080500@isode.com> <42594879.3B92@xyzzy.claranet.de> <425A8A63.268A@xyzzy.claranet.de> <425C607E.1B5A@xyzzy.claranet.de> <425D8031.D57@xyzzy.claranet.de> <426BB63D.3060506@isode.com> <426F6AC8.5045@xyzzy.claranet.de> Date: Thu, 28 Apr 2005 15:18:48 GMT Lines: 27 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In <426F6AC8.5045@xyzzy.claranet.de> Frank Ellermann writes: >Alexey Melnikov wrote: >>> the general principle that we do not change anything in RFC >>> 2822 that we do not need to change. >> I concur with your preference. >I don't. But you asked for someone to come along and provide a casting vote to settle the difference between you (wanting @) and me (wanting to retain and ). Now you have it, and it puts you in a minority (even if it does not amount to a ruling from the Chair). You are the only person wanting @. -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 From owner-ietf-usefor@mail.imc.org Thu Apr 28 12:15:49 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id MAA01002 for ; Thu, 28 Apr 2005 12:15:48 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3SGEh6x058335 for ; Thu, 28 Apr 2005 09:14:43 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3SGEhL4058334 for ietf-usefor-skb; Thu, 28 Apr 2005 09:14:43 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from lon-mail-1.gradwell.net (lon-mail-1.gradwell.net [193.111.201.125]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3SGEgum058327 for ; Thu, 28 Apr 2005 09:14:43 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from host81-144-74-47.midband.mdip.bt.net ([81.144.74.47]) by lon-mail-1.gradwell.net with esmtp (Gradwell gwh-smtpd 1.181) id 42710bf1.165c1.5f for ietf-usefor@imc.org; Thu, 28 Apr 2005 17:14:41 +0100 (envelope-sender ) Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3SGCRv17028 for ietf-usefor@imc.org; Thu, 28 Apr 2005 17:12:27 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20771 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: Again broken (was: Fixed) Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: <4255153E.3080500@isode.com> <42594879.3B92@xyzzy.claranet.de> <425A8A63.268A@xyzzy.claranet.de> <425C607E.1B5A@xyzzy.claranet.de> <425D8031.D57@xyzzy.claranet.de> <426BB63D.3060506@isode.com> <426F6107.666C@xyzzy.claranet.de> Date: Thu, 28 Apr 2005 15:03:19 GMT Lines: 155 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In <426F6107.666C@xyzzy.claranet.de> Frank Ellermann writes: >Charles Lindsey wrote: >~~~ begin quoted ABNF ~~~ >message-id = "Message-ID:" SP [FWS] msg-id [FWS] CRLF ........... >~~~ end of quoted sytax ~~~ >Is that as you want it ? Yes. Do you see any technical error in it? > I'm not sure about Alexey's external >expert for 2822bis, so I think the most simple solution to get >an expert review for the two proposals is to post them on the >rfc-822 mailing list. I don't think there is current activity on 2822bis. As soon as there is, I will be trying to persuade them into adopting our syntax, or preferebaly some less bizarre subset of it. Their present msg-id definition should just go into their obs-syntax. But that is a battle for another day and another place. >> The msg-id-core MUST NOT be more than 250 octets in length. >s/msg-id-core/msg-id/g Oops! >>| Also note that this updated ABNF applies wherever is used, >>| including the References header discussed in Section 3.2.1 and the >>| Supersedes header discussed in Section 3.2.5. >IMHO that's obvious and needs no "note", It's in the present text. I would have no worries if Ken decided to take it out. >BTW, in Control: you have the same syntax problem as in the >References:, you separate the arguments by an _optional_ [FWS]: >Control: cancel Hold on! You aren't allowed to cancel more than one article at a time. Yes, I know CNews does it, but this WG decided long ago not to allow it, since most other serving agents don't accept it. >Control: newgroupthis.is.new >Control: rmgroupthis.is.old But yes, you have spotted a real bug there (but fortunately the error is not repeated in the syntax of the actual s in USEPRO). But whilst we are here, there are two other errors in the Control syntax. A. The FWS should be CFWS in all places. The WG agreed early on that comments should be acceptable in all places allowed by RFC 2822, but on a "MUST accept, do not generate yet" basis. Later on, we agreed that they should be disallowed wherevoer there was a severe performance penalty, so they were removed from Newsgroups, Message-ID, Path, Followup-To and Distribution (these were intended to be permanent exclusions, not just until "yet" happened). But it was never agreed for Control, because Control headers are so rare that performance issues can be neglected. B. s/control-message/control-command/ This is to accord with a change I made in USEPRO. The problem is that the term "control message" is already defined to mean the whole article containing a Control header, and to have the syntactic element meaning something different seemed a bad idea. So I believe the complete syntax for Control headers should be: control = "Control:" SP [CFWS] control-command [CFWS] CRLF control-command = verb *( CFWS argument ) verb = token argument = value On top of that, a "do not generate comments yet" wording would be in order, as for the References header, but I would prefer a generic statement to that effect covering all headers inherited from RFC 1036. >... "value" is apparently not >defined in usefor-03 or 2822. "token" is apparently also not >defined in usefor-03 or 2822. , and are all defined in RFC 2045, and used at various places in USEFOR. This needs to be stated somewhere, but there is more to it than that because RFC 2045, being RFC 822 based, does not tell you where to put the [CFWS]s in them, so we need some wording to cover that. Also, in Injection-Info, where it says post-host-param = "posting-host" "=" host-value it actually means post-host-param = [CFWS] "posting-host" [CFWS] "=" [CFWS] host-value [CFWS] and similarly for the other <*-param>s. So that needs to be covered by explicit syntax or suitable verbiage. >>| NOTE: It in RECOMMENDED in [RFC 2822] that, for ensuring >>| global uniqueness, the be some domain identifier >It's unnecessary to quote that RECOMMENDATION, least of all the >"some domain identifier". Out of its original very convoluted >RfC 2822 context this quote is in fact _worse_ than RfC 2822. There is a problem in RFC 2822 (which you yourself spotted) in that there is wording in 3.4.1 to the effect that a is meant to be an IP address (whether IPv4 or IPv6 or any other IPvxx), in spite of a syntax apparently allowing much more, but they omitted to say the same thing about a . The text I wrote was intended to plug this whole, but it was only a NOTE, so had no normative effect. If you still want this hole plugged, then please propose an alternative text. >IIRC we had a "pseudo-consensus" (= you and me) about domain- >literals in addresses, Message-IDs, and Path-identifiers, i.e. >say something like STD 11 without shouting: >| Note: THE USE OF DOMAIN-LITERALS IS STRONGLY DISCOURAGED. Yes, but there is more to say than that, because even a is discouraged (does RFC 2822 still say that?) you want even more discouragement for other than IP addresses inside them, and you still need to extend it to s. -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 From owner-ietf-usefor@mail.imc.org Thu Apr 28 12:15:51 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id MAA01020 for ; Thu, 28 Apr 2005 12:15:50 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3SGEbtH058276 for ; Thu, 28 Apr 2005 09:14:37 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3SGEbdn058275 for ietf-usefor-skb; Thu, 28 Apr 2005 09:14:37 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from lon-mail-1.gradwell.net (lon-mail-1.gradwell.net [193.111.201.125]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3SGEZPf058248 for ; Thu, 28 Apr 2005 09:14:36 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from host81-144-74-47.midband.mdip.bt.net ([81.144.74.47]) by lon-mail-1.gradwell.net with esmtp (Gradwell gwh-smtpd 1.181) id 42710bea.165c1.59 for ietf-usefor@imc.org; Thu, 28 Apr 2005 17:14:34 +0100 (envelope-sender ) Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3SGCTL17037 for ietf-usefor@imc.org; Thu, 28 Apr 2005 17:12:29 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20773 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: Differences between RFC 2822 and Usefor (fwd) Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: Date: Thu, 28 Apr 2005 15:36:33 GMT Lines: 51 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In Richard Clayton writes: >So it seems to me that this is a genuine inter-operability issue -- not >at the protocol levels : the articles get to the far end just fine and >the servers will serve them perfectly well. The issue occurs at the >level at which the readers view the newsgroups. For that reason I think >that a standardised way of expressing which pile an article belongs in >is of very significant value. Yes at present some software is a bit >flaky, but that doesn't mean we should fail to say what should be done. I agree entirely. RFC 2119 is very badly written, and is wide open to those who claim that "interoperability" means only that "it must be delivered correctly at the far end". The WG has long striven to use "MUST" wording for cases where "it does not behave as it is intended to behave if you don't say 'MUST'". Anyway, to deal with this present case, you will find some wording in USEPRO 7.7 entitled "Duties of a Reading Agent" where it says, in effect, that the agent MAY take note of the References, Subject and Date headers in order to display list of articles sensibly. This makes such displaying an optional part of the protocol, which in turn means that reading agents are entitled to rely upon the presence and correctness of those headers. Yes, it is pretty meaningless, and it is only there to satisfy the pedants who object to the use of MUST and SHOULD when dealing with the behaviour of followup agents. But if that is the price we have to pay to get this wording in, .......... >sounds like a cock-up :( It WAS a cock-up. >so best fix it then! ASAP!! Exactly. >Signature by unknown keyid: 0xD8D683E2 Eh? I though I knew all your public keys. -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 From owner-ietf-usefor@mail.imc.org Thu Apr 28 12:15:53 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id MAA01039 for ; Thu, 28 Apr 2005 12:15:53 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3SGEenR058308 for ; Thu, 28 Apr 2005 09:14:40 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3SGEePO058307 for ietf-usefor-skb; Thu, 28 Apr 2005 09:14:40 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from lon-mail-1.gradwell.net (lon-mail-1.gradwell.net [193.111.201.125]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3SGEcLf058290 for ; Thu, 28 Apr 2005 09:14:39 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from host81-144-74-47.midband.mdip.bt.net ([81.144.74.47]) by lon-mail-1.gradwell.net with esmtp (Gradwell gwh-smtpd 1.181) id 42710bed.165c1.5c for ietf-usefor@imc.org; Thu, 28 Apr 2005 17:14:37 +0100 (envelope-sender ) Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3SGCHP16973 for ietf-usefor@imc.org; Thu, 28 Apr 2005 17:12:17 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20765 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: References Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: <426E579F.1060809@oceana.com> <426F3B14.54C1@xyzzy.claranet.de> <426F7411.2010903@epix.net> Date: Thu, 28 Apr 2005 11:29:40 GMT Lines: 70 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In <426F7411.2010903@epix.net> "Forrest J. Cavalier III" writes: >I think it is a requirement that a news client be able to determine >"followup" by inspecting the article headers. OK. Then you need to tell us exactly WHY that requirement is needed. As opposed to a requirement "that a news client be able to determine articles that the poster intends to be treated the same way as followups ...". >If it is not a requirement, then the term "followup" is meaningless >and should not appear in any document. The term "followup" has to appear in our standards, if only because we have to explain the purpose of the Followup-To header. And we have to explain the duties of a followup agent. And USEAGE (which is not standards-track) also explains how to contruct attributions and quotations in followups. >In support of this, I point that the References header is for >the use of the READER, not the POSTER. The poster already knows how >all the articles are related. Indeed, but the POSTER gets some say in how the READER gets to view his articles by including (or not including) a References header. >Frank, you allowed that followup => references. If you don't >hold that references => followup, then does the standard allow an >article to reference more than one thread? I think that >allowing References to appear in an article which is not a reply >to a single article (which I thought was the term "followup"), >leads to undecidable actions about threading. What is the READER >agent supposed to do in that situation? Both USEPRO and RFC 2822 state explicitly that they make no provision for followups/replies to multiple precursors, though they do not explicitly forbid them. Some future extension to either or both documents may do so - various suggestions have been made as to how to do it, but there is no agreed method yet. Nevertheless, people will and do do it, and try to construct a References header to suit. In fact, if this is done sensibly, existing threading agents will produce a tolerable behaviour (by which I mean that they will not list an article as available for reading earlier than any of its precursors). In my suggested text for the semantics of the References header, I tried to word it so as to afford the maximum flexibility for such future extensions, whilst always ensuring correct behaviour in the single precursor case. >Is there a term for articles that can have references but are >not followups? If not, why the special term "followup"? What >does the apperance of References header mean, from a READER perspective? I once coined the term "pseudo-followup" just for use in our discussions in this WG (not with any intent of using that term in any document). I was rewarded with the usual multi-page rant from John :-( . -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 From owner-ietf-usefor@mail.imc.org Thu Apr 28 12:15:56 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id MAA01073 for ; Thu, 28 Apr 2005 12:15:55 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3SGEccR058291 for ; Thu, 28 Apr 2005 09:14:38 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3SGEcrw058288 for ietf-usefor-skb; Thu, 28 Apr 2005 09:14:38 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from lon-mail-1.gradwell.net (lon-mail-1.gradwell.net [193.111.201.125]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3SGEYdL058233 for ; Thu, 28 Apr 2005 09:14:34 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from host81-144-74-47.midband.mdip.bt.net ([81.144.74.47]) by lon-mail-1.gradwell.net with esmtp (Gradwell gwh-smtpd 1.181) id 42710be9.165c1.57 for ietf-usefor@imc.org; Thu, 28 Apr 2005 17:14:33 +0100 (envelope-sender ) Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3SGCI116983 for ietf-usefor@imc.org; Thu, 28 Apr 2005 17:12:18 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20766 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: References Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: <87fyxcksel.fsf@windlord.stanford.edu> <426FDD09.5030102@oceana.com> <426FE5EB.5030302@isode.com> <426FE83E.3040503@isode.com> Date: Thu, 28 Apr 2005 11:40:18 GMT Lines: 28 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In <426FE83E.3040503@isode.com> Alexey Melnikov writes: >Alexey Melnikov wrote: >> I do, but the devil is in details: what exactly should be said and where. >The USEFOR document can say that certain types of messages MAY require >the References header. But I just think this should be obvious. I think we say the minimum necessary to prevent other readers from misinterpreting what the document says in the way that John has been misinterpreting it. A definition of "followup" that clearly distinguishes them from RFC 2822 "replies" is a good start (and I think we already agreed to that). Also, if we intend that References may sometimes be used with non-followups, it would be useful to say so. -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 From owner-ietf-usefor@mail.imc.org Thu Apr 28 12:15:59 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id MAA01091 for ; Thu, 28 Apr 2005 12:15:58 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3SGEZgD058252 for ; Thu, 28 Apr 2005 09:14:35 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3SGEZCe058249 for ietf-usefor-skb; Thu, 28 Apr 2005 09:14:35 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from lon-mail-1.gradwell.net (lon-mail-1.gradwell.net [193.111.201.125]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3SGEYAk058232 for ; Thu, 28 Apr 2005 09:14:34 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from host81-144-74-47.midband.mdip.bt.net ([81.144.74.47]) by lon-mail-1.gradwell.net with esmtp (Gradwell gwh-smtpd 1.181) id 42710be8.165c1.56 for ietf-usefor@imc.org; Thu, 28 Apr 2005 17:14:32 +0100 (envelope-sender ) Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3SGCUs17041 for ietf-usefor@imc.org; Thu, 28 Apr 2005 17:12:31 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20774 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: Suggested References texts Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: <87br80ks7i.fsf@windlord.stanford.edu> Date: Thu, 28 Apr 2005 15:42:13 GMT Lines: 29 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In <87br80ks7i.fsf@windlord.stanford.edu> Russ Allbery writes: >Either we define followups by poster intent and some articles with >Reference headers aren't followups, or we define followups by the presence >of a Reference header and we can't make a strong statement about intent. >We can't have it both ways; Reference headers are used in practice for >things other than followups, and I don't expect that practice to change >(nor do I think we really *want* it to change). Exactly. But still John believes that the real world can be changed by smoke and mirrors. I have explained it to him. Seth has explained it to him. Shmuel has explained it to him. And now you have explained it to him. But he still does not see it. I should not waste any more time on it. -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 From owner-ietf-usefor@mail.imc.org Thu Apr 28 12:16:01 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id MAA01113 for ; Thu, 28 Apr 2005 12:16:01 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3SGEc1j058298 for ; Thu, 28 Apr 2005 09:14:38 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3SGEchC058296 for ietf-usefor-skb; Thu, 28 Apr 2005 09:14:38 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from lon-mail-1.gradwell.net (lon-mail-1.gradwell.net [193.111.201.125]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3SGEbFF058278 for ; Thu, 28 Apr 2005 09:14:38 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from host81-144-74-47.midband.mdip.bt.net ([81.144.74.47]) by lon-mail-1.gradwell.net with esmtp (Gradwell gwh-smtpd 1.181) id 42710bec.165c1.5b for ietf-usefor@imc.org; Thu, 28 Apr 2005 17:14:36 +0100 (envelope-sender ) Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3SGCPU17017 for ietf-usefor@imc.org; Thu, 28 Apr 2005 17:12:25 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20770 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: Obsolete vs. obsolescent Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: <200504111212.j3BCChM14516@clerew.man.ac.uk> <87psx1gnjc.fsf@windlord.stanford.edu> <87fyxvx1m6.fsf@windlord.stanford.edu> <87mzs2a1hu.fsf@windlord.stanford.edu> <87wtqxui6s.fsf@windlord.stanford.edu> <87d5so2bpk.fsf@windlord.stanford.edu> <42697BB6.6074@xyzzy.claranet.de> <426BC455.2040108@isode.com> <426E4B45.7010205@oceana.com> <426F4F7C.7385@xyzzy.claranet.de> <87k6moksps.fsf@windlord.stanford.edu> Date: Thu, 28 Apr 2005 14:09:23 GMT Lines: 55 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In <87k6moksps.fsf@windlord.stanford.edu> Russ Allbery writes: >Frank Ellermann writes: >> IMHO all that should be done is to update this statement: >> | As a result, this header is to be regarded as obsolete, and >> | it will likely be removed entirely in a future version of >> | this standard. >> As a result, this header is to be regarded as obsolescent. >> Servers and UAs MUST accept but SHOULD ignore it, and SHOULD >> NOT generate it. >I dislike saying that servers MUST accept a header, since servers are not >required to accept *any* article. I would also leave in the bit about >removing it in a future version of the standard. I agree (though there is a general caveat in USEPRO that the right of any server to drop any article on the floor overrides any "MUST" requirement). But I am happy with the s/obsolete/obsolescent/. >> Subject: cmsg is not mentioned anywhere in usefor-03, that's a >> serious bug. It's mentioned in usepro-03, but it only says >> that this is now obsolete. That's of course FUBAR, it's a >> severe security violation. We cannot jump from RfC 1036 to a >> completely different procedure at date X. AFAIK RfCs must not >> have a flag day. >I think USEFOR does need an explicit mention that subjects starting with >cmsg should not be considered control messages, to document the change >from RFC 1036. No, I think it is a protocol issue, and the mention in USEPRO is sufficient. Any I do not share Frank's worries about a flag day. That usage is already ignored by most servers, so nobody should be relying on its working anymore. If the few remaining servers that honour it suddenly become non-compliant the day our RFC gets published, then so be it. The sky will not immediately fall in, but anyone continuing to use or accept the practice will be severaly sat upon. In any case, I have not seen or heard of anybody actually trying to use the feature in earnest for many years. -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 From owner-ietf-usefor@mail.imc.org Thu Apr 28 13:02:51 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id MAA01040 for ; Thu, 28 Apr 2005 12:15:53 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3SGEgv0058326 for ; Thu, 28 Apr 2005 09:14:42 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3SGEgFV058325 for ietf-usefor-skb; Thu, 28 Apr 2005 09:14:42 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from lon-mail-1.gradwell.net (lon-mail-1.gradwell.net [193.111.201.125]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3SGEfii058315 for ; Thu, 28 Apr 2005 09:14:41 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from host81-144-74-47.midband.mdip.bt.net ([81.144.74.47]) by lon-mail-1.gradwell.net with esmtp (Gradwell gwh-smtpd 1.181) id 42710bef.165c1.5e for ietf-usefor@imc.org; Thu, 28 Apr 2005 17:14:39 +0100 (envelope-sender ) Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3SGCLE16994 for ietf-usefor@imc.org; Thu, 28 Apr 2005 17:12:21 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20768 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: Suggested References texts Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: <426F443C.6561@xyzzy.claranet.de> Date: Thu, 28 Apr 2005 13:30:42 GMT Lines: 110 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In <426F443C.6561@xyzzy.claranet.de> Frank Ellermann writes: >Charles Lindsey wrote: >> A "followup" is an article containing a response to the >> contents of an earlier article. It will always include a >> References header pointing to that earlier article and any >> other "precursors". >Strong NAK to any change of <426E4FC9.7020903@oceana.com> >We know that References do _not_ always include 'any other >"precursors"', in fact there's a complete chapter explaining >how to trim the References. OK. Consider the word "any" struck. >> An article is a "precursor" of some later article which is a >> followup to it >> or which is otherwise intended to be grouped with it for >> purposes of display (e.g. as a part of a multipart posting >> such as a FAQ). >Strong NAK, ... >A "precursor" is simply the opposite of a "followup", so if >A is the precursor of B, then B is a followup of A. There's >no case where A is the precursor of B, but B is no followup >of A, that would be madness. Whilst I do not want the term "followup" to include those "pseudo-followups" (if I can use that term without precipitating another rant from John :-) ), I defined "precursor" to include the "pseudo-precursors" because it simplified the wording of the semantics further down. If you want me to limit "precursor" as you suggest, then please suggest alternative wording for those semantics. The term "followup" is widely used within the Usenet community, and it would be artificial, misleading and unwise to define it in any other way. But "precursor" is not so widely used and would not mislead in the same way. >> Within USEFOR: >>| references = "References:" SP [CFWS] 1*(msg-id [CFWS]) CRLF Oops! That is still wrong. > references = "References:" SP msg-id-list CRLF > msg-id-list = [CFWS] msg-id *( CFWS msg-id ) [CFWS] or, following the style in draft-13, references = "References:" SP [CFWS] msg-id 1*(CFWS msg-id ) [CFWS] CRLF >That's the simple variant where a comment is enough to separate >the Message-IDs as in (c) but adjacent Message-IDs as >in aren't allowed. Agreed. Those s are "MUST accept, don't generate yet", so the issue does not arise until "yet" happens, by which time most agents will probably be able to manage without even a WSP. >Strong NAK to anything below this line. So you want me to omit | The list is composed of message identifiers of precursors of the | current article, sorted so that no article precedes any of its own | precursors. It SHOULD include both the earliest and the immediate | precursors on the current article, even if some of the intermediate | ones are omitted. A given message identifier MUST NOT appear more | than once. But that is the semantics (meaning) of the header, which is a proper matter for USEFOR. It tells reading agents just what they can deduce from the header, and it guides those who generate References for "pseudo-followups". RFC 2822 and our earlier ARTICLE drafts got away by just providing an algorithm for generating this header, but that is not enough if the header is not restricted to followups/replies. > [Usepro] >> If the resulting References header is excessively long, it >> MAY be trimmed, but the first and the last two message >> identifiers MUST NOT be removed. >I'd go for a SHOULD here, because some news servers and tools >really hate it if the References are too long, I've seen cases >where the References were truncated I think we agreed earlier that the actual trimming details belonged in USEAGE. But please see my remark in the latest USEAGE where I would like to keep the whole header within 998 octets (rather than 21 entries), because some followup/reply agents unfold the old references header and then conveniently "forget" to refold the new. -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 From owner-ietf-usefor@mail.imc.org Thu Apr 28 13:07:30 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id NAA05087 for ; Thu, 28 Apr 2005 13:07:30 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3SH67Zu065540 for ; Thu, 28 Apr 2005 10:06:07 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3SH67mv065539 for ietf-usefor-skb; Thu, 28 Apr 2005 10:06:07 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtp3.Stanford.EDU (smtp3.Stanford.EDU [171.67.16.138]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3SH650w065531 for ; Thu, 28 Apr 2005 10:06:06 -0700 (PDT) (envelope-from rra@stanford.edu) Received: from windlord.stanford.edu (windlord.Stanford.EDU [171.64.19.147]) by smtp3.Stanford.EDU (8.12.11/8.12.11) with SMTP id j3SH65MB008207 for ; Thu, 28 Apr 2005 10:06:05 -0700 Received: (qmail 31887 invoked by uid 1000); 28 Apr 2005 17:06:04 -0000 To: ietf-usefor@imc.org Subject: Re: Obsolete vs. obsolescent In-Reply-To: (Charles Lindsey's message of "Thu, 28 Apr 2005 14:09:23 GMT") References: <200504111212.j3BCChM14516@clerew.man.ac.uk> <87psx1gnjc.fsf@windlord.stanford.edu> <87fyxvx1m6.fsf@windlord.stanford.edu> <87mzs2a1hu.fsf@windlord.stanford.edu> <87wtqxui6s.fsf@windlord.stanford.edu> <87d5so2bpk.fsf@windlord.stanford.edu> <42697BB6.6074@xyzzy.claranet.de> <426BC455.2040108@isode.com> <426E4B45.7010205@oceana.com> <426F4F7C.7385@xyzzy.claranet.de> <87k6moksps.fsf@windlord.stanford.edu> From: Russ Allbery Organization: The Eyrie Date: Thu, 28 Apr 2005 10:06:04 -0700 Message-ID: <87br7yg7b7.fsf@windlord.stanford.edu> User-Agent: Gnus/5.1007 (Gnus v5.10.7) XEmacs/21.4 (Jumbo Shrimp, linux) MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Charles Lindsey writes: > Russ Allbery writes: >> I think USEFOR does need an explicit mention that subjects starting >> with cmsg should not be considered control messages, to document the >> change from RFC 1036. > No, I think it is a protocol issue, and the mention in USEPRO is > sufficient. The transformation of a message from a regular news message to a control message because of the presence of five characters at the beginning of the Subject line sounds like a syntax and semantics issue to me, even if we're just warning that it used to be interpreted that way. cmsg changes the semantic meaning of the message entirely, if honored. I don't feel too strongly about it, though. -- Russ Allbery (rra@stanford.edu) From owner-ietf-usefor@mail.imc.org Thu Apr 28 19:18:17 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id TAA06845 for ; Thu, 28 Apr 2005 19:18:17 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3SNG3mF027787 for ; Thu, 28 Apr 2005 16:16:03 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3SNG3kt027786 for ietf-usefor-skb; Thu, 28 Apr 2005 16:16:03 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from ciao.gmane.org (main.gmane.org [80.91.229.2]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3SNG0Du027780 for ; Thu, 28 Apr 2005 16:16:01 -0700 (PDT) (envelope-from usenet-format@gmane.org) Received: from list by ciao.gmane.org with local (Exim 4.43) id 1DRI99-0006R6-A9 for ietf-usefor@imc.org; Fri, 29 Apr 2005 01:10:15 +0200 Received: from 62.80.58.28 ([62.80.58.28]) by main.gmane.org with esmtp (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Fri, 29 Apr 2005 01:10:15 +0200 Received: from nobody by 62.80.58.28 with local (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Fri, 29 Apr 2005 01:10:15 +0200 X-Injected-Via-Gmane: http://gmane.org/ To: ietf-usefor@imc.org From: Frank Ellermann Subject: Re: Suggested References texts Date: Fri, 29 Apr 2005 01:05:14 +0200 Organization: Lines: 85 Message-ID: <42716C2A.4F53@xyzzy.claranet.de> References: <426F443C.6561@xyzzy.claranet.de> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Complaints-To: usenet@sea.gmane.org X-Gmane-NNTP-Posting-Host: 62.80.58.28 X-Mailer: Mozilla 3.0 (OS/2; U) Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Content-Transfer-Encoding: 7bit Charles Lindsey wrote: > If you want me to limit "precursor" as you suggest, then > please suggest alternative wording for those semantics. I just don't want any additional wording in usefor-04, it's one of these cases where each word is too much, unless you plan to integrate the 2004 discussions with the complete history of this header field in RfCs, news, mail, and UAs as told by $number_of_USEFOR_participants_since_99. > references = "References:" SP msg-id-list CRLF >> msg-id-list = [CFWS] msg-id *( CFWS msg-id ) [CFWS] > or, following the style in draft-13, > references = "References:" SP [CFWS] msg-id 1*(CFWS msg-id ) > [CFWS] CRLF ACK, one production less, OTOH a rather long rule. I thought that you could use the msg-id-list also in Control: cancel or Supersedes:, but that's not the case, our Supersedes: has only one msg-id, and Control: has no CFWS. > So you want me to omit [... some text about precursors in usefor-04 ...] Yes, that's what I didn't want because it's essentially the same as in 2822, but now that you ask I'm not so sure about: >| A given message identifier MUST NOT appear more than once. Is that a problem anywhere, bad enough for a MUST NOT ? It's obviously stupid, but is it potentially harmful ? > that is the semantics (meaning) of the header, which is a > proper matter for USEFOR. Not if it's essentially the same as in 2822. Or that's what I thought the plan is: Anything in RfC 2822 (excl. obs-stuff) that's not explicity changed in usefor-03/04 is implicitly a part of usefor-03/04. That's why we don't explain mailbox, addr-spec, From:, Sender:, etc. We only explain what's different, e.g. the msg-id syntax or something with the Date: timezone. And of course all header fields found only in news, among others Followup-To:. Unless you start again with message/partial and other oddities I don't see where the semantics of "normal" References: is so different from RfC 2822 that it needs an explanation. Sure, we don't have In-Reply-To, that's a minor difference, UAs aren't expected to derive missing References: from In-Reply-To: But that's an issue for usepro, not usefor, like trimming etc. > It tells reading agents just what they can deduce from the > header If developers of reading agents need more info they can find it where deelopers of followup-agents find it, in usepro. What's so special for news reading-agents, in relation to RfC 2822 ? Or do you want some "security considerations" about References: "Beware, the References in news can be wild and wonderful, and for mail it's the same problem. But we wanted to document it explicitly. so take care." Seriously, I don't get it. > I think we agreed earlier that the actual trimming details > belonged in USEAGE. Quite possible, that's something between you and Alexey, this split still confuses me. But at least I got the idea what's usefor-xx and what not: usefor-xx is only the article format. Not all why-s, just the minimum to describe what's actually transmitted. Form but not function. Plus basic definitions. Bye, Frank From owner-ietf-usefor@mail.imc.org Thu Apr 28 19:37:44 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id TAA08205 for ; Thu, 28 Apr 2005 19:37:43 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3SNape4030708 for ; Thu, 28 Apr 2005 16:36:51 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3SNapH4030707 for ietf-usefor-skb; Thu, 28 Apr 2005 16:36:51 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtp1.Stanford.EDU (smtp1.Stanford.EDU [171.67.16.123]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3SNaoDA030687 for ; Thu, 28 Apr 2005 16:36:50 -0700 (PDT) (envelope-from rra@stanford.edu) Received: from windlord.stanford.edu (windlord.Stanford.EDU [171.64.19.147]) by smtp1.Stanford.EDU (8.12.11/8.12.11) with SMTP id j3SNan6V012490 for ; Thu, 28 Apr 2005 16:36:49 -0700 Received: (qmail 14645 invoked by uid 1000); 28 Apr 2005 23:36:49 -0000 To: ietf-usefor@imc.org Subject: Re: Suggested References texts In-Reply-To: <42716C2A.4F53@xyzzy.claranet.de> (Frank Ellermann's message of "Fri, 29 Apr 2005 01:05:14 +0200") References: <426F443C.6561@xyzzy.claranet.de> <42716C2A.4F53@xyzzy.claranet.de> From: Russ Allbery Organization: The Eyrie Date: Thu, 28 Apr 2005 16:36:49 -0700 Message-ID: <871x8ubhim.fsf@windlord.stanford.edu> User-Agent: Gnus/5.1007 (Gnus v5.10.7) XEmacs/21.4 (Jumbo Shrimp, linux) MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Frank Ellermann writes: > I just don't want any additional wording in usefor-04, it's > one of these cases where each word is too much, unless you > plan to integrate the 2004 discussions with the complete > history of this header field in RfCs, news, mail, and UAs > as told by $number_of_USEFOR_participants_since_99. Not if it's essentially the same as in 2822. Or that's what I > thought the plan is: Anything in RfC 2822 (excl. obs-stuff) > that's not explicity changed in usefor-03/04 is implicitly a > part of usefor-03/04. > That's why we don't explain mailbox, addr-spec, From:, Sender:, > etc. We only explain what's different, e.g. the msg-id syntax > or something with the Date: timezone. And of course all header > fields found only in news, among others Followup-To:. > Unless you start again with message/partial and other oddities > I don't see where the semantics of "normal" References: is so > different from RfC 2822 that it needs an explanation. > Sure, we don't have In-Reply-To, that's a minor difference, UAs > aren't expected to derive missing References: from In-Reply-To: > But that's an issue for usepro, not usefor, like trimming etc. The only thing that I can see mentioning, given this whole discussion, is saying that References is required for followups, although I don't think we should say much more than RFC 1036 does in that area. For the rest, I agree with you wholeheartedly. I *don't* want to give the impression, even in USEFOR, that all the rules in RFC 2822 for replies apply unmodified to followups. Replies and followups *aren't* the same thing, and I think it's worth making that clear. Among other things, correctly constructing a followup requires dealing with fields that aren't in RFC 2822 at all. >> I think we agreed earlier that the actual trimming details >> belonged in USEAGE. If we want to strongly encourage (SHOULD) a particular trimming strategy, that goes into USEPRO as it's part of the protocol. Given the number of bad trimming strategies that people use that cause interoperability problems, I think this is probably a good idea. > Quite possible, that's something between you and Alexey, this > split still confuses me. But at least I got the idea what's > usefor-xx and what not: usefor-xx is only the article format. > Not all why-s, just the minimum to describe what's actually > transmitted. Form but not function. Plus basic definitions. Yup. -- Russ Allbery (rra@stanford.edu) From owner-ietf-usefor@mail.imc.org Thu Apr 28 19:39:39 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id TAA08289 for ; Thu, 28 Apr 2005 19:39:39 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3SNcmhP030975 for ; Thu, 28 Apr 2005 16:38:48 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3SNcmXk030974 for ietf-usefor-skb; Thu, 28 Apr 2005 16:38:48 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from ciao.gmane.org (main.gmane.org [80.91.229.2]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3SNckJn030968 for ; Thu, 28 Apr 2005 16:38:47 -0700 (PDT) (envelope-from usenet-format@gmane.org) Received: from list by ciao.gmane.org with local (Exim 4.43) id 1DRIUy-0008Pe-JA for ietf-usefor@imc.org; Fri, 29 Apr 2005 01:32:48 +0200 Received: from 62.80.58.28 ([62.80.58.28]) by main.gmane.org with esmtp (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Fri, 29 Apr 2005 01:32:48 +0200 Received: from nobody by 62.80.58.28 with local (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Fri, 29 Apr 2005 01:32:48 +0200 X-Injected-Via-Gmane: http://gmane.org/ To: ietf-usefor@imc.org From: Frank Ellermann Subject: Re: Obsolete vs. obsolescent Date: Fri, 29 Apr 2005 01:37:12 +0200 Organization: Lines: 25 Message-ID: <427173A8.63D5@xyzzy.claranet.de> References: <200504111212.j3BCChM14516@clerew.man.ac.uk> <87psx1gnjc.fsf@windlord.stanford.edu> <87fyxvx1m6.fsf@windlord.stanford.edu> <87mzs2a1hu.fsf@windlord.stanford.edu> <87wtqxui6s.fsf@windlord.stanford.edu> <87d5so2bpk.fsf@windlord.stanford.edu> <42697BB6.6074@xyzzy.claranet.de> <426BC455.2040108@isode.com> <426E4B45.7010205@oceana.com> <426F4F7C.7385@xyzzy.claranet.de> <87k6moksps.fsf@windlord.stanford.edu> <87br7yg7b7.fsf@windlord.stanford.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Complaints-To: usenet@sea.gmane.org X-Gmane-NNTP-Posting-Host: 62.80.58.28 X-Mailer: Mozilla 3.0 (OS/2; U) Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Content-Transfer-Encoding: 7bit Russ Allbery wrote: [cmsg] > I don't feel too strongly about it, though. Okay, but it's not nice when it has serious problems with the propagation. An error message "administrative restriction" is also not exactly clear. Or if we'd expect gateways to do something special, or better nothing unless they are forced to bypass an "administrative restriction" we have to say so somewhere. Not in usefor-x, that was probably a bad idea. But the usepro-03 text pretends that there's actually no problem at all with Subject: cmsg. And that's not the case, The attempts to avoid obsolete problems are the problem - not some old servers trying to act on Subject: cmsg, but the many servers trying to protect old servers are a pain. And they won't stop this immediately only because a new RfC wants it. Let's forget Also-Control unless you're sure that it could be a similar problem. Bye, Frank From owner-ietf-usefor@mail.imc.org Thu Apr 28 19:46:58 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id TAA08917 for ; Thu, 28 Apr 2005 19:46:57 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3SNjvvH032588 for ; Thu, 28 Apr 2005 16:45:57 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3SNjvrS032587 for ietf-usefor-skb; Thu, 28 Apr 2005 16:45:57 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtp2.Stanford.EDU (smtp2.Stanford.EDU [171.67.16.125]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3SNjumh032577 for ; Thu, 28 Apr 2005 16:45:56 -0700 (PDT) (envelope-from rra@stanford.edu) Received: from windlord.stanford.edu (windlord.Stanford.EDU [171.64.19.147]) by smtp2.Stanford.EDU (8.12.11/8.12.11) with SMTP id j3SNjtDf014386 for ; Thu, 28 Apr 2005 16:45:56 -0700 Received: (qmail 14800 invoked by uid 1000); 28 Apr 2005 23:45:55 -0000 To: ietf-usefor@imc.org Subject: Re: Obsolete vs. obsolescent In-Reply-To: <427173A8.63D5@xyzzy.claranet.de> (Frank Ellermann's message of "Fri, 29 Apr 2005 01:37:12 +0200") References: <200504111212.j3BCChM14516@clerew.man.ac.uk> <87psx1gnjc.fsf@windlord.stanford.edu> <87fyxvx1m6.fsf@windlord.stanford.edu> <87mzs2a1hu.fsf@windlord.stanford.edu> <87wtqxui6s.fsf@windlord.stanford.edu> <87d5so2bpk.fsf@windlord.stanford.edu> <42697BB6.6074@xyzzy.claranet.de> <426BC455.2040108@isode.com> <426E4B45.7010205@oceana.com> <426F4F7C.7385@xyzzy.claranet.de> <87k6moksps.fsf@windlord.stanford.edu> <87br7yg7b7.fsf@windlord.stanford.edu> <427173A8.63D5@xyzzy.claranet.de> From: Russ Allbery Organization: The Eyrie Date: Thu, 28 Apr 2005 16:45:55 -0700 Message-ID: <87zmvijwi4.fsf@windlord.stanford.edu> User-Agent: Gnus/5.1007 (Gnus v5.10.7) XEmacs/21.4 (Jumbo Shrimp, linux) MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Frank Ellermann writes: > Okay, but it's not nice when it has serious problems with the > propagation. An error message "administrative restriction" > is also not exactly clear. Or if we'd expect gateways to do > something special, or better nothing unless they are forced > to bypass an "administrative restriction" we have to say so > somewhere. Not in usefor-x, that was probably a bad idea. A subject starting with "cmsg " needs to be prohibited in anything that isn't a control message. I don't feel strongly about whether that should be in USEFOR or USEPRO, although to me it feels like a format issue, but it needs to be in there somewhere. It's premature to assume that cmsg no longer receives special handling; we're not done with that transition. We might never be, frankly. Using cmsg in the Subject header of regular posts will cause significant interoperability problems. > And they won't stop this immediately only because a new RfC > wants it. Let's forget Also-Control unless you're sure that > it could be a similar problem. Also-Control simply SHOULD NOT be used, but that's more USEPRO. -- Russ Allbery (rra@stanford.edu) From owner-ietf-usefor@mail.imc.org Thu Apr 28 20:09:15 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id UAA10995 for ; Thu, 28 Apr 2005 20:09:15 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3T07mfq035067 for ; Thu, 28 Apr 2005 17:07:48 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3T07mD2035066 for ietf-usefor-skb; Thu, 28 Apr 2005 17:07:48 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from ciao.gmane.org (main.gmane.org [80.91.229.2]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3T07k1E035052 for ; Thu, 28 Apr 2005 17:07:46 -0700 (PDT) (envelope-from usenet-format@gmane.org) Received: from list by ciao.gmane.org with local (Exim 4.43) id 1DRIxA-0002cb-U1 for ietf-usefor@imc.org; Fri, 29 Apr 2005 02:01:56 +0200 Received: from 62.80.58.28 ([62.80.58.28]) by main.gmane.org with esmtp (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Fri, 29 Apr 2005 02:01:56 +0200 Received: from nobody by 62.80.58.28 with local (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Fri, 29 Apr 2005 02:01:56 +0200 X-Injected-Via-Gmane: http://gmane.org/ To: ietf-usefor@imc.org From: Frank Ellermann Subject: Re: Fixed Date: Fri, 29 Apr 2005 02:01:51 +0200 Organization: Lines: 23 Message-ID: <4271796F.2884@xyzzy.claranet.de> References: <4255153E.3080500@isode.com> <42594879.3B92@xyzzy.claranet.de> <425A8A63.268A@xyzzy.claranet.de> <425C607E.1B5A@xyzzy.claranet.de> <425D8031.D57@xyzzy.claranet.de> <426BB63D.3060506@isode.com> <426F6AC8.5045@xyzzy.claranet.de> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Complaints-To: usenet@sea.gmane.org X-Gmane-NNTP-Posting-Host: 62.80.58.28 X-Mailer: Mozilla 3.0 (OS/2; U) Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Content-Transfer-Encoding: 7bit Charles Lindsey wrote: > But you asked for someone to come along and provide a > casting vote True. I wasn't happy with how Alexey did it, no indication that he understood why we agree on a syntax but disagree on the stupid names of the productions. As chair he's supposed to handle such obscure problems with tact. > You are the only person wanting @. That's not the point, I'm the person who seriously does not like "id-right", I think that your idea of "id-right" is not the "id-domain" in my last proposal. If you like to reopen negotiations, you can have "id-left". The critical part for Usenet is the RHS with a domain. The LHS is then only a problem of *the* domain owner. For your "id-right" that's very unclear. Bye, Frank From owner-ietf-usefor@mail.imc.org Thu Apr 28 21:56:19 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id VAA17742 for ; Thu, 28 Apr 2005 21:56:18 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3T1t6X5050389 for ; Thu, 28 Apr 2005 18:55:06 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3T1t69Y050388 for ietf-usefor-skb; Thu, 28 Apr 2005 18:55:06 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from ciao.gmane.org (main.gmane.org [80.91.229.2]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3T1t4Ir050381 for ; Thu, 28 Apr 2005 18:55:05 -0700 (PDT) (envelope-from usenet-format@gmane.org) Received: from list by ciao.gmane.org with local (Exim 4.43) id 1DRKcZ-0002Jm-1j for ietf-usefor@imc.org; Fri, 29 Apr 2005 03:48:47 +0200 Received: from 62.80.58.28 ([62.80.58.28]) by main.gmane.org with esmtp (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Fri, 29 Apr 2005 03:48:47 +0200 Received: from nobody by 62.80.58.28 with local (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Fri, 29 Apr 2005 03:48:47 +0200 X-Injected-Via-Gmane: http://gmane.org/ To: ietf-usefor@imc.org From: Frank Ellermann Subject: Re: Again broken Date: Fri, 29 Apr 2005 03:49:55 +0200 Organization: Lines: 112 Message-ID: <427192C3.301E@xyzzy.claranet.de> References: <4255153E.3080500@isode.com> <42594879.3B92@xyzzy.claranet.de> <425A8A63.268A@xyzzy.claranet.de> <425C607E.1B5A@xyzzy.claranet.de> <425D8031.D57@xyzzy.claranet.de> <426BB63D.3060506@isode.com> <426F6107.666C@xyzzy.claranet.de> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Complaints-To: usenet@sea.gmane.org X-Gmane-NNTP-Posting-Host: 62.80.58.28 X-Mailer: Mozilla 3.0 (OS/2; U) Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Content-Transfer-Encoding: 7bit Charles Lindsey wrote: >> Is that as you want it ? > Yes. Do you see any technical error in it? No, I just wanted to be sure that I can quote it as is. > I will be trying to persuade them into adopting our > syntax, or preferebaly some less bizarre subset of it. The excluded ">" is bizarre, but allowing it elsewhere would be a pain for gateways. This syntax _is_ the Message-ID for mail and news, bizarre or not. I thought about putting it into an I-D, "updates 2822", ready, problem solved. >> Control: cancel > Hold on! You aren't allowed to cancel more than one > article at a time. Sorry, I wrote it after the very similar References issue. > A. The FWS should be CFWS in all places. I thought the server folks don't want CFWS in their headers, because they don't want to waste time to parse comments. > The WG agreed early on that comments should be acceptable > in all places allowed by RFC 2822 Especially where stupid users do stupid things, but Control ? > Control headers are so rare that performance issues can be > neglected. Okay, if you say so. Convince Russ that it's a good idea. > control = "Control:" SP [CFWS] control-command [CFWS] CRLF Often you have the leading / trailing [CFWS] in the "inner" rule with shorter lines, i.e. control = "Control:" SP control-command CRLF control-command = [CFWS] verb *( CFWS argument ) [CFWS} Of course irrelevant; I only see it when I try to quote it. > On top of that, a "do not generate comments yet" wording > would be in order, as for the References header If you think that comments are stupid, and they are, and only needed for users who trust that the rules are everywhere the same, then using [FWS] and FWS without any additional wording is much clearer than [CFWS] and CFWS with a rule to pray not yet use comments in an "administrative" header like Control. > I would prefer a generic statement to that effect covering > all headers inherited from RFC 1036. That's a trend, you love to say A in syntax, and then to add long qualifications which are near to NOT A. I'd just say NOT A in syntax without qualification, and be done with it. > , and are all defined in RFC 2045 Yes. So far I didn't know that usefor-03 implicitly includes all 2045 terms. I only knew that it includes 2822 minus obs, and of course 2234bis. 2045 is older than 2234. IMHO it's better to copy (and check) what you need, instead of implicitly importing it. Checking: token := 1* Whatever that is, it's no proper ABNF. So why did Keith do it in 3834, Bruce must have seen this, I'm stunned. 2231 doesn't fix it. We shouldn't import this quasi-ABNF, even if Bruce promises to let it pass. > There is a problem in RFC 2822 (which you yourself spotted) A _real_ problem in 2822 ? So far I tended to hallucinate problems, but the worst I'm sure of was an "I don't like it". > If you still want this hole plugged, then please propose an > alternative text. That's simple, just say "id-domain" and "id-literal" _without_ additional wording. Or say "address-literal". Not the unclear "id-right" and "no-fold-literal" stuff, where nobody knows what it's supposed to mean. Or add a note with this effect: Note: id-right is a domain, and no-fold-literal is the special case of a domain-literal (a.k.a. address-literal, see RfC 2821) >>| Note: THE USE OF DOMAIN-LITERALS IS STRONGLY DISCOURAGED. > Yes, but there is more to say than that, because even a > is discouraged (does RFC 2822 still say > that?) you want even more discouragement for other than IP > addresses inside them, and you still need to extend it to > s. More to say ? That statement is perfectly clear, and we have serious technical reasons why we want it. No, AFAIK RfC 2882 does not say so, otherwise we wouldn't need it in usefor-xx. Bye, Frank From owner-ietf-usefor@mail.imc.org Thu Apr 28 22:03:19 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id WAA18260 for ; Thu, 28 Apr 2005 22:03:19 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3T21cd4051310 for ; Thu, 28 Apr 2005 19:01:38 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3T21cep051309 for ietf-usefor-skb; Thu, 28 Apr 2005 19:01:38 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtp1.Stanford.EDU (smtp1.Stanford.EDU [171.67.16.123]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3T21bD6051302 for ; Thu, 28 Apr 2005 19:01:37 -0700 (PDT) (envelope-from rra@stanford.edu) Received: from windlord.stanford.edu (windlord.Stanford.EDU [171.64.19.147]) by smtp1.Stanford.EDU (8.12.11/8.12.11) with SMTP id j3T21aTx016233 for ; Thu, 28 Apr 2005 19:01:36 -0700 Received: (qmail 19657 invoked by uid 1000); 29 Apr 2005 02:01:36 -0000 To: ietf-usefor@imc.org Subject: Re: Again broken In-Reply-To: <427192C3.301E@xyzzy.claranet.de> (Frank Ellermann's message of "Fri, 29 Apr 2005 03:49:55 +0200") References: <4255153E.3080500@isode.com> <42594879.3B92@xyzzy.claranet.de> <425A8A63.268A@xyzzy.claranet.de> <425C607E.1B5A@xyzzy.claranet.de> <425D8031.D57@xyzzy.claranet.de> <426BB63D.3060506@isode.com> <426F6107.666C@xyzzy.claranet.de> <427192C3.301E@xyzzy.claranet.de> From: Russ Allbery Organization: The Eyrie Date: Thu, 28 Apr 2005 19:01:36 -0700 Message-ID: <87sm1apchr.fsf@windlord.stanford.edu> User-Agent: Gnus/5.1007 (Gnus v5.10.7) XEmacs/21.4 (Jumbo Shrimp, linux) MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Frank Ellermann writes: > Charles Lindsey wrote: >> The WG agreed early on that comments should be acceptable >> in all places allowed by RFC 2822 > Especially where stupid users do stupid things, but Control ? >> Control headers are so rare that performance issues can be >> neglected. > Okay, if you say so. Convince Russ that it's a good idea. RFC 2822 doesn't say anything about Control headers, for obvious reasons. What are we gaining by adding comments to the header and breaking existing software? -- Russ Allbery (rra@stanford.edu) From owner-ietf-usefor@mail.imc.org Fri Apr 29 12:15:43 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id MAA10962 for ; Fri, 29 Apr 2005 12:15:42 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3TGECwO028386 for ; Fri, 29 Apr 2005 09:14:12 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3TGECjq028385 for ietf-usefor-skb; Fri, 29 Apr 2005 09:14:12 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from lon-mail-3.gradwell.net (lon-mail-3.gradwell.net [193.111.201.127]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3TGEB7H028374 for ; Fri, 29 Apr 2005 09:14:12 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from host81-144-75-3.midband.mdip.bt.net ([81.144.75.3]) by lon-mail-3.gradwell.net with esmtp (Gradwell gwh-smtpd 1.181) id 42725d52.17ebc.215 for ietf-usefor@imc.org; Fri, 29 Apr 2005 17:14:10 +0100 (envelope-sender ) Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3TGCZc26915 for ietf-usefor@imc.org; Fri, 29 Apr 2005 17:12:35 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20786 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: Suggested References texts Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: <426F443C.6561@xyzzy.claranet.de> <42716C2A.4F53@xyzzy.claranet.de> Date: Fri, 29 Apr 2005 15:14:24 GMT Lines: 100 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In <42716C2A.4F53@xyzzy.claranet.de> Frank Ellermann writes: >I just don't want any additional wording in usefor-04, it's >one of these cases where each word is too much, unless you >plan to integrate the 2004 discussions with the complete >history of this header field in RfCs, news, mail, and UAs >as told by $number_of_USEFOR_participants_since_99. It's like an attempt to define spam with all exceptions and >details, it's hopeless. You really tried. Now let's please >just get away with the bare minimum in usefor-04, for the >technical details usepro-03 should be good enough. But usepro-03 doesn't provide those technical details. It tells you how to construct a followup to an article that is already in front of you (and RFC 2822 does the same for email replies, but is a bit more complicated because it has to allow for cases where References headers were not provided - that "SHOULD" - and it deals with In-Reply-To as well). But suppose you are a multipart FAQ writer trying to use References to refer back to the earlier parts, how do you know what you can do? Or you are writing a threading agent. How do you know what you can rely on in a properly formed References header? If all you are given is the bare syntax (which is what you seem to suggest), then what is to stop you including s of articles that never existed? What is to stop you listing s in the wrong order? Where is the wording to say these things are wrong? Now indeed, most FAQ writers are clueful enough to construct sensible References headers, but in fact they are merely relying on "folklore" - and the whole purpose of standards is to replace folklore by somethng with more precision. Or else they have to "reverse engineer" the algorithm given for followups in USEPRO, and deduce the underlying semantics from that (with the danger of reading more into the algorithm than is strictly justified). Likewise for those who write threading agents. >> So you want me to omit >[... some text about precursors in usefor-04 ...] >Yes, that's what I didn't want because it's essentially the >same as in 2822, but now that you ask I'm not so sure about: >>| A given message identifier MUST NOT appear more than once. >Is that a problem anywhere, bad enough for a MUST NOT ? It's >obviously stupid, but is it potentially harmful ? Yes, that sentence might go. >> that is the semantics (meaning) of the header, which is a >> proper matter for USEFOR. >Not if it's essentially the same as in 2822. Or that's what I >thought the plan is: Anything in RfC 2822 (excl. obs-stuff) >that's not explicity changed in usefor-03/04 is implicitly a >part of usefor-03/04. But RFC 2888 doesn't explain the _meaning_ of the References header. It just tells you how to contruct one when creating a Reply, which is not what we want to use it for. >That's why we don't explain mailbox, addr-spec, From:, Sender:, >etc. Indeed, because RFC 2822 explains all that is needed for those. >Unless you start again with message/partial and other oddities >I don't see where the semantics of "normal" References: is so >different from RfC 2822 that it needs an explanation. Exactly. It is the "other oddities" which necessitate some extra explanation. >> It tells reading agents just what they can deduce from the >> header >If developers of reading agents need more info they can find it >where deelopers of followup-agents find it, in usepro. What's >so special for news reading-agents, in relation to RfC 2822 ? They are going to encounter References headers in articles that are not followups (or replies in the 2822 sense). So there is nothing in USEPRO to help them (except guesswork based on what followup agents do). -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 From owner-ietf-usefor@mail.imc.org Fri Apr 29 12:15:43 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id MAA10964 for ; Fri, 29 Apr 2005 12:15:42 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3TGEBRF028376 for ; Fri, 29 Apr 2005 09:14:11 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3TGEBcb028375 for ietf-usefor-skb; Fri, 29 Apr 2005 09:14:11 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from lon-mail-3.gradwell.net (lon-mail-3.gradwell.net [193.111.201.127]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3TGEAF3028364 for ; Fri, 29 Apr 2005 09:14:10 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from host81-144-75-3.midband.mdip.bt.net ([81.144.75.3]) by lon-mail-3.gradwell.net with esmtp (Gradwell gwh-smtpd 1.181) id 42725d51.17ebc.214 for ietf-usefor@imc.org; Fri, 29 Apr 2005 17:14:09 +0100 (envelope-sender ) Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3TGCdc26926 for ietf-usefor@imc.org; Fri, 29 Apr 2005 17:12:39 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20788 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: Obsolete vs. obsolescent Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: <200504111212.j3BCChM14516@clerew.man.ac.uk> <87psx1gnjc.fsf@windlord.stanford.edu> <87fyxvx1m6.fsf@windlord.stanford.edu> <87mzs2a1hu.fsf@windlord.stanford.edu> <87wtqxui6s.fsf@windlord.stanford.edu> <87d5so2bpk.fsf@windlord.stanford.edu> <42697BB6.6074@xyzzy.claranet.de> <426BC455.2040108@isode.com> <426E4B45.7010205@oceana.com> <426F4F7C.7385@xyzzy.claranet.de> <87k6moksps.fsf@windlord.stanford.edu> <87br7yg7b7.fsf@windlord.stanford.edu> <427173A8.63D5@xyzzy.claranet.de> <87zmvijwi4.fsf@windlord.stanford.edu> Date: Fri, 29 Apr 2005 15:33:26 GMT Lines: 53 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In <87zmvijwi4.fsf@windlord.stanford.edu> Russ Allbery writes: >Frank Ellermann writes: >> Okay, but it's not nice when it has serious problems with the >> propagation. An error message "administrative restriction" >> is also not exactly clear. Or if we'd expect gateways to do >> something special, or better nothing unless they are forced >> to bypass an "administrative restriction" we have to say so >> somewhere. Not in usefor-x, that was probably a bad idea. >A subject starting with "cmsg " needs to be prohibited in anything that >isn't a control message. I don't feel strongly about whether that should >be in USEFOR or USEPRO, although to me it feels like a format issue, but >it needs to be in there somewhere. If you want to say that, then USEFOR is clearly the right place to say it. But evidently Frank doesn't want that to be said (he is complaining about current systems that object when you try to do it), and I don't think I want to do it either. Better just to let the usage die away (which USEPRO requires). Otherwise we shall find new software coming on stream that objects to 'cmsg' in Subjects (just to comply with the new USEFOR) just as the last remaining servers that might have interpreted that 'cmsg' are being withdrawn. Eventually, we shall have wide deployment of a mandatory requirement for Subject header which is there to solve a problem that no longer exists. > It's premature to assume that cmsg no >longer receives special handling; we're not done with that transition. We >might never be, frankly. Using cmsg in the Subject header of regular >posts will cause significant interoperability problems. How likely is it that people will 'accidentally' write Subject headers starting with 'cmsg'? How much of a problem are such headers causing at the present time? And if they do write such a header (without any intent to cause a Control action), how likely is it that it will be meaningful enough to cause some untoward effect? If it ain't actually broke at the present time, then don't waste time trying to fix it. But if the WG really wants this prohibition including, then let it say so, in which case USEFOR is the place to do it. -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 From owner-ietf-usefor@mail.imc.org Fri Apr 29 12:15:44 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id MAA10998 for ; Fri, 29 Apr 2005 12:15:44 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3TGE9Lj028356 for ; Fri, 29 Apr 2005 09:14:09 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3TGE9kY028354 for ietf-usefor-skb; Fri, 29 Apr 2005 09:14:09 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from lon-mail-3.gradwell.net (lon-mail-3.gradwell.net [193.111.201.127]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3TGE8Yh028340 for ; Fri, 29 Apr 2005 09:14:08 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from host81-144-75-3.midband.mdip.bt.net ([81.144.75.3]) by lon-mail-3.gradwell.net with esmtp (Gradwell gwh-smtpd 1.181) id 42725d4e.17ebc.211 for ietf-usefor@imc.org; Fri, 29 Apr 2005 17:14:06 +0100 (envelope-sender ) Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3TGCbM26921 for ietf-usefor@imc.org; Fri, 29 Apr 2005 17:12:37 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20787 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: Suggested References texts Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: <426F443C.6561@xyzzy.claranet.de> <42716C2A.4F53@xyzzy.claranet.de> <871x8ubhim.fsf@windlord.stanford.edu> Date: Fri, 29 Apr 2005 15:18:12 GMT Lines: 20 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In <871x8ubhim.fsf@windlord.stanford.edu> Russ Allbery writes: >I *don't* want to give the impression, even in USEFOR, that all the rules >in RFC 2822 for replies apply unmodified to followups. Replies and >followups *aren't* the same thing, and I think it's worth making that >clear. Among other things, correctly constructing a followup requires >dealing with fields that aren't in RFC 2822 at all. Eh? Which fields might those be? -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 From owner-ietf-usefor@mail.imc.org Fri Apr 29 12:15:46 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id MAA11016 for ; Fri, 29 Apr 2005 12:15:45 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3TGE9df028355 for ; Fri, 29 Apr 2005 09:14:09 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3TGE9S0028353 for ietf-usefor-skb; Fri, 29 Apr 2005 09:14:09 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from lon-mail-3.gradwell.net (lon-mail-3.gradwell.net [193.111.201.127]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3TGE8wt028341 for ; Fri, 29 Apr 2005 09:14:08 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from host81-144-75-3.midband.mdip.bt.net ([81.144.75.3]) by lon-mail-3.gradwell.net with esmtp (Gradwell gwh-smtpd 1.181) id 42725d4f.17ebc.212 for ietf-usefor@imc.org; Fri, 29 Apr 2005 17:14:07 +0100 (envelope-sender ) Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3TGCeD26930 for ietf-usefor@imc.org; Fri, 29 Apr 2005 17:12:40 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20789 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: Again broken Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: <4255153E.3080500@isode.com> <42594879.3B92@xyzzy.claranet.de> <425A8A63.268A@xyzzy.claranet.de> <425C607E.1B5A@xyzzy.claranet.de> <425D8031.D57@xyzzy.claranet.de> <426BB63D.3060506@isode.com> <426F6107.666C@xyzzy.claranet.de> <427192C3.301E@xyzzy.claranet.de> <87sm1apchr.fsf@windlord.stanford.edu> Date: Fri, 29 Apr 2005 15:48:22 GMT Lines: 39 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In <87sm1apchr.fsf@windlord.stanford.edu> Russ Allbery writes: >RFC 2822 doesn't say anything about Control headers, for obvious reasons. >What are we gaining by adding comments to the header and breaking existing >software? Consistency with conventions for headers elsewhere. Users don't want to have to remember which headers allow comments and which don't (except where there are strong and permanent reasons to the contrary, such as performance issues that are never going to go away). Sooner or later, some enterprising person is going to write Control: cancel <12345@example.com> (unapproved article in moderated group) and will be most miffed if it is rejected or doesn't work. Yes, it may not work with many current servers, but there is no reason why it should not work long term. It is not rocket science to preprocess all headers to remove all comments and foldings before acting upon them. Anyway, the WG has never asked for this particular comment to be forbidden (and it _did_ decide very early on that comments were normally to be accepted everywhere, but maybe not generated "yet"). It was not forbidden in any of the earlier drafts, so it is premature for USEFOR to have removed it. If the WG wants this change, then it ought to be discussed first. -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 From owner-ietf-usefor@mail.imc.org Fri Apr 29 12:15:47 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id MAA11024 for ; Fri, 29 Apr 2005 12:15:46 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3TGEAh9028366 for ; Fri, 29 Apr 2005 09:14:10 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3TGEArp028365 for ietf-usefor-skb; Fri, 29 Apr 2005 09:14:10 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from lon-mail-3.gradwell.net (lon-mail-3.gradwell.net [193.111.201.127]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3TGE95u028342 for ; Fri, 29 Apr 2005 09:14:09 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from host81-144-75-3.midband.mdip.bt.net ([81.144.75.3]) by lon-mail-3.gradwell.net with esmtp (Gradwell gwh-smtpd 1.181) id 42725d50.17ebc.213 for ietf-usefor@imc.org; Fri, 29 Apr 2005 17:14:08 +0100 (envelope-sender ) Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3TGCg426935 for ietf-usefor@imc.org; Fri, 29 Apr 2005 17:12:42 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20790 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: Again broken Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: <4255153E.3080500@isode.com> <42594879.3B92@xyzzy.claranet.de> <425A8A63.268A@xyzzy.claranet.de> <425C607E.1B5A@xyzzy.claranet.de> <425D8031.D57@xyzzy.claranet.de> <426BB63D.3060506@isode.com> <426F6107.666C@xyzzy.claranet.de> <427192C3.301E@xyzzy.claranet.de> Date: Fri, 29 Apr 2005 16:08:38 GMT Lines: 56 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In <427192C3.301E@xyzzy.claranet.de> Frank Ellermann writes: >Charles Lindsey wrote: >> , and are all defined in RFC 2045 >Yes. So far I didn't know that usefor-03 implicitly includes >all 2045 terms. I only knew that it includes 2822 minus obs, >and of course 2234bis. It needs to be stated somewhere. >2045 is older than 2234. IMHO it's better to copy (and check) >what you need, instead of implicitly importing it. Checking: > token := 1* or tspecials> >Whatever that is, it's no proper ABNF. So why did Keith do it >in 3834, Because that is exactly what RFC 2045 says. It even appeared (but not normatively) in my old draft-13. But the real problems are and . In draft-13, I actually had (simplifying slightly, because there were also RFC 2231 complications): parameter = [CFWS] token [CFWS] "=" value value = [CFWS] token [CFWS] / quoted-string (again, not normatively) But I was forbidden from having that stuff incorporated into USEFOR. As a result, USEFOR does not adequately explain where [CFWS] can appear in Injection-Info. >> There is a problem in RFC 2822 (which you yourself spotted) >A _real_ problem in 2822 ? So far I tended to hallucinate >problems, but the worst I'm sure of was an "I don't like it". OK, if you don't want it fixed, let it go. But just using in place of will not fix it either, without further verbiage. -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 From owner-ietf-usefor@mail.imc.org Fri Apr 29 12:28:54 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id MAA12145 for ; Fri, 29 Apr 2005 12:28:53 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3TGSCUP031320 for ; Fri, 29 Apr 2005 09:28:12 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3TGSCO9031319 for ietf-usefor-skb; Fri, 29 Apr 2005 09:28:12 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtp2.Stanford.EDU (smtp2.Stanford.EDU [171.67.16.125]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3TGSBhi031311 for ; Fri, 29 Apr 2005 09:28:11 -0700 (PDT) (envelope-from rra@stanford.edu) Received: from windlord.stanford.edu (windlord.Stanford.EDU [171.64.19.147]) by smtp2.Stanford.EDU (8.12.11/8.12.11) with SMTP id j3TGSBoe030840 for ; Fri, 29 Apr 2005 09:28:11 -0700 Received: (qmail 16647 invoked by uid 1000); 29 Apr 2005 16:28:10 -0000 To: ietf-usefor@imc.org Subject: Re: Suggested References texts In-Reply-To: (Charles Lindsey's message of "Fri, 29 Apr 2005 15:18:12 GMT") References: <426F443C.6561@xyzzy.claranet.de> <42716C2A.4F53@xyzzy.claranet.de> <871x8ubhim.fsf@windlord.stanford.edu> From: Russ Allbery Organization: The Eyrie Date: Fri, 29 Apr 2005 09:28:10 -0700 Message-ID: <87vf651rad.fsf@windlord.stanford.edu> User-Agent: Gnus/5.1007 (Gnus v5.10.7) XEmacs/21.4 (Jumbo Shrimp, linux) MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Charles Lindsey writes: > Russ Allbery writes: >> I *don't* want to give the impression, even in USEFOR, that all the >> rules in RFC 2822 for replies apply unmodified to followups. Replies >> and followups *aren't* the same thing, and I think it's worth making >> that clear. Among other things, correctly constructing a followup >> requires dealing with fields that aren't in RFC 2822 at all. > Eh? Which fields might those be? Followup-To and Newsgroups. -- Russ Allbery (rra@stanford.edu) From owner-ietf-usefor@mail.imc.org Fri Apr 29 12:33:07 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id MAA12476 for ; Fri, 29 Apr 2005 12:33:07 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3TGWUrA031980 for ; Fri, 29 Apr 2005 09:32:30 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3TGWU0Q031979 for ietf-usefor-skb; Fri, 29 Apr 2005 09:32:30 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtp1.Stanford.EDU (smtp1.Stanford.EDU [171.67.16.123]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3TGWTUp031964 for ; Fri, 29 Apr 2005 09:32:29 -0700 (PDT) (envelope-from rra@stanford.edu) Received: from windlord.stanford.edu (windlord.Stanford.EDU [171.64.19.147]) by smtp1.Stanford.EDU (8.12.11/8.12.11) with SMTP id j3TGWTfW032369 for ; Fri, 29 Apr 2005 09:32:29 -0700 Received: (qmail 16732 invoked by uid 1000); 29 Apr 2005 16:32:28 -0000 To: ietf-usefor@imc.org Subject: Re: Obsolete vs. obsolescent In-Reply-To: (Charles Lindsey's message of "Fri, 29 Apr 2005 15:33:26 GMT") References: <87psx1gnjc.fsf@windlord.stanford.edu> <87fyxvx1m6.fsf@windlord.stanford.edu> <87mzs2a1hu.fsf@windlord.stanford.edu> <87wtqxui6s.fsf@windlord.stanford.edu> <87d5so2bpk.fsf@windlord.stanford.edu> <42697BB6.6074@xyzzy.claranet.de> <426BC455.2040108@isode.com> <426E4B45.7010205@oceana.com> <426F4F7C.7385@xyzzy.claranet.de> <87k6moksps.fsf@windlord.stanford.edu> <87br7yg7b7.fsf@windlord.stanford.edu> <427173A8.63D5@xyzzy.claranet.de> <87zmvijwi4.fsf@windlord.stanford.edu> From: Russ Allbery Organization: The Eyrie Date: Fri, 29 Apr 2005 09:32:28 -0700 Message-ID: <87r7gt1r37.fsf@windlord.stanford.edu> User-Agent: Gnus/5.1007 (Gnus v5.10.7) XEmacs/21.4 (Jumbo Shrimp, linux) MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Charles Lindsey writes: > But evidently Frank doesn't want that to be said (he is complaining > about current systems that object when you try to do it), and I don't > think I want to do it either. Better just to let the usage die away > (which USEPRO requires). That isn't going to actually happen in any sort of reasonable time frame. > Otherwise we shall find new software coming on stream that objects to > 'cmsg' in Subjects (just to comply with the new USEFOR) just as the last > remaining servers that might have interpreted that 'cmsg' are being > withdrawn. I wonder if any of us will live long enough to see the last remaining servers that interpreted cmsg be withdrawn. > How likely is it that people will 'accidentally' write Subject headers > starting with 'cmsg'? This doesn't seem like a relevant question to me. Standards aren't concerned only with things that can happen accidentally. > How much of a problem are such headers causing at the present time? They aren't propagated. They have huge interoperability problems at the present time. > And if they do write such a header (without any intent to cause a > Control action), how likely is it that it will be meaningful enough to > cause some untoward effect? That's not the breakage that I'm worried about (control messages should be authenticated anyway instead of cancels, and it's pretty hard to write a cancel control message accidentally). The article will mysteriously disappear, which is a different kind of breakage. > If it ain't actually broke at the present time, then don't waste time > trying to fix it. It's very broken at the present time. -- Russ Allbery (rra@stanford.edu) From owner-ietf-usefor@mail.imc.org Fri Apr 29 12:48:53 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id MAA13910 for ; Fri, 29 Apr 2005 12:48:52 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3TGli1Y034566 for ; Fri, 29 Apr 2005 09:47:44 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3TGliUc034565 for ietf-usefor-skb; Fri, 29 Apr 2005 09:47:44 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from pine.epix.net (pine.epix.net [199.224.64.53]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3TGlhgP034559 for ; Fri, 29 Apr 2005 09:47:44 -0700 (PDT) (envelope-from mibsoft@epix.net) Received: from [192.168.2.11] (hrbg-216-108-206-236-pppoe.dsl.hrbg.epix.net [216.108.206.236]) by pine.epix.net (8.12.10/2004120601/PL) with ESMTP id j3TGlZov008073 for ; Fri, 29 Apr 2005 12:47:40 -0400 (EDT) Message-ID: <4272652A.9030101@epix.net> Date: Fri, 29 Apr 2005 12:47:38 -0400 From: "Forrest J. Cavalier III" User-Agent: Mozilla Thunderbird 0.7 (Windows/20040616) X-Accept-Language: en-us, en MIME-Version: 1.0 To: ietf-usefor@imc.org Subject: Re: Again broken References: <4255153E.3080500@isode.com> <42594879.3B92@xyzzy.claranet.de> <425A8A63.268A@xyzzy.claranet.de> <425C607E.1B5A@xyzzy.claranet.de> <425D8031.D57@xyzzy.claranet.de> <426BB63D.3060506@isode.com> <426F6107.666C@xyzzy.claranet.de> <427192C3.301E@xyzzy.claranet.de> <87sm1apchr.fsf@windlord.stanford.edu> In-Reply-To: Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit X-Scanned-By: MIMEDefang 2.51 on 199.224.89.153 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Content-Transfer-Encoding: 7bit Charles Lindsey wrote: > If the WG wants this change, then it ought to be discussed first. There you go again. Charles, the only meaning of "change" that makes sense is a change from current published standards and practice. I believe you have once again used it to mean "an edit to the draft." Please stop it. If something appeared in the draft that was never discussed in the WG, it ought to be removed. Removing such things do not need to be discussed first, because removing them is not a change, having them in is a change. The draft is fluid, it isn't a big deal to delete and edit words. Modifying servers to parse comments in control headers is a big deal. No matter what ideas of future scenarios anyone can conjure up, unless it is a current request, or existing practice, or a forseeable need (not a "would be nice.") it ought to be cut from the drafts. It is outside the charter. From owner-ietf-usefor@mail.imc.org Fri Apr 29 20:56:21 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id UAA06248 for ; Fri, 29 Apr 2005 20:56:20 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3U0tCZT014472 for ; Fri, 29 Apr 2005 17:55:12 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3U0tCI3014471 for ietf-usefor-skb; Fri, 29 Apr 2005 17:55:12 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from ciao.gmane.org (main.gmane.org [80.91.229.2]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3U0tAoD014462 for ; Fri, 29 Apr 2005 17:55:10 -0700 (PDT) (envelope-from usenet-format@gmane.org) Received: from list by ciao.gmane.org with local (Exim 4.43) id 1DRgAI-00068a-S8 for ietf-usefor@imc.org; Sat, 30 Apr 2005 02:49:02 +0200 Received: from 212.82.251.236 ([212.82.251.236]) by main.gmane.org with esmtp (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Sat, 30 Apr 2005 02:49:02 +0200 Received: from nobody by 212.82.251.236 with local (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Sat, 30 Apr 2005 02:49:02 +0200 X-Injected-Via-Gmane: http://gmane.org/ To: ietf-usefor@imc.org From: Frank Ellermann Subject: Re: Suggested References texts Date: Sat, 30 Apr 2005 02:44:16 +0200 Organization: Lines: 107 Message-ID: <4272D4E0.3BB8@xyzzy.claranet.de> References: <426F443C.6561@xyzzy.claranet.de> <42716C2A.4F53@xyzzy.claranet.de> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Complaints-To: usenet@sea.gmane.org X-Gmane-NNTP-Posting-Host: 212.82.251.236 X-Mailer: Mozilla 3.0 (OS/2; U) Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Content-Transfer-Encoding: 7bit Charles Lindsey wrote: > suppose you are a multipart FAQ writer trying to use > References to refer back to the earlier parts, how do > you know what you can do? There are many possible and correct ways to do this, and they are all obvious. It's 2005. If somebody posts a multipart FAQ in an NG where I don't like it I'm tempted to report it as net abuse. Please no UUCP arguments, and of course there's nothing wrong with reading news offline, but URLs really exist in this millennium. Too many "FAQs" are in fact only auto-posted propagada from self-appointed authorities. > Or you are writing a threading agent. How do you know > what you can rely on in a properly formed References > header? There will be two documented ways to create it, 2822 and Usepro-xx, so I'd assume that that's "properly formed". I'd test this assumption with some real References, find that it's generally true, and use it. Minimal testing would convince me that a reading agent does not always "see" all referenced articles. Maybe I'd try a RTL strategy to find a visible precursor. If that doesn't work I'd try to attach the article to an existing thread with another strategy, this could be a set operation on known Message-IDs. If it still doesn't work it's a new thread, e.g. the result of a Followup-To from an invisible NG. That takes care of anything with References. The other case, no References, is also simple, some regexp on the Subject (Re: etc.) with a plausibility test on the Date. If it's an UA also working for mail I'd add In-Reply-To as another source for this algorithm (details specified in 2822), or maybe I'd do this anyway. There are several variants like allowing threads to be broken by new subjects, but I'd stay away from it unless my algorithm to compare Subjects supports 2231 / 2047. > what is to stop you including s of articles that > never existed? Certainly irrelevant for a reading agent, it cannot know whether referenced articles are invisible or fictitious. For the author of a multipart FAQ it could be an idea to base his parts on a fictitious root. And after the tenth question about an apparently lost root he knows that it's a bad idea. But in theory it's "allowed", a reading agent crashing in this situation is broken. > What is to stop you listing s in the wrong > order? Solved by induction, there's essentially one documented way to create a 2822-reply or news-followup, the result is "properly formed". Even if the older References were garbage, the last could be enough to find the precursor. If not it's a case of garbage in garbage out, and there's nothing Usepro-xx or one of its many agents can do about it. Okay, they should remove / ignore syntactical garbage. > most FAQ writers are clueful enough to construct sensible > References headers, but in fact they are merely relying > on "folklore" What's so difficult with these sets and trees ? If I cannot guarantee date( part( n+1 )) > date( part( n )) then I should simply add part( n ) to the References in part( n+1 ). That's no urban legend or rocket science, it's common sense. And Usenet would survive it if I get wrong and let all parts only reference part( 1 ) or other dubious ideas. > they have to "reverse engineer" the algorithm given for > followups in USEPRO, and deduce the underlying semantics > from that There are several more or less working algorithms. What you're proposing is to define THE algorithm. But we know that there is no single valid algorithm, or rather that we are unable to agree on whatever single algorithm. > RFC 2888 doesn't explain the _meaning_ of the References > header. No surprise if I look into the 2822 credits. Was it exactly the same stalemate as here with the same combattants ? At least 2822 mumbles something about "threads", good enough. And it also doesn't forbid a more literal interpretation of "reference". But OTOH it doesn't mention this, if that's a clue I got it. Bye, Frank From owner-ietf-usefor@mail.imc.org Fri Apr 29 22:01:28 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id WAA09545 for ; Fri, 29 Apr 2005 22:01:28 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3U205Hh023773 for ; Fri, 29 Apr 2005 19:00:05 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3U205Xp023772 for ietf-usefor-skb; Fri, 29 Apr 2005 19:00:05 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from ciao.gmane.org (main.gmane.org [80.91.229.2]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3U2031D023754 for ; Fri, 29 Apr 2005 19:00:04 -0700 (PDT) (envelope-from usenet-format@gmane.org) Received: from list by ciao.gmane.org with local (Exim 4.43) id 1DRhB1-0000vI-EY for ietf-usefor@imc.org; Sat, 30 Apr 2005 03:53:51 +0200 Received: from 212.82.251.236 ([212.82.251.236]) by main.gmane.org with esmtp (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Sat, 30 Apr 2005 03:53:51 +0200 Received: from nobody by 212.82.251.236 with local (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Sat, 30 Apr 2005 03:53:51 +0200 X-Injected-Via-Gmane: http://gmane.org/ To: ietf-usefor@imc.org From: Frank Ellermann Subject: Re: Again broken Date: Sat, 30 Apr 2005 03:32:34 +0200 Organization: Lines: 20 Message-ID: <4272E032.4BCB@xyzzy.claranet.de> References: <4255153E.3080500@isode.com> <42594879.3B92@xyzzy.claranet.de> <425A8A63.268A@xyzzy.claranet.de> <425C607E.1B5A@xyzzy.claranet.de> <425D8031.D57@xyzzy.claranet.de> <426BB63D.3060506@isode.com> <426F6107.666C@xyzzy.claranet.de> <427192C3.301E@xyzzy.claranet.de> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Complaints-To: usenet@sea.gmane.org X-Gmane-NNTP-Posting-Host: 212.82.251.236 X-Mailer: Mozilla 3.0 (OS/2; U) Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Content-Transfer-Encoding: 7bit Charles Lindsey wrote: > OK, if you don't want it fixed, let it go. I've proposed many ways to fix it, e.g. by clear names for the msg-id productions, or the last proposal was this note: | Note: id-right is a domain, and no-fold-literal is the | special case of a domain-literal (a.k.a. address-literal, | see RfC 2821) Plus a general note about avoiding domain-literal whereever possible, i.e. the same idea as in STD 11. > But just using in place of will not > fix it either, without further verbiage. IBTD. If the msg-id is broken in relation to 1036 / s-o-1036 usefor-xx MUST NOT fly as far as I'm concerned. Bye, Frank From owner-ietf-usefor@mail.imc.org Fri Apr 29 22:08:20 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id WAA10004 for ; Fri, 29 Apr 2005 22:08:19 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3U27S7D024824 for ; Fri, 29 Apr 2005 19:07:28 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3U27SEn024823 for ietf-usefor-skb; Fri, 29 Apr 2005 19:07:28 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from ciao.gmane.org (main.gmane.org [80.91.229.2]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3U27RnE024814 for ; Fri, 29 Apr 2005 19:07:27 -0700 (PDT) (envelope-from usenet-format@gmane.org) Received: from root by ciao.gmane.org with local (Exim 4.43) id 1DRhHc-0002Gp-Qg for ietf-usefor@imc.org; Sat, 30 Apr 2005 04:00:41 +0200 Received: from 212.82.251.236 ([212.82.251.236]) by main.gmane.org with esmtp (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Sat, 30 Apr 2005 04:00:40 +0200 Received: from nobody by 212.82.251.236 with local (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Sat, 30 Apr 2005 04:00:40 +0200 X-Injected-Via-Gmane: http://gmane.org/ To: ietf-usefor@imc.org From: Frank Ellermann Subject: CFWS (was: Again broken) Date: Sat, 30 Apr 2005 03:11:26 +0200 Organization: Lines: 12 Message-ID: <4272DB3E.74CA@xyzzy.claranet.de> References: <4255153E.3080500@isode.com> <42594879.3B92@xyzzy.claranet.de> <425A8A63.268A@xyzzy.claranet.de> <425C607E.1B5A@xyzzy.claranet.de> <425D8031.D57@xyzzy.claranet.de> <426BB63D.3060506@isode.com> <426F6107.666C@xyzzy.claranet.de> <427192C3.301E@xyzzy.claranet.de> <87sm1apchr.fsf@windlord.stanford.edu> Mime-Version: 1.0 Content-Type: text/plain; charset="windows-1252" Content-Transfer-Encoding: quoted-printable X-Complaints-To: usenet@sea.gmane.org X-Gmane-NNTP-Posting-Host: 212.82.251.236 X-Mailer: Mozilla 3.0 (OS/2; U) Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Content-Transfer-Encoding: quoted-printable Charles Lindsey wrote: > it ought to be discussed Kill all CFWS in pre-2045 news-only-headers where it doesn't contradict 2045 / 2822. For new headers like User-Agent or Injection-Info adopt common practice or 2045 / 2822 style. I know that it's not what the WG once wanted, but you asked. =80 0.02, Frank From owner-ietf-usefor@mail.imc.org Fri Apr 29 23:28:55 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id XAA14539 for ; Fri, 29 Apr 2005 23:28:54 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3U3RfOi036281 for ; Fri, 29 Apr 2005 20:27:41 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3U3Rfuk036280 for ietf-usefor-skb; Fri, 29 Apr 2005 20:27:41 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from ciao.gmane.org (main.gmane.org [80.91.229.2]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3U3Rd4U036253 for ; Fri, 29 Apr 2005 20:27:39 -0700 (PDT) (envelope-from usenet-format@gmane.org) Received: from list by ciao.gmane.org with local (Exim 4.43) id 1DRiY3-0007Di-Bm for ietf-usefor@imc.org; Sat, 30 Apr 2005 05:21:43 +0200 Received: from 212.82.251.236 ([212.82.251.236]) by main.gmane.org with esmtp (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Sat, 30 Apr 2005 05:21:43 +0200 Received: from nobody by 212.82.251.236 with local (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Sat, 30 Apr 2005 05:21:43 +0200 X-Injected-Via-Gmane: http://gmane.org/ To: ietf-usefor@imc.org From: Frank Ellermann Subject: 2045 ABNF (was: Again broken) Date: Sat, 30 Apr 2005 05:14:59 +0200 Organization: Lines: 71 Message-ID: <4272F833.324@xyzzy.claranet.de> References: <4255153E.3080500@isode.com> <42594879.3B92@xyzzy.claranet.de> <425A8A63.268A@xyzzy.claranet.de> <425C607E.1B5A@xyzzy.claranet.de> <425D8031.D57@xyzzy.claranet.de> <426BB63D.3060506@isode.com> <426F6107.666C@xyzzy.claranet.de> <427192C3.301E@xyzzy.claranet.de> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Complaints-To: usenet@sea.gmane.org X-Gmane-NNTP-Posting-Host: 212.82.251.236 X-Mailer: Mozilla 3.0 (OS/2; U) Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Content-Transfer-Encoding: 7bit Charles Lindsey wrote: >> So far I didn't know that usefor-03 implicitly includes >> all 2045 terms. I only knew that it includes 2822 minus >> obs, and of course 2234bis. > It needs to be stated somewhere. It is, I was blind. But we cannot use this pseudo-ABNF, it has to be fixed for the three (?) 2045 terms needed in usefor-xx: value, token, parameter, anything else ? > It even appeared (but not normatively) in my old draft-13. Yes, the "collected ABNF" was nice, but I never checked all the details in 2003 for draft 7 and its "last last last call". Fortunately Bruce was an inspiration how to get these details right, but that was later from my POV. 1) value = token / quoted-string No "value =" in 2822, no "value :=" in 2231, so that should be no problem as soon as you have a valid "token" elsewhere. 2) parameter = attribute "=" value Trouble, "parameter" is redefined in 2231. While I'd love to nullify 2231 that's not excatly realistic. Maybe we could nominate 2231 by a joint agreement of LTRU and USEFOR as the most painful RfC. :-( Better remove the one "archive-param". 2a) attribute = token Same problem as with "parameter", but you only (?) need it indirectly for "archive-param". In theory it's very easy to translate the 2231 gibberish to proper ABNF: attribute = 1*attribute-char attribute-char = A proper subset of US-ASCII, "tspecials".defined in 2045, and 2045 took it from STD 11. But not the same as the "specials" in 2822. I need some ASCII-subset tools. 3) token = 1* Okay, that's the same as attribute-char plus "*", "'", "%". > parameter = [CFWS] token [CFWS] "=" value > value = [CFWS] token [CFWS] / quoted-string Where have you found these [CFWS] ? Because 2045 mentions 822 and its comments, and 2822 claims to update 822 ? > But I was forbidden from having that stuff incorporated into > USEFOR. Do you have a Message-ID for this obscure decree ? I see the potential problem with the optional CFWS, but otherwise your "value" is just a copy from 2045. Okay, "parameter" is a bad case of 2231, see above. > as a result, USEFOR does not adequately explain where [CFWS] > can appear in Injection-Info. Okay, then let's reinvent the wheel, no 2231 and no CFWS. Bye From owner-ietf-usefor@mail.imc.org Sat Apr 30 03:20:14 2005 Received: from above.proper.com (above.proper.com [208.184.76.39]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id DAA16697 for ; Sat, 30 Apr 2005 03:20:13 -0400 (EDT) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3U7Iq1M014460 for ; Sat, 30 Apr 2005 00:18:52 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3U7IqEH014459 for ietf-usefor-skb; Sat, 30 Apr 2005 00:18:52 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtps-vbr1.xs4all.nl (smtps-vbr1.xs4all.nl [194.109.24.19]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3U7IoQC014427 for ; Sat, 30 Apr 2005 00:18:51 -0700 (PDT) (envelope-from rvtol@isolution.nl) Received: from isop10 (velvet.isolution.nl [194.109.164.102]) (authenticated bits=0) by smtps-vbr1.xs4all.nl (8.12.11/8.12.11) with ESMTP id j3U7IZoD052467 (version=TLSv1/SSLv3 cipher=RC4-MD5 bits=128 verify=NO); Sat, 30 Apr 2005 09:18:41 +0200 (CEST) (envelope-from rvtol@isolution.nl) Message-ID: <075201c54d54$d7b2d460$0b01a8c0@isolution.nl> From: "Ruud H.G. van Tol" To: Cc: "Charles Lindsey" References: <426F443C.6561@xyzzy.claranet.de> Subject: Re: Suggested References texts Date: Sat, 30 Apr 2005 09:18:29 +0200 Organization: Chaos rules. MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-15" Content-Transfer-Encoding: 7bit X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 6.00.2800.1478 X-MIMEOLE: Produced By Microsoft MimeOLE V6.00.2800.1478 X-Virus-Scanned: by XS4ALL Virus Scanner Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Content-Transfer-Encoding: 7bit Charles Lindsey: > references = "References:" SP [CFWS] msg-id 1*(CFWS msg-id ) > [CFWS] CRLF I suppose that was meant as references = "References:" SP [CFWS] msg-id *( CFWS msg-id ) [CFWS] CRLF (changed the '1*' into '*'; inserted a space) -- Grtz, Ruud Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3U7Iq1M014460 for ; Sat, 30 Apr 2005 00:18:52 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3U7IqEH014459 for ietf-usefor-skb; Sat, 30 Apr 2005 00:18:52 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtps-vbr1.xs4all.nl (smtps-vbr1.xs4all.nl [194.109.24.19]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3U7IoQC014427 for ; Sat, 30 Apr 2005 00:18:51 -0700 (PDT) (envelope-from rvtol@isolution.nl) Received: from isop10 (velvet.isolution.nl [194.109.164.102]) (authenticated bits=0) by smtps-vbr1.xs4all.nl (8.12.11/8.12.11) with ESMTP id j3U7IZoD052467 (version=TLSv1/SSLv3 cipher=RC4-MD5 bits=128 verify=NO); Sat, 30 Apr 2005 09:18:41 +0200 (CEST) (envelope-from rvtol@isolution.nl) Message-ID: <075201c54d54$d7b2d460$0b01a8c0@isolution.nl> From: "Ruud H.G. van Tol" To: Cc: "Charles Lindsey" References: <426F443C.6561@xyzzy.claranet.de> Subject: Re: Suggested References texts Date: Sat, 30 Apr 2005 09:18:29 +0200 Organization: Chaos rules. MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-15" Content-Transfer-Encoding: 7bit X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 6.00.2800.1478 X-MIMEOLE: Produced By Microsoft MimeOLE V6.00.2800.1478 X-Virus-Scanned: by XS4ALL Virus Scanner Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Charles Lindsey: > references = "References:" SP [CFWS] msg-id 1*(CFWS msg-id ) > [CFWS] CRLF I suppose that was meant as references = "References:" SP [CFWS] msg-id *( CFWS msg-id ) [CFWS] CRLF (changed the '1*' into '*'; inserted a space) -- Grtz, Ruud Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3U3RfOi036281 for ; Fri, 29 Apr 2005 20:27:41 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3U3Rfuk036280 for ietf-usefor-skb; Fri, 29 Apr 2005 20:27:41 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from ciao.gmane.org (main.gmane.org [80.91.229.2]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3U3Rd4U036253 for ; Fri, 29 Apr 2005 20:27:39 -0700 (PDT) (envelope-from usenet-format@gmane.org) Received: from list by ciao.gmane.org with local (Exim 4.43) id 1DRiY3-0007Di-Bm for ietf-usefor@imc.org; Sat, 30 Apr 2005 05:21:43 +0200 Received: from 212.82.251.236 ([212.82.251.236]) by main.gmane.org with esmtp (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Sat, 30 Apr 2005 05:21:43 +0200 Received: from nobody by 212.82.251.236 with local (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Sat, 30 Apr 2005 05:21:43 +0200 X-Injected-Via-Gmane: http://gmane.org/ To: ietf-usefor@imc.org From: Frank Ellermann Subject: 2045 ABNF (was: Again broken) Date: Sat, 30 Apr 2005 05:14:59 +0200 Organization: Lines: 71 Message-ID: <4272F833.324@xyzzy.claranet.de> References: <4255153E.3080500@isode.com> <42594879.3B92@xyzzy.claranet.de> <425A8A63.268A@xyzzy.claranet.de> <425C607E.1B5A@xyzzy.claranet.de> <425D8031.D57@xyzzy.claranet.de> <426BB63D.3060506@isode.com> <426F6107.666C@xyzzy.claranet.de> <427192C3.301E@xyzzy.claranet.de> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Complaints-To: usenet@sea.gmane.org X-Gmane-NNTP-Posting-Host: 212.82.251.236 X-Mailer: Mozilla 3.0 (OS/2; U) Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Charles Lindsey wrote: >> So far I didn't know that usefor-03 implicitly includes >> all 2045 terms. I only knew that it includes 2822 minus >> obs, and of course 2234bis. > It needs to be stated somewhere. It is, I was blind. But we cannot use this pseudo-ABNF, it has to be fixed for the three (?) 2045 terms needed in usefor-xx: value, token, parameter, anything else ? > It even appeared (but not normatively) in my old draft-13. Yes, the "collected ABNF" was nice, but I never checked all the details in 2003 for draft 7 and its "last last last call". Fortunately Bruce was an inspiration how to get these details right, but that was later from my POV. 1) value = token / quoted-string No "value =" in 2822, no "value :=" in 2231, so that should be no problem as soon as you have a valid "token" elsewhere. 2) parameter = attribute "=" value Trouble, "parameter" is redefined in 2231. While I'd love to nullify 2231 that's not excatly realistic. Maybe we could nominate 2231 by a joint agreement of LTRU and USEFOR as the most painful RfC. :-( Better remove the one "archive-param". 2a) attribute = token Same problem as with "parameter", but you only (?) need it indirectly for "archive-param". In theory it's very easy to translate the 2231 gibberish to proper ABNF: attribute = 1*attribute-char attribute-char = A proper subset of US-ASCII, "tspecials".defined in 2045, and 2045 took it from STD 11. But not the same as the "specials" in 2822. I need some ASCII-subset tools. 3) token = 1* Okay, that's the same as attribute-char plus "*", "'", "%". > parameter = [CFWS] token [CFWS] "=" value > value = [CFWS] token [CFWS] / quoted-string Where have you found these [CFWS] ? Because 2045 mentions 822 and its comments, and 2822 claims to update 822 ? > But I was forbidden from having that stuff incorporated into > USEFOR. Do you have a Message-ID for this obscure decree ? I see the potential problem with the optional CFWS, but otherwise your "value" is just a copy from 2045. Okay, "parameter" is a bad case of 2231, see above. > as a result, USEFOR does not adequately explain where [CFWS] > can appear in Injection-Info. Okay, then let's reinvent the wheel, no 2231 and no CFWS. Bye Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3U27S7D024824 for ; Fri, 29 Apr 2005 19:07:28 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3U27SEn024823 for ietf-usefor-skb; Fri, 29 Apr 2005 19:07:28 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from ciao.gmane.org (main.gmane.org [80.91.229.2]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3U27RnE024814 for ; Fri, 29 Apr 2005 19:07:27 -0700 (PDT) (envelope-from usenet-format@gmane.org) Received: from root by ciao.gmane.org with local (Exim 4.43) id 1DRhHc-0002Gp-Qg for ietf-usefor@imc.org; Sat, 30 Apr 2005 04:00:41 +0200 Received: from 212.82.251.236 ([212.82.251.236]) by main.gmane.org with esmtp (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Sat, 30 Apr 2005 04:00:40 +0200 Received: from nobody by 212.82.251.236 with local (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Sat, 30 Apr 2005 04:00:40 +0200 X-Injected-Via-Gmane: http://gmane.org/ To: ietf-usefor@imc.org From: Frank Ellermann Subject: CFWS (was: Again broken) Date: Sat, 30 Apr 2005 03:11:26 +0200 Organization: Lines: 12 Message-ID: <4272DB3E.74CA@xyzzy.claranet.de> References: <4255153E.3080500@isode.com> <42594879.3B92@xyzzy.claranet.de> <425A8A63.268A@xyzzy.claranet.de> <425C607E.1B5A@xyzzy.claranet.de> <425D8031.D57@xyzzy.claranet.de> <426BB63D.3060506@isode.com> <426F6107.666C@xyzzy.claranet.de> <427192C3.301E@xyzzy.claranet.de> <87sm1apchr.fsf@windlord.stanford.edu> Mime-Version: 1.0 Content-Type: text/plain; charset="windows-1252" Content-Transfer-Encoding: quoted-printable X-Complaints-To: usenet@sea.gmane.org X-Gmane-NNTP-Posting-Host: 212.82.251.236 X-Mailer: Mozilla 3.0 (OS/2; U) Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Charles Lindsey wrote: > it ought to be discussed Kill all CFWS in pre-2045 news-only-headers where it doesn't contradict 2045 / 2822. For new headers like User-Agent or Injection-Info adopt common practice or 2045 / 2822 style. I know that it's not what the WG once wanted, but you asked. =80 0.02, Frank Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3U205Hh023773 for ; Fri, 29 Apr 2005 19:00:05 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3U205Xp023772 for ietf-usefor-skb; Fri, 29 Apr 2005 19:00:05 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from ciao.gmane.org (main.gmane.org [80.91.229.2]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3U2031D023754 for ; Fri, 29 Apr 2005 19:00:04 -0700 (PDT) (envelope-from usenet-format@gmane.org) Received: from list by ciao.gmane.org with local (Exim 4.43) id 1DRhB1-0000vI-EY for ietf-usefor@imc.org; Sat, 30 Apr 2005 03:53:51 +0200 Received: from 212.82.251.236 ([212.82.251.236]) by main.gmane.org with esmtp (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Sat, 30 Apr 2005 03:53:51 +0200 Received: from nobody by 212.82.251.236 with local (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Sat, 30 Apr 2005 03:53:51 +0200 X-Injected-Via-Gmane: http://gmane.org/ To: ietf-usefor@imc.org From: Frank Ellermann Subject: Re: Again broken Date: Sat, 30 Apr 2005 03:32:34 +0200 Organization: Lines: 20 Message-ID: <4272E032.4BCB@xyzzy.claranet.de> References: <4255153E.3080500@isode.com> <42594879.3B92@xyzzy.claranet.de> <425A8A63.268A@xyzzy.claranet.de> <425C607E.1B5A@xyzzy.claranet.de> <425D8031.D57@xyzzy.claranet.de> <426BB63D.3060506@isode.com> <426F6107.666C@xyzzy.claranet.de> <427192C3.301E@xyzzy.claranet.de> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Complaints-To: usenet@sea.gmane.org X-Gmane-NNTP-Posting-Host: 212.82.251.236 X-Mailer: Mozilla 3.0 (OS/2; U) Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Charles Lindsey wrote: > OK, if you don't want it fixed, let it go. I've proposed many ways to fix it, e.g. by clear names for the msg-id productions, or the last proposal was this note: | Note: id-right is a domain, and no-fold-literal is the | special case of a domain-literal (a.k.a. address-literal, | see RfC 2821) Plus a general note about avoiding domain-literal whereever possible, i.e. the same idea as in STD 11. > But just using in place of will not > fix it either, without further verbiage. IBTD. If the msg-id is broken in relation to 1036 / s-o-1036 usefor-xx MUST NOT fly as far as I'm concerned. Bye, Frank Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3U0tCZT014472 for ; Fri, 29 Apr 2005 17:55:12 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3U0tCI3014471 for ietf-usefor-skb; Fri, 29 Apr 2005 17:55:12 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from ciao.gmane.org (main.gmane.org [80.91.229.2]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3U0tAoD014462 for ; Fri, 29 Apr 2005 17:55:10 -0700 (PDT) (envelope-from usenet-format@gmane.org) Received: from list by ciao.gmane.org with local (Exim 4.43) id 1DRgAI-00068a-S8 for ietf-usefor@imc.org; Sat, 30 Apr 2005 02:49:02 +0200 Received: from 212.82.251.236 ([212.82.251.236]) by main.gmane.org with esmtp (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Sat, 30 Apr 2005 02:49:02 +0200 Received: from nobody by 212.82.251.236 with local (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Sat, 30 Apr 2005 02:49:02 +0200 X-Injected-Via-Gmane: http://gmane.org/ To: ietf-usefor@imc.org From: Frank Ellermann Subject: Re: Suggested References texts Date: Sat, 30 Apr 2005 02:44:16 +0200 Organization: Lines: 107 Message-ID: <4272D4E0.3BB8@xyzzy.claranet.de> References: <426F443C.6561@xyzzy.claranet.de> <42716C2A.4F53@xyzzy.claranet.de> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Complaints-To: usenet@sea.gmane.org X-Gmane-NNTP-Posting-Host: 212.82.251.236 X-Mailer: Mozilla 3.0 (OS/2; U) Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Charles Lindsey wrote: > suppose you are a multipart FAQ writer trying to use > References to refer back to the earlier parts, how do > you know what you can do? There are many possible and correct ways to do this, and they are all obvious. It's 2005. If somebody posts a multipart FAQ in an NG where I don't like it I'm tempted to report it as net abuse. Please no UUCP arguments, and of course there's nothing wrong with reading news offline, but URLs really exist in this millennium. Too many "FAQs" are in fact only auto-posted propagada from self-appointed authorities. > Or you are writing a threading agent. How do you know > what you can rely on in a properly formed References > header? There will be two documented ways to create it, 2822 and Usepro-xx, so I'd assume that that's "properly formed". I'd test this assumption with some real References, find that it's generally true, and use it. Minimal testing would convince me that a reading agent does not always "see" all referenced articles. Maybe I'd try a RTL strategy to find a visible precursor. If that doesn't work I'd try to attach the article to an existing thread with another strategy, this could be a set operation on known Message-IDs. If it still doesn't work it's a new thread, e.g. the result of a Followup-To from an invisible NG. That takes care of anything with References. The other case, no References, is also simple, some regexp on the Subject (Re: etc.) with a plausibility test on the Date. If it's an UA also working for mail I'd add In-Reply-To as another source for this algorithm (details specified in 2822), or maybe I'd do this anyway. There are several variants like allowing threads to be broken by new subjects, but I'd stay away from it unless my algorithm to compare Subjects supports 2231 / 2047. > what is to stop you including s of articles that > never existed? Certainly irrelevant for a reading agent, it cannot know whether referenced articles are invisible or fictitious. For the author of a multipart FAQ it could be an idea to base his parts on a fictitious root. And after the tenth question about an apparently lost root he knows that it's a bad idea. But in theory it's "allowed", a reading agent crashing in this situation is broken. > What is to stop you listing s in the wrong > order? Solved by induction, there's essentially one documented way to create a 2822-reply or news-followup, the result is "properly formed". Even if the older References were garbage, the last could be enough to find the precursor. If not it's a case of garbage in garbage out, and there's nothing Usepro-xx or one of its many agents can do about it. Okay, they should remove / ignore syntactical garbage. > most FAQ writers are clueful enough to construct sensible > References headers, but in fact they are merely relying > on "folklore" What's so difficult with these sets and trees ? If I cannot guarantee date( part( n+1 )) > date( part( n )) then I should simply add part( n ) to the References in part( n+1 ). That's no urban legend or rocket science, it's common sense. And Usenet would survive it if I get wrong and let all parts only reference part( 1 ) or other dubious ideas. > they have to "reverse engineer" the algorithm given for > followups in USEPRO, and deduce the underlying semantics > from that There are several more or less working algorithms. What you're proposing is to define THE algorithm. But we know that there is no single valid algorithm, or rather that we are unable to agree on whatever single algorithm. > RFC 2888 doesn't explain the _meaning_ of the References > header. No surprise if I look into the 2822 credits. Was it exactly the same stalemate as here with the same combattants ? At least 2822 mumbles something about "threads", good enough. And it also doesn't forbid a more literal interpretation of "reference". But OTOH it doesn't mention this, if that's a clue I got it. Bye, Frank Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3TGli1Y034566 for ; Fri, 29 Apr 2005 09:47:44 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3TGliUc034565 for ietf-usefor-skb; Fri, 29 Apr 2005 09:47:44 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from pine.epix.net (pine.epix.net [199.224.64.53]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3TGlhgP034559 for ; Fri, 29 Apr 2005 09:47:44 -0700 (PDT) (envelope-from mibsoft@epix.net) Received: from [192.168.2.11] (hrbg-216-108-206-236-pppoe.dsl.hrbg.epix.net [216.108.206.236]) by pine.epix.net (8.12.10/2004120601/PL) with ESMTP id j3TGlZov008073 for ; Fri, 29 Apr 2005 12:47:40 -0400 (EDT) Message-ID: <4272652A.9030101@epix.net> Date: Fri, 29 Apr 2005 12:47:38 -0400 From: "Forrest J. Cavalier III" User-Agent: Mozilla Thunderbird 0.7 (Windows/20040616) X-Accept-Language: en-us, en MIME-Version: 1.0 To: ietf-usefor@imc.org Subject: Re: Again broken References: <4255153E.3080500@isode.com> <42594879.3B92@xyzzy.claranet.de> <425A8A63.268A@xyzzy.claranet.de> <425C607E.1B5A@xyzzy.claranet.de> <425D8031.D57@xyzzy.claranet.de> <426BB63D.3060506@isode.com> <426F6107.666C@xyzzy.claranet.de> <427192C3.301E@xyzzy.claranet.de> <87sm1apchr.fsf@windlord.stanford.edu> In-Reply-To: Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit X-Scanned-By: MIMEDefang 2.51 on 199.224.89.153 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Charles Lindsey wrote: > If the WG wants this change, then it ought to be discussed first. There you go again. Charles, the only meaning of "change" that makes sense is a change from current published standards and practice. I believe you have once again used it to mean "an edit to the draft." Please stop it. If something appeared in the draft that was never discussed in the WG, it ought to be removed. Removing such things do not need to be discussed first, because removing them is not a change, having them in is a change. The draft is fluid, it isn't a big deal to delete and edit words. Modifying servers to parse comments in control headers is a big deal. No matter what ideas of future scenarios anyone can conjure up, unless it is a current request, or existing practice, or a forseeable need (not a "would be nice.") it ought to be cut from the drafts. It is outside the charter. Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3TGWUrA031980 for ; Fri, 29 Apr 2005 09:32:30 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3TGWU0Q031979 for ietf-usefor-skb; Fri, 29 Apr 2005 09:32:30 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtp1.Stanford.EDU (smtp1.Stanford.EDU [171.67.16.123]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3TGWTUp031964 for ; Fri, 29 Apr 2005 09:32:29 -0700 (PDT) (envelope-from rra@stanford.edu) Received: from windlord.stanford.edu (windlord.Stanford.EDU [171.64.19.147]) by smtp1.Stanford.EDU (8.12.11/8.12.11) with SMTP id j3TGWTfW032369 for ; Fri, 29 Apr 2005 09:32:29 -0700 Received: (qmail 16732 invoked by uid 1000); 29 Apr 2005 16:32:28 -0000 To: ietf-usefor@imc.org Subject: Re: Obsolete vs. obsolescent In-Reply-To: (Charles Lindsey's message of "Fri, 29 Apr 2005 15:33:26 GMT") References: <87psx1gnjc.fsf@windlord.stanford.edu> <87fyxvx1m6.fsf@windlord.stanford.edu> <87mzs2a1hu.fsf@windlord.stanford.edu> <87wtqxui6s.fsf@windlord.stanford.edu> <87d5so2bpk.fsf@windlord.stanford.edu> <42697BB6.6074@xyzzy.claranet.de> <426BC455.2040108@isode.com> <426E4B45.7010205@oceana.com> <426F4F7C.7385@xyzzy.claranet.de> <87k6moksps.fsf@windlord.stanford.edu> <87br7yg7b7.fsf@windlord.stanford.edu> <427173A8.63D5@xyzzy.claranet.de> <87zmvijwi4.fsf@windlord.stanford.edu> From: Russ Allbery Organization: The Eyrie Date: Fri, 29 Apr 2005 09:32:28 -0700 Message-ID: <87r7gt1r37.fsf@windlord.stanford.edu> User-Agent: Gnus/5.1007 (Gnus v5.10.7) XEmacs/21.4 (Jumbo Shrimp, linux) MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Charles Lindsey writes: > But evidently Frank doesn't want that to be said (he is complaining > about current systems that object when you try to do it), and I don't > think I want to do it either. Better just to let the usage die away > (which USEPRO requires). That isn't going to actually happen in any sort of reasonable time frame. > Otherwise we shall find new software coming on stream that objects to > 'cmsg' in Subjects (just to comply with the new USEFOR) just as the last > remaining servers that might have interpreted that 'cmsg' are being > withdrawn. I wonder if any of us will live long enough to see the last remaining servers that interpreted cmsg be withdrawn. > How likely is it that people will 'accidentally' write Subject headers > starting with 'cmsg'? This doesn't seem like a relevant question to me. Standards aren't concerned only with things that can happen accidentally. > How much of a problem are such headers causing at the present time? They aren't propagated. They have huge interoperability problems at the present time. > And if they do write such a header (without any intent to cause a > Control action), how likely is it that it will be meaningful enough to > cause some untoward effect? That's not the breakage that I'm worried about (control messages should be authenticated anyway instead of cancels, and it's pretty hard to write a cancel control message accidentally). The article will mysteriously disappear, which is a different kind of breakage. > If it ain't actually broke at the present time, then don't waste time > trying to fix it. It's very broken at the present time. -- Russ Allbery (rra@stanford.edu) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3TGSCUP031320 for ; Fri, 29 Apr 2005 09:28:12 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3TGSCO9031319 for ietf-usefor-skb; Fri, 29 Apr 2005 09:28:12 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtp2.Stanford.EDU (smtp2.Stanford.EDU [171.67.16.125]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3TGSBhi031311 for ; Fri, 29 Apr 2005 09:28:11 -0700 (PDT) (envelope-from rra@stanford.edu) Received: from windlord.stanford.edu (windlord.Stanford.EDU [171.64.19.147]) by smtp2.Stanford.EDU (8.12.11/8.12.11) with SMTP id j3TGSBoe030840 for ; Fri, 29 Apr 2005 09:28:11 -0700 Received: (qmail 16647 invoked by uid 1000); 29 Apr 2005 16:28:10 -0000 To: ietf-usefor@imc.org Subject: Re: Suggested References texts In-Reply-To: (Charles Lindsey's message of "Fri, 29 Apr 2005 15:18:12 GMT") References: <426F443C.6561@xyzzy.claranet.de> <42716C2A.4F53@xyzzy.claranet.de> <871x8ubhim.fsf@windlord.stanford.edu> From: Russ Allbery Organization: The Eyrie Date: Fri, 29 Apr 2005 09:28:10 -0700 Message-ID: <87vf651rad.fsf@windlord.stanford.edu> User-Agent: Gnus/5.1007 (Gnus v5.10.7) XEmacs/21.4 (Jumbo Shrimp, linux) MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Charles Lindsey writes: > Russ Allbery writes: >> I *don't* want to give the impression, even in USEFOR, that all the >> rules in RFC 2822 for replies apply unmodified to followups. Replies >> and followups *aren't* the same thing, and I think it's worth making >> that clear. Among other things, correctly constructing a followup >> requires dealing with fields that aren't in RFC 2822 at all. > Eh? Which fields might those be? Followup-To and Newsgroups. -- Russ Allbery (rra@stanford.edu) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3TGECwO028386 for ; Fri, 29 Apr 2005 09:14:12 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3TGECjq028385 for ietf-usefor-skb; Fri, 29 Apr 2005 09:14:12 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from lon-mail-3.gradwell.net (lon-mail-3.gradwell.net [193.111.201.127]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3TGEB7H028374 for ; Fri, 29 Apr 2005 09:14:12 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from host81-144-75-3.midband.mdip.bt.net ([81.144.75.3]) by lon-mail-3.gradwell.net with esmtp (Gradwell gwh-smtpd 1.181) id 42725d52.17ebc.215 for ietf-usefor@imc.org; Fri, 29 Apr 2005 17:14:10 +0100 (envelope-sender ) Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3TGCZc26915 for ietf-usefor@imc.org; Fri, 29 Apr 2005 17:12:35 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20786 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: Suggested References texts Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: <426F443C.6561@xyzzy.claranet.de> <42716C2A.4F53@xyzzy.claranet.de> Date: Fri, 29 Apr 2005 15:14:24 GMT Lines: 100 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In <42716C2A.4F53@xyzzy.claranet.de> Frank Ellermann writes: >I just don't want any additional wording in usefor-04, it's >one of these cases where each word is too much, unless you >plan to integrate the 2004 discussions with the complete >history of this header field in RfCs, news, mail, and UAs >as told by $number_of_USEFOR_participants_since_99. It's like an attempt to define spam with all exceptions and >details, it's hopeless. You really tried. Now let's please >just get away with the bare minimum in usefor-04, for the >technical details usepro-03 should be good enough. But usepro-03 doesn't provide those technical details. It tells you how to construct a followup to an article that is already in front of you (and RFC 2822 does the same for email replies, but is a bit more complicated because it has to allow for cases where References headers were not provided - that "SHOULD" - and it deals with In-Reply-To as well). But suppose you are a multipart FAQ writer trying to use References to refer back to the earlier parts, how do you know what you can do? Or you are writing a threading agent. How do you know what you can rely on in a properly formed References header? If all you are given is the bare syntax (which is what you seem to suggest), then what is to stop you including s of articles that never existed? What is to stop you listing s in the wrong order? Where is the wording to say these things are wrong? Now indeed, most FAQ writers are clueful enough to construct sensible References headers, but in fact they are merely relying on "folklore" - and the whole purpose of standards is to replace folklore by somethng with more precision. Or else they have to "reverse engineer" the algorithm given for followups in USEPRO, and deduce the underlying semantics from that (with the danger of reading more into the algorithm than is strictly justified). Likewise for those who write threading agents. >> So you want me to omit >[... some text about precursors in usefor-04 ...] >Yes, that's what I didn't want because it's essentially the >same as in 2822, but now that you ask I'm not so sure about: >>| A given message identifier MUST NOT appear more than once. >Is that a problem anywhere, bad enough for a MUST NOT ? It's >obviously stupid, but is it potentially harmful ? Yes, that sentence might go. >> that is the semantics (meaning) of the header, which is a >> proper matter for USEFOR. >Not if it's essentially the same as in 2822. Or that's what I >thought the plan is: Anything in RfC 2822 (excl. obs-stuff) >that's not explicity changed in usefor-03/04 is implicitly a >part of usefor-03/04. But RFC 2888 doesn't explain the _meaning_ of the References header. It just tells you how to contruct one when creating a Reply, which is not what we want to use it for. >That's why we don't explain mailbox, addr-spec, From:, Sender:, >etc. Indeed, because RFC 2822 explains all that is needed for those. >Unless you start again with message/partial and other oddities >I don't see where the semantics of "normal" References: is so >different from RfC 2822 that it needs an explanation. Exactly. It is the "other oddities" which necessitate some extra explanation. >> It tells reading agents just what they can deduce from the >> header >If developers of reading agents need more info they can find it >where deelopers of followup-agents find it, in usepro. What's >so special for news reading-agents, in relation to RfC 2822 ? They are going to encounter References headers in articles that are not followups (or replies in the 2822 sense). So there is nothing in USEPRO to help them (except guesswork based on what followup agents do). -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3TGEBRF028376 for ; Fri, 29 Apr 2005 09:14:11 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3TGEBcb028375 for ietf-usefor-skb; Fri, 29 Apr 2005 09:14:11 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from lon-mail-3.gradwell.net (lon-mail-3.gradwell.net [193.111.201.127]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3TGEAF3028364 for ; Fri, 29 Apr 2005 09:14:10 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from host81-144-75-3.midband.mdip.bt.net ([81.144.75.3]) by lon-mail-3.gradwell.net with esmtp (Gradwell gwh-smtpd 1.181) id 42725d51.17ebc.214 for ietf-usefor@imc.org; Fri, 29 Apr 2005 17:14:09 +0100 (envelope-sender ) Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3TGCdc26926 for ietf-usefor@imc.org; Fri, 29 Apr 2005 17:12:39 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20788 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: Obsolete vs. obsolescent Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: <200504111212.j3BCChM14516@clerew.man.ac.uk> <87psx1gnjc.fsf@windlord.stanford.edu> <87fyxvx1m6.fsf@windlord.stanford.edu> <87mzs2a1hu.fsf@windlord.stanford.edu> <87wtqxui6s.fsf@windlord.stanford.edu> <87d5so2bpk.fsf@windlord.stanford.edu> <42697BB6.6074@xyzzy.claranet.de> <426BC455.2040108@isode.com> <426E4B45.7010205@oceana.com> <426F4F7C.7385@xyzzy.claranet.de> <87k6moksps.fsf@windlord.stanford.edu> <87br7yg7b7.fsf@windlord.stanford.edu> <427173A8.63D5@xyzzy.claranet.de> <87zmvijwi4.fsf@windlord.stanford.edu> Date: Fri, 29 Apr 2005 15:33:26 GMT Lines: 53 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In <87zmvijwi4.fsf@windlord.stanford.edu> Russ Allbery writes: >Frank Ellermann writes: >> Okay, but it's not nice when it has serious problems with the >> propagation. An error message "administrative restriction" >> is also not exactly clear. Or if we'd expect gateways to do >> something special, or better nothing unless they are forced >> to bypass an "administrative restriction" we have to say so >> somewhere. Not in usefor-x, that was probably a bad idea. >A subject starting with "cmsg " needs to be prohibited in anything that >isn't a control message. I don't feel strongly about whether that should >be in USEFOR or USEPRO, although to me it feels like a format issue, but >it needs to be in there somewhere. If you want to say that, then USEFOR is clearly the right place to say it. But evidently Frank doesn't want that to be said (he is complaining about current systems that object when you try to do it), and I don't think I want to do it either. Better just to let the usage die away (which USEPRO requires). Otherwise we shall find new software coming on stream that objects to 'cmsg' in Subjects (just to comply with the new USEFOR) just as the last remaining servers that might have interpreted that 'cmsg' are being withdrawn. Eventually, we shall have wide deployment of a mandatory requirement for Subject header which is there to solve a problem that no longer exists. > It's premature to assume that cmsg no >longer receives special handling; we're not done with that transition. We >might never be, frankly. Using cmsg in the Subject header of regular >posts will cause significant interoperability problems. How likely is it that people will 'accidentally' write Subject headers starting with 'cmsg'? How much of a problem are such headers causing at the present time? And if they do write such a header (without any intent to cause a Control action), how likely is it that it will be meaningful enough to cause some untoward effect? If it ain't actually broke at the present time, then don't waste time trying to fix it. But if the WG really wants this prohibition including, then let it say so, in which case USEFOR is the place to do it. -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3TGEAh9028366 for ; Fri, 29 Apr 2005 09:14:10 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3TGEArp028365 for ietf-usefor-skb; Fri, 29 Apr 2005 09:14:10 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from lon-mail-3.gradwell.net (lon-mail-3.gradwell.net [193.111.201.127]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3TGE95u028342 for ; Fri, 29 Apr 2005 09:14:09 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from host81-144-75-3.midband.mdip.bt.net ([81.144.75.3]) by lon-mail-3.gradwell.net with esmtp (Gradwell gwh-smtpd 1.181) id 42725d50.17ebc.213 for ietf-usefor@imc.org; Fri, 29 Apr 2005 17:14:08 +0100 (envelope-sender ) Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3TGCg426935 for ietf-usefor@imc.org; Fri, 29 Apr 2005 17:12:42 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20790 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: Again broken Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: <4255153E.3080500@isode.com> <42594879.3B92@xyzzy.claranet.de> <425A8A63.268A@xyzzy.claranet.de> <425C607E.1B5A@xyzzy.claranet.de> <425D8031.D57@xyzzy.claranet.de> <426BB63D.3060506@isode.com> <426F6107.666C@xyzzy.claranet.de> <427192C3.301E@xyzzy.claranet.de> Date: Fri, 29 Apr 2005 16:08:38 GMT Lines: 56 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In <427192C3.301E@xyzzy.claranet.de> Frank Ellermann writes: >Charles Lindsey wrote: >> , and are all defined in RFC 2045 >Yes. So far I didn't know that usefor-03 implicitly includes >all 2045 terms. I only knew that it includes 2822 minus obs, >and of course 2234bis. It needs to be stated somewhere. >2045 is older than 2234. IMHO it's better to copy (and check) >what you need, instead of implicitly importing it. Checking: > token := 1* or tspecials> >Whatever that is, it's no proper ABNF. So why did Keith do it >in 3834, Because that is exactly what RFC 2045 says. It even appeared (but not normatively) in my old draft-13. But the real problems are and . In draft-13, I actually had (simplifying slightly, because there were also RFC 2231 complications): parameter = [CFWS] token [CFWS] "=" value value = [CFWS] token [CFWS] / quoted-string (again, not normatively) But I was forbidden from having that stuff incorporated into USEFOR. As a result, USEFOR does not adequately explain where [CFWS] can appear in Injection-Info. >> There is a problem in RFC 2822 (which you yourself spotted) >A _real_ problem in 2822 ? So far I tended to hallucinate >problems, but the worst I'm sure of was an "I don't like it". OK, if you don't want it fixed, let it go. But just using in place of will not fix it either, without further verbiage. -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3TGE9Lj028356 for ; Fri, 29 Apr 2005 09:14:09 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3TGE9kY028354 for ietf-usefor-skb; Fri, 29 Apr 2005 09:14:09 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from lon-mail-3.gradwell.net (lon-mail-3.gradwell.net [193.111.201.127]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3TGE8Yh028340 for ; Fri, 29 Apr 2005 09:14:08 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from host81-144-75-3.midband.mdip.bt.net ([81.144.75.3]) by lon-mail-3.gradwell.net with esmtp (Gradwell gwh-smtpd 1.181) id 42725d4e.17ebc.211 for ietf-usefor@imc.org; Fri, 29 Apr 2005 17:14:06 +0100 (envelope-sender ) Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3TGCbM26921 for ietf-usefor@imc.org; Fri, 29 Apr 2005 17:12:37 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20787 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: Suggested References texts Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: <426F443C.6561@xyzzy.claranet.de> <42716C2A.4F53@xyzzy.claranet.de> <871x8ubhim.fsf@windlord.stanford.edu> Date: Fri, 29 Apr 2005 15:18:12 GMT Lines: 20 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In <871x8ubhim.fsf@windlord.stanford.edu> Russ Allbery writes: >I *don't* want to give the impression, even in USEFOR, that all the rules >in RFC 2822 for replies apply unmodified to followups. Replies and >followups *aren't* the same thing, and I think it's worth making that >clear. Among other things, correctly constructing a followup requires >dealing with fields that aren't in RFC 2822 at all. Eh? Which fields might those be? -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3TGE9df028355 for ; Fri, 29 Apr 2005 09:14:09 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3TGE9S0028353 for ietf-usefor-skb; Fri, 29 Apr 2005 09:14:09 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from lon-mail-3.gradwell.net (lon-mail-3.gradwell.net [193.111.201.127]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3TGE8wt028341 for ; Fri, 29 Apr 2005 09:14:08 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from host81-144-75-3.midband.mdip.bt.net ([81.144.75.3]) by lon-mail-3.gradwell.net with esmtp (Gradwell gwh-smtpd 1.181) id 42725d4f.17ebc.212 for ietf-usefor@imc.org; Fri, 29 Apr 2005 17:14:07 +0100 (envelope-sender ) Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3TGCeD26930 for ietf-usefor@imc.org; Fri, 29 Apr 2005 17:12:40 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20789 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: Again broken Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: <4255153E.3080500@isode.com> <42594879.3B92@xyzzy.claranet.de> <425A8A63.268A@xyzzy.claranet.de> <425C607E.1B5A@xyzzy.claranet.de> <425D8031.D57@xyzzy.claranet.de> <426BB63D.3060506@isode.com> <426F6107.666C@xyzzy.claranet.de> <427192C3.301E@xyzzy.claranet.de> <87sm1apchr.fsf@windlord.stanford.edu> Date: Fri, 29 Apr 2005 15:48:22 GMT Lines: 39 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In <87sm1apchr.fsf@windlord.stanford.edu> Russ Allbery writes: >RFC 2822 doesn't say anything about Control headers, for obvious reasons. >What are we gaining by adding comments to the header and breaking existing >software? Consistency with conventions for headers elsewhere. Users don't want to have to remember which headers allow comments and which don't (except where there are strong and permanent reasons to the contrary, such as performance issues that are never going to go away). Sooner or later, some enterprising person is going to write Control: cancel <12345@example.com> (unapproved article in moderated group) and will be most miffed if it is rejected or doesn't work. Yes, it may not work with many current servers, but there is no reason why it should not work long term. It is not rocket science to preprocess all headers to remove all comments and foldings before acting upon them. Anyway, the WG has never asked for this particular comment to be forbidden (and it _did_ decide very early on that comments were normally to be accepted everywhere, but maybe not generated "yet"). It was not forbidden in any of the earlier drafts, so it is premature for USEFOR to have removed it. If the WG wants this change, then it ought to be discussed first. -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3T21cd4051310 for ; Thu, 28 Apr 2005 19:01:38 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3T21cep051309 for ietf-usefor-skb; Thu, 28 Apr 2005 19:01:38 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtp1.Stanford.EDU (smtp1.Stanford.EDU [171.67.16.123]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3T21bD6051302 for ; Thu, 28 Apr 2005 19:01:37 -0700 (PDT) (envelope-from rra@stanford.edu) Received: from windlord.stanford.edu (windlord.Stanford.EDU [171.64.19.147]) by smtp1.Stanford.EDU (8.12.11/8.12.11) with SMTP id j3T21aTx016233 for ; Thu, 28 Apr 2005 19:01:36 -0700 Received: (qmail 19657 invoked by uid 1000); 29 Apr 2005 02:01:36 -0000 To: ietf-usefor@imc.org Subject: Re: Again broken In-Reply-To: <427192C3.301E@xyzzy.claranet.de> (Frank Ellermann's message of "Fri, 29 Apr 2005 03:49:55 +0200") References: <4255153E.3080500@isode.com> <42594879.3B92@xyzzy.claranet.de> <425A8A63.268A@xyzzy.claranet.de> <425C607E.1B5A@xyzzy.claranet.de> <425D8031.D57@xyzzy.claranet.de> <426BB63D.3060506@isode.com> <426F6107.666C@xyzzy.claranet.de> <427192C3.301E@xyzzy.claranet.de> From: Russ Allbery Organization: The Eyrie Date: Thu, 28 Apr 2005 19:01:36 -0700 Message-ID: <87sm1apchr.fsf@windlord.stanford.edu> User-Agent: Gnus/5.1007 (Gnus v5.10.7) XEmacs/21.4 (Jumbo Shrimp, linux) MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Frank Ellermann writes: > Charles Lindsey wrote: >> The WG agreed early on that comments should be acceptable >> in all places allowed by RFC 2822 > Especially where stupid users do stupid things, but Control ? >> Control headers are so rare that performance issues can be >> neglected. > Okay, if you say so. Convince Russ that it's a good idea. RFC 2822 doesn't say anything about Control headers, for obvious reasons. What are we gaining by adding comments to the header and breaking existing software? -- Russ Allbery (rra@stanford.edu) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3T1t6X5050389 for ; Thu, 28 Apr 2005 18:55:06 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3T1t69Y050388 for ietf-usefor-skb; Thu, 28 Apr 2005 18:55:06 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from ciao.gmane.org (main.gmane.org [80.91.229.2]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3T1t4Ir050381 for ; Thu, 28 Apr 2005 18:55:05 -0700 (PDT) (envelope-from usenet-format@gmane.org) Received: from list by ciao.gmane.org with local (Exim 4.43) id 1DRKcZ-0002Jm-1j for ietf-usefor@imc.org; Fri, 29 Apr 2005 03:48:47 +0200 Received: from 62.80.58.28 ([62.80.58.28]) by main.gmane.org with esmtp (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Fri, 29 Apr 2005 03:48:47 +0200 Received: from nobody by 62.80.58.28 with local (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Fri, 29 Apr 2005 03:48:47 +0200 X-Injected-Via-Gmane: http://gmane.org/ To: ietf-usefor@imc.org From: Frank Ellermann Subject: Re: Again broken Date: Fri, 29 Apr 2005 03:49:55 +0200 Organization: Lines: 112 Message-ID: <427192C3.301E@xyzzy.claranet.de> References: <4255153E.3080500@isode.com> <42594879.3B92@xyzzy.claranet.de> <425A8A63.268A@xyzzy.claranet.de> <425C607E.1B5A@xyzzy.claranet.de> <425D8031.D57@xyzzy.claranet.de> <426BB63D.3060506@isode.com> <426F6107.666C@xyzzy.claranet.de> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Complaints-To: usenet@sea.gmane.org X-Gmane-NNTP-Posting-Host: 62.80.58.28 X-Mailer: Mozilla 3.0 (OS/2; U) Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Charles Lindsey wrote: >> Is that as you want it ? > Yes. Do you see any technical error in it? No, I just wanted to be sure that I can quote it as is. > I will be trying to persuade them into adopting our > syntax, or preferebaly some less bizarre subset of it. The excluded ">" is bizarre, but allowing it elsewhere would be a pain for gateways. This syntax _is_ the Message-ID for mail and news, bizarre or not. I thought about putting it into an I-D, "updates 2822", ready, problem solved. >> Control: cancel > Hold on! You aren't allowed to cancel more than one > article at a time. Sorry, I wrote it after the very similar References issue. > A. The FWS should be CFWS in all places. I thought the server folks don't want CFWS in their headers, because they don't want to waste time to parse comments. > The WG agreed early on that comments should be acceptable > in all places allowed by RFC 2822 Especially where stupid users do stupid things, but Control ? > Control headers are so rare that performance issues can be > neglected. Okay, if you say so. Convince Russ that it's a good idea. > control = "Control:" SP [CFWS] control-command [CFWS] CRLF Often you have the leading / trailing [CFWS] in the "inner" rule with shorter lines, i.e. control = "Control:" SP control-command CRLF control-command = [CFWS] verb *( CFWS argument ) [CFWS} Of course irrelevant; I only see it when I try to quote it. > On top of that, a "do not generate comments yet" wording > would be in order, as for the References header If you think that comments are stupid, and they are, and only needed for users who trust that the rules are everywhere the same, then using [FWS] and FWS without any additional wording is much clearer than [CFWS] and CFWS with a rule to pray not yet use comments in an "administrative" header like Control. > I would prefer a generic statement to that effect covering > all headers inherited from RFC 1036. That's a trend, you love to say A in syntax, and then to add long qualifications which are near to NOT A. I'd just say NOT A in syntax without qualification, and be done with it. > , and are all defined in RFC 2045 Yes. So far I didn't know that usefor-03 implicitly includes all 2045 terms. I only knew that it includes 2822 minus obs, and of course 2234bis. 2045 is older than 2234. IMHO it's better to copy (and check) what you need, instead of implicitly importing it. Checking: token := 1* Whatever that is, it's no proper ABNF. So why did Keith do it in 3834, Bruce must have seen this, I'm stunned. 2231 doesn't fix it. We shouldn't import this quasi-ABNF, even if Bruce promises to let it pass. > There is a problem in RFC 2822 (which you yourself spotted) A _real_ problem in 2822 ? So far I tended to hallucinate problems, but the worst I'm sure of was an "I don't like it". > If you still want this hole plugged, then please propose an > alternative text. That's simple, just say "id-domain" and "id-literal" _without_ additional wording. Or say "address-literal". Not the unclear "id-right" and "no-fold-literal" stuff, where nobody knows what it's supposed to mean. Or add a note with this effect: Note: id-right is a domain, and no-fold-literal is the special case of a domain-literal (a.k.a. address-literal, see RfC 2821) >>| Note: THE USE OF DOMAIN-LITERALS IS STRONGLY DISCOURAGED. > Yes, but there is more to say than that, because even a > is discouraged (does RFC 2822 still say > that?) you want even more discouragement for other than IP > addresses inside them, and you still need to extend it to > s. More to say ? That statement is perfectly clear, and we have serious technical reasons why we want it. No, AFAIK RfC 2882 does not say so, otherwise we wouldn't need it in usefor-xx. Bye, Frank Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3T07mfq035067 for ; Thu, 28 Apr 2005 17:07:48 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3T07mD2035066 for ietf-usefor-skb; Thu, 28 Apr 2005 17:07:48 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from ciao.gmane.org (main.gmane.org [80.91.229.2]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3T07k1E035052 for ; Thu, 28 Apr 2005 17:07:46 -0700 (PDT) (envelope-from usenet-format@gmane.org) Received: from list by ciao.gmane.org with local (Exim 4.43) id 1DRIxA-0002cb-U1 for ietf-usefor@imc.org; Fri, 29 Apr 2005 02:01:56 +0200 Received: from 62.80.58.28 ([62.80.58.28]) by main.gmane.org with esmtp (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Fri, 29 Apr 2005 02:01:56 +0200 Received: from nobody by 62.80.58.28 with local (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Fri, 29 Apr 2005 02:01:56 +0200 X-Injected-Via-Gmane: http://gmane.org/ To: ietf-usefor@imc.org From: Frank Ellermann Subject: Re: Fixed Date: Fri, 29 Apr 2005 02:01:51 +0200 Organization: Lines: 23 Message-ID: <4271796F.2884@xyzzy.claranet.de> References: <4255153E.3080500@isode.com> <42594879.3B92@xyzzy.claranet.de> <425A8A63.268A@xyzzy.claranet.de> <425C607E.1B5A@xyzzy.claranet.de> <425D8031.D57@xyzzy.claranet.de> <426BB63D.3060506@isode.com> <426F6AC8.5045@xyzzy.claranet.de> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Complaints-To: usenet@sea.gmane.org X-Gmane-NNTP-Posting-Host: 62.80.58.28 X-Mailer: Mozilla 3.0 (OS/2; U) Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Charles Lindsey wrote: > But you asked for someone to come along and provide a > casting vote True. I wasn't happy with how Alexey did it, no indication that he understood why we agree on a syntax but disagree on the stupid names of the productions. As chair he's supposed to handle such obscure problems with tact. > You are the only person wanting @. That's not the point, I'm the person who seriously does not like "id-right", I think that your idea of "id-right" is not the "id-domain" in my last proposal. If you like to reopen negotiations, you can have "id-left". The critical part for Usenet is the RHS with a domain. The LHS is then only a problem of *the* domain owner. For your "id-right" that's very unclear. Bye, Frank Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3SNjvvH032588 for ; Thu, 28 Apr 2005 16:45:57 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3SNjvrS032587 for ietf-usefor-skb; Thu, 28 Apr 2005 16:45:57 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtp2.Stanford.EDU (smtp2.Stanford.EDU [171.67.16.125]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3SNjumh032577 for ; Thu, 28 Apr 2005 16:45:56 -0700 (PDT) (envelope-from rra@stanford.edu) Received: from windlord.stanford.edu (windlord.Stanford.EDU [171.64.19.147]) by smtp2.Stanford.EDU (8.12.11/8.12.11) with SMTP id j3SNjtDf014386 for ; Thu, 28 Apr 2005 16:45:56 -0700 Received: (qmail 14800 invoked by uid 1000); 28 Apr 2005 23:45:55 -0000 To: ietf-usefor@imc.org Subject: Re: Obsolete vs. obsolescent In-Reply-To: <427173A8.63D5@xyzzy.claranet.de> (Frank Ellermann's message of "Fri, 29 Apr 2005 01:37:12 +0200") References: <200504111212.j3BCChM14516@clerew.man.ac.uk> <87psx1gnjc.fsf@windlord.stanford.edu> <87fyxvx1m6.fsf@windlord.stanford.edu> <87mzs2a1hu.fsf@windlord.stanford.edu> <87wtqxui6s.fsf@windlord.stanford.edu> <87d5so2bpk.fsf@windlord.stanford.edu> <42697BB6.6074@xyzzy.claranet.de> <426BC455.2040108@isode.com> <426E4B45.7010205@oceana.com> <426F4F7C.7385@xyzzy.claranet.de> <87k6moksps.fsf@windlord.stanford.edu> <87br7yg7b7.fsf@windlord.stanford.edu> <427173A8.63D5@xyzzy.claranet.de> From: Russ Allbery Organization: The Eyrie Date: Thu, 28 Apr 2005 16:45:55 -0700 Message-ID: <87zmvijwi4.fsf@windlord.stanford.edu> User-Agent: Gnus/5.1007 (Gnus v5.10.7) XEmacs/21.4 (Jumbo Shrimp, linux) MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Frank Ellermann writes: > Okay, but it's not nice when it has serious problems with the > propagation. An error message "administrative restriction" > is also not exactly clear. Or if we'd expect gateways to do > something special, or better nothing unless they are forced > to bypass an "administrative restriction" we have to say so > somewhere. Not in usefor-x, that was probably a bad idea. A subject starting with "cmsg " needs to be prohibited in anything that isn't a control message. I don't feel strongly about whether that should be in USEFOR or USEPRO, although to me it feels like a format issue, but it needs to be in there somewhere. It's premature to assume that cmsg no longer receives special handling; we're not done with that transition. We might never be, frankly. Using cmsg in the Subject header of regular posts will cause significant interoperability problems. > And they won't stop this immediately only because a new RfC > wants it. Let's forget Also-Control unless you're sure that > it could be a similar problem. Also-Control simply SHOULD NOT be used, but that's more USEPRO. -- Russ Allbery (rra@stanford.edu) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3SNcmhP030975 for ; Thu, 28 Apr 2005 16:38:48 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3SNcmXk030974 for ietf-usefor-skb; Thu, 28 Apr 2005 16:38:48 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from ciao.gmane.org (main.gmane.org [80.91.229.2]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3SNckJn030968 for ; Thu, 28 Apr 2005 16:38:47 -0700 (PDT) (envelope-from usenet-format@gmane.org) Received: from list by ciao.gmane.org with local (Exim 4.43) id 1DRIUy-0008Pe-JA for ietf-usefor@imc.org; Fri, 29 Apr 2005 01:32:48 +0200 Received: from 62.80.58.28 ([62.80.58.28]) by main.gmane.org with esmtp (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Fri, 29 Apr 2005 01:32:48 +0200 Received: from nobody by 62.80.58.28 with local (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Fri, 29 Apr 2005 01:32:48 +0200 X-Injected-Via-Gmane: http://gmane.org/ To: ietf-usefor@imc.org From: Frank Ellermann Subject: Re: Obsolete vs. obsolescent Date: Fri, 29 Apr 2005 01:37:12 +0200 Organization: Lines: 25 Message-ID: <427173A8.63D5@xyzzy.claranet.de> References: <200504111212.j3BCChM14516@clerew.man.ac.uk> <87psx1gnjc.fsf@windlord.stanford.edu> <87fyxvx1m6.fsf@windlord.stanford.edu> <87mzs2a1hu.fsf@windlord.stanford.edu> <87wtqxui6s.fsf@windlord.stanford.edu> <87d5so2bpk.fsf@windlord.stanford.edu> <42697BB6.6074@xyzzy.claranet.de> <426BC455.2040108@isode.com> <426E4B45.7010205@oceana.com> <426F4F7C.7385@xyzzy.claranet.de> <87k6moksps.fsf@windlord.stanford.edu> <87br7yg7b7.fsf@windlord.stanford.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Complaints-To: usenet@sea.gmane.org X-Gmane-NNTP-Posting-Host: 62.80.58.28 X-Mailer: Mozilla 3.0 (OS/2; U) Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Russ Allbery wrote: [cmsg] > I don't feel too strongly about it, though. Okay, but it's not nice when it has serious problems with the propagation. An error message "administrative restriction" is also not exactly clear. Or if we'd expect gateways to do something special, or better nothing unless they are forced to bypass an "administrative restriction" we have to say so somewhere. Not in usefor-x, that was probably a bad idea. But the usepro-03 text pretends that there's actually no problem at all with Subject: cmsg. And that's not the case, The attempts to avoid obsolete problems are the problem - not some old servers trying to act on Subject: cmsg, but the many servers trying to protect old servers are a pain. And they won't stop this immediately only because a new RfC wants it. Let's forget Also-Control unless you're sure that it could be a similar problem. Bye, Frank Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3SNape4030708 for ; Thu, 28 Apr 2005 16:36:51 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3SNapH4030707 for ietf-usefor-skb; Thu, 28 Apr 2005 16:36:51 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtp1.Stanford.EDU (smtp1.Stanford.EDU [171.67.16.123]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3SNaoDA030687 for ; Thu, 28 Apr 2005 16:36:50 -0700 (PDT) (envelope-from rra@stanford.edu) Received: from windlord.stanford.edu (windlord.Stanford.EDU [171.64.19.147]) by smtp1.Stanford.EDU (8.12.11/8.12.11) with SMTP id j3SNan6V012490 for ; Thu, 28 Apr 2005 16:36:49 -0700 Received: (qmail 14645 invoked by uid 1000); 28 Apr 2005 23:36:49 -0000 To: ietf-usefor@imc.org Subject: Re: Suggested References texts In-Reply-To: <42716C2A.4F53@xyzzy.claranet.de> (Frank Ellermann's message of "Fri, 29 Apr 2005 01:05:14 +0200") References: <426F443C.6561@xyzzy.claranet.de> <42716C2A.4F53@xyzzy.claranet.de> From: Russ Allbery Organization: The Eyrie Date: Thu, 28 Apr 2005 16:36:49 -0700 Message-ID: <871x8ubhim.fsf@windlord.stanford.edu> User-Agent: Gnus/5.1007 (Gnus v5.10.7) XEmacs/21.4 (Jumbo Shrimp, linux) MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Frank Ellermann writes: > I just don't want any additional wording in usefor-04, it's > one of these cases where each word is too much, unless you > plan to integrate the 2004 discussions with the complete > history of this header field in RfCs, news, mail, and UAs > as told by $number_of_USEFOR_participants_since_99. Not if it's essentially the same as in 2822. Or that's what I > thought the plan is: Anything in RfC 2822 (excl. obs-stuff) > that's not explicity changed in usefor-03/04 is implicitly a > part of usefor-03/04. > That's why we don't explain mailbox, addr-spec, From:, Sender:, > etc. We only explain what's different, e.g. the msg-id syntax > or something with the Date: timezone. And of course all header > fields found only in news, among others Followup-To:. > Unless you start again with message/partial and other oddities > I don't see where the semantics of "normal" References: is so > different from RfC 2822 that it needs an explanation. > Sure, we don't have In-Reply-To, that's a minor difference, UAs > aren't expected to derive missing References: from In-Reply-To: > But that's an issue for usepro, not usefor, like trimming etc. The only thing that I can see mentioning, given this whole discussion, is saying that References is required for followups, although I don't think we should say much more than RFC 1036 does in that area. For the rest, I agree with you wholeheartedly. I *don't* want to give the impression, even in USEFOR, that all the rules in RFC 2822 for replies apply unmodified to followups. Replies and followups *aren't* the same thing, and I think it's worth making that clear. Among other things, correctly constructing a followup requires dealing with fields that aren't in RFC 2822 at all. >> I think we agreed earlier that the actual trimming details >> belonged in USEAGE. If we want to strongly encourage (SHOULD) a particular trimming strategy, that goes into USEPRO as it's part of the protocol. Given the number of bad trimming strategies that people use that cause interoperability problems, I think this is probably a good idea. > Quite possible, that's something between you and Alexey, this > split still confuses me. But at least I got the idea what's > usefor-xx and what not: usefor-xx is only the article format. > Not all why-s, just the minimum to describe what's actually > transmitted. Form but not function. Plus basic definitions. Yup. -- Russ Allbery (rra@stanford.edu) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3SNG3mF027787 for ; Thu, 28 Apr 2005 16:16:03 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3SNG3kt027786 for ietf-usefor-skb; Thu, 28 Apr 2005 16:16:03 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from ciao.gmane.org (main.gmane.org [80.91.229.2]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3SNG0Du027780 for ; Thu, 28 Apr 2005 16:16:01 -0700 (PDT) (envelope-from usenet-format@gmane.org) Received: from list by ciao.gmane.org with local (Exim 4.43) id 1DRI99-0006R6-A9 for ietf-usefor@imc.org; Fri, 29 Apr 2005 01:10:15 +0200 Received: from 62.80.58.28 ([62.80.58.28]) by main.gmane.org with esmtp (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Fri, 29 Apr 2005 01:10:15 +0200 Received: from nobody by 62.80.58.28 with local (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Fri, 29 Apr 2005 01:10:15 +0200 X-Injected-Via-Gmane: http://gmane.org/ To: ietf-usefor@imc.org From: Frank Ellermann Subject: Re: Suggested References texts Date: Fri, 29 Apr 2005 01:05:14 +0200 Organization: Lines: 85 Message-ID: <42716C2A.4F53@xyzzy.claranet.de> References: <426F443C.6561@xyzzy.claranet.de> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Complaints-To: usenet@sea.gmane.org X-Gmane-NNTP-Posting-Host: 62.80.58.28 X-Mailer: Mozilla 3.0 (OS/2; U) Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Charles Lindsey wrote: > If you want me to limit "precursor" as you suggest, then > please suggest alternative wording for those semantics. I just don't want any additional wording in usefor-04, it's one of these cases where each word is too much, unless you plan to integrate the 2004 discussions with the complete history of this header field in RfCs, news, mail, and UAs as told by $number_of_USEFOR_participants_since_99. > references = "References:" SP msg-id-list CRLF >> msg-id-list = [CFWS] msg-id *( CFWS msg-id ) [CFWS] > or, following the style in draft-13, > references = "References:" SP [CFWS] msg-id 1*(CFWS msg-id ) > [CFWS] CRLF ACK, one production less, OTOH a rather long rule. I thought that you could use the msg-id-list also in Control: cancel or Supersedes:, but that's not the case, our Supersedes: has only one msg-id, and Control: has no CFWS. > So you want me to omit [... some text about precursors in usefor-04 ...] Yes, that's what I didn't want because it's essentially the same as in 2822, but now that you ask I'm not so sure about: >| A given message identifier MUST NOT appear more than once. Is that a problem anywhere, bad enough for a MUST NOT ? It's obviously stupid, but is it potentially harmful ? > that is the semantics (meaning) of the header, which is a > proper matter for USEFOR. Not if it's essentially the same as in 2822. Or that's what I thought the plan is: Anything in RfC 2822 (excl. obs-stuff) that's not explicity changed in usefor-03/04 is implicitly a part of usefor-03/04. That's why we don't explain mailbox, addr-spec, From:, Sender:, etc. We only explain what's different, e.g. the msg-id syntax or something with the Date: timezone. And of course all header fields found only in news, among others Followup-To:. Unless you start again with message/partial and other oddities I don't see where the semantics of "normal" References: is so different from RfC 2822 that it needs an explanation. Sure, we don't have In-Reply-To, that's a minor difference, UAs aren't expected to derive missing References: from In-Reply-To: But that's an issue for usepro, not usefor, like trimming etc. > It tells reading agents just what they can deduce from the > header If developers of reading agents need more info they can find it where deelopers of followup-agents find it, in usepro. What's so special for news reading-agents, in relation to RfC 2822 ? Or do you want some "security considerations" about References: "Beware, the References in news can be wild and wonderful, and for mail it's the same problem. But we wanted to document it explicitly. so take care." Seriously, I don't get it. > I think we agreed earlier that the actual trimming details > belonged in USEAGE. Quite possible, that's something between you and Alexey, this split still confuses me. But at least I got the idea what's usefor-xx and what not: usefor-xx is only the article format. Not all why-s, just the minimum to describe what's actually transmitted. Form but not function. Plus basic definitions. Bye, Frank Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3SH67Zu065540 for ; Thu, 28 Apr 2005 10:06:07 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3SH67mv065539 for ietf-usefor-skb; Thu, 28 Apr 2005 10:06:07 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtp3.Stanford.EDU (smtp3.Stanford.EDU [171.67.16.138]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3SH650w065531 for ; Thu, 28 Apr 2005 10:06:06 -0700 (PDT) (envelope-from rra@stanford.edu) Received: from windlord.stanford.edu (windlord.Stanford.EDU [171.64.19.147]) by smtp3.Stanford.EDU (8.12.11/8.12.11) with SMTP id j3SH65MB008207 for ; Thu, 28 Apr 2005 10:06:05 -0700 Received: (qmail 31887 invoked by uid 1000); 28 Apr 2005 17:06:04 -0000 To: ietf-usefor@imc.org Subject: Re: Obsolete vs. obsolescent In-Reply-To: (Charles Lindsey's message of "Thu, 28 Apr 2005 14:09:23 GMT") References: <200504111212.j3BCChM14516@clerew.man.ac.uk> <87psx1gnjc.fsf@windlord.stanford.edu> <87fyxvx1m6.fsf@windlord.stanford.edu> <87mzs2a1hu.fsf@windlord.stanford.edu> <87wtqxui6s.fsf@windlord.stanford.edu> <87d5so2bpk.fsf@windlord.stanford.edu> <42697BB6.6074@xyzzy.claranet.de> <426BC455.2040108@isode.com> <426E4B45.7010205@oceana.com> <426F4F7C.7385@xyzzy.claranet.de> <87k6moksps.fsf@windlord.stanford.edu> From: Russ Allbery Organization: The Eyrie Date: Thu, 28 Apr 2005 10:06:04 -0700 Message-ID: <87br7yg7b7.fsf@windlord.stanford.edu> User-Agent: Gnus/5.1007 (Gnus v5.10.7) XEmacs/21.4 (Jumbo Shrimp, linux) MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Charles Lindsey writes: > Russ Allbery writes: >> I think USEFOR does need an explicit mention that subjects starting >> with cmsg should not be considered control messages, to document the >> change from RFC 1036. > No, I think it is a protocol issue, and the mention in USEPRO is > sufficient. The transformation of a message from a regular news message to a control message because of the presence of five characters at the beginning of the Subject line sounds like a syntax and semantics issue to me, even if we're just warning that it used to be interpreted that way. cmsg changes the semantic meaning of the message entirely, if honored. I don't feel too strongly about it, though. -- Russ Allbery (rra@stanford.edu) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3SGEh6x058335 for ; Thu, 28 Apr 2005 09:14:43 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3SGEhL4058334 for ietf-usefor-skb; Thu, 28 Apr 2005 09:14:43 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from lon-mail-1.gradwell.net (lon-mail-1.gradwell.net [193.111.201.125]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3SGEgum058327 for ; Thu, 28 Apr 2005 09:14:43 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from host81-144-74-47.midband.mdip.bt.net ([81.144.74.47]) by lon-mail-1.gradwell.net with esmtp (Gradwell gwh-smtpd 1.181) id 42710bf1.165c1.5f for ietf-usefor@imc.org; Thu, 28 Apr 2005 17:14:41 +0100 (envelope-sender ) Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3SGCRv17028 for ietf-usefor@imc.org; Thu, 28 Apr 2005 17:12:27 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20771 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: Again broken (was: Fixed) Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: <4255153E.3080500@isode.com> <42594879.3B92@xyzzy.claranet.de> <425A8A63.268A@xyzzy.claranet.de> <425C607E.1B5A@xyzzy.claranet.de> <425D8031.D57@xyzzy.claranet.de> <426BB63D.3060506@isode.com> <426F6107.666C@xyzzy.claranet.de> Date: Thu, 28 Apr 2005 15:03:19 GMT Lines: 155 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In <426F6107.666C@xyzzy.claranet.de> Frank Ellermann writes: >Charles Lindsey wrote: >~~~ begin quoted ABNF ~~~ >message-id = "Message-ID:" SP [FWS] msg-id [FWS] CRLF ........... >~~~ end of quoted sytax ~~~ >Is that as you want it ? Yes. Do you see any technical error in it? > I'm not sure about Alexey's external >expert for 2822bis, so I think the most simple solution to get >an expert review for the two proposals is to post them on the >rfc-822 mailing list. I don't think there is current activity on 2822bis. As soon as there is, I will be trying to persuade them into adopting our syntax, or preferebaly some less bizarre subset of it. Their present msg-id definition should just go into their obs-syntax. But that is a battle for another day and another place. >> The msg-id-core MUST NOT be more than 250 octets in length. >s/msg-id-core/msg-id/g Oops! >>| Also note that this updated ABNF applies wherever is used, >>| including the References header discussed in Section 3.2.1 and the >>| Supersedes header discussed in Section 3.2.5. >IMHO that's obvious and needs no "note", It's in the present text. I would have no worries if Ken decided to take it out. >BTW, in Control: you have the same syntax problem as in the >References:, you separate the arguments by an _optional_ [FWS]: >Control: cancel Hold on! You aren't allowed to cancel more than one article at a time. Yes, I know CNews does it, but this WG decided long ago not to allow it, since most other serving agents don't accept it. >Control: newgroupthis.is.new >Control: rmgroupthis.is.old But yes, you have spotted a real bug there (but fortunately the error is not repeated in the syntax of the actual s in USEPRO). But whilst we are here, there are two other errors in the Control syntax. A. The FWS should be CFWS in all places. The WG agreed early on that comments should be acceptable in all places allowed by RFC 2822, but on a "MUST accept, do not generate yet" basis. Later on, we agreed that they should be disallowed wherevoer there was a severe performance penalty, so they were removed from Newsgroups, Message-ID, Path, Followup-To and Distribution (these were intended to be permanent exclusions, not just until "yet" happened). But it was never agreed for Control, because Control headers are so rare that performance issues can be neglected. B. s/control-message/control-command/ This is to accord with a change I made in USEPRO. The problem is that the term "control message" is already defined to mean the whole article containing a Control header, and to have the syntactic element meaning something different seemed a bad idea. So I believe the complete syntax for Control headers should be: control = "Control:" SP [CFWS] control-command [CFWS] CRLF control-command = verb *( CFWS argument ) verb = token argument = value On top of that, a "do not generate comments yet" wording would be in order, as for the References header, but I would prefer a generic statement to that effect covering all headers inherited from RFC 1036. >... "value" is apparently not >defined in usefor-03 or 2822. "token" is apparently also not >defined in usefor-03 or 2822. , and are all defined in RFC 2045, and used at various places in USEFOR. This needs to be stated somewhere, but there is more to it than that because RFC 2045, being RFC 822 based, does not tell you where to put the [CFWS]s in them, so we need some wording to cover that. Also, in Injection-Info, where it says post-host-param = "posting-host" "=" host-value it actually means post-host-param = [CFWS] "posting-host" [CFWS] "=" [CFWS] host-value [CFWS] and similarly for the other <*-param>s. So that needs to be covered by explicit syntax or suitable verbiage. >>| NOTE: It in RECOMMENDED in [RFC 2822] that, for ensuring >>| global uniqueness, the be some domain identifier >It's unnecessary to quote that RECOMMENDATION, least of all the >"some domain identifier". Out of its original very convoluted >RfC 2822 context this quote is in fact _worse_ than RfC 2822. There is a problem in RFC 2822 (which you yourself spotted) in that there is wording in 3.4.1 to the effect that a is meant to be an IP address (whether IPv4 or IPv6 or any other IPvxx), in spite of a syntax apparently allowing much more, but they omitted to say the same thing about a . The text I wrote was intended to plug this whole, but it was only a NOTE, so had no normative effect. If you still want this hole plugged, then please propose an alternative text. >IIRC we had a "pseudo-consensus" (= you and me) about domain- >literals in addresses, Message-IDs, and Path-identifiers, i.e. >say something like STD 11 without shouting: >| Note: THE USE OF DOMAIN-LITERALS IS STRONGLY DISCOURAGED. Yes, but there is more to say than that, because even a is discouraged (does RFC 2822 still say that?) you want even more discouragement for other than IP addresses inside them, and you still need to extend it to s. -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3SGEgv0058326 for ; Thu, 28 Apr 2005 09:14:42 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3SGEgFV058325 for ietf-usefor-skb; Thu, 28 Apr 2005 09:14:42 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from lon-mail-1.gradwell.net (lon-mail-1.gradwell.net [193.111.201.125]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3SGEfii058315 for ; Thu, 28 Apr 2005 09:14:41 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from host81-144-74-47.midband.mdip.bt.net ([81.144.74.47]) by lon-mail-1.gradwell.net with esmtp (Gradwell gwh-smtpd 1.181) id 42710bef.165c1.5e for ietf-usefor@imc.org; Thu, 28 Apr 2005 17:14:39 +0100 (envelope-sender ) Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3SGCLE16994 for ietf-usefor@imc.org; Thu, 28 Apr 2005 17:12:21 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20768 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: Suggested References texts Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: <426F443C.6561@xyzzy.claranet.de> Date: Thu, 28 Apr 2005 13:30:42 GMT Lines: 110 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In <426F443C.6561@xyzzy.claranet.de> Frank Ellermann writes: >Charles Lindsey wrote: >> A "followup" is an article containing a response to the >> contents of an earlier article. It will always include a >> References header pointing to that earlier article and any >> other "precursors". >Strong NAK to any change of <426E4FC9.7020903@oceana.com> >We know that References do _not_ always include 'any other >"precursors"', in fact there's a complete chapter explaining >how to trim the References. OK. Consider the word "any" struck. >> An article is a "precursor" of some later article which is a >> followup to it >> or which is otherwise intended to be grouped with it for >> purposes of display (e.g. as a part of a multipart posting >> such as a FAQ). >Strong NAK, ... >A "precursor" is simply the opposite of a "followup", so if >A is the precursor of B, then B is a followup of A. There's >no case where A is the precursor of B, but B is no followup >of A, that would be madness. Whilst I do not want the term "followup" to include those "pseudo-followups" (if I can use that term without precipitating another rant from John :-) ), I defined "precursor" to include the "pseudo-precursors" because it simplified the wording of the semantics further down. If you want me to limit "precursor" as you suggest, then please suggest alternative wording for those semantics. The term "followup" is widely used within the Usenet community, and it would be artificial, misleading and unwise to define it in any other way. But "precursor" is not so widely used and would not mislead in the same way. >> Within USEFOR: >>| references = "References:" SP [CFWS] 1*(msg-id [CFWS]) CRLF Oops! That is still wrong. > references = "References:" SP msg-id-list CRLF > msg-id-list = [CFWS] msg-id *( CFWS msg-id ) [CFWS] or, following the style in draft-13, references = "References:" SP [CFWS] msg-id 1*(CFWS msg-id ) [CFWS] CRLF >That's the simple variant where a comment is enough to separate >the Message-IDs as in (c) but adjacent Message-IDs as >in aren't allowed. Agreed. Those s are "MUST accept, don't generate yet", so the issue does not arise until "yet" happens, by which time most agents will probably be able to manage without even a WSP. >Strong NAK to anything below this line. So you want me to omit | The list is composed of message identifiers of precursors of the | current article, sorted so that no article precedes any of its own | precursors. It SHOULD include both the earliest and the immediate | precursors on the current article, even if some of the intermediate | ones are omitted. A given message identifier MUST NOT appear more | than once. But that is the semantics (meaning) of the header, which is a proper matter for USEFOR. It tells reading agents just what they can deduce from the header, and it guides those who generate References for "pseudo-followups". RFC 2822 and our earlier ARTICLE drafts got away by just providing an algorithm for generating this header, but that is not enough if the header is not restricted to followups/replies. > [Usepro] >> If the resulting References header is excessively long, it >> MAY be trimmed, but the first and the last two message >> identifiers MUST NOT be removed. >I'd go for a SHOULD here, because some news servers and tools >really hate it if the References are too long, I've seen cases >where the References were truncated I think we agreed earlier that the actual trimming details belonged in USEAGE. But please see my remark in the latest USEAGE where I would like to keep the whole header within 998 octets (rather than 21 entries), because some followup/reply agents unfold the old references header and then conveniently "forget" to refold the new. -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3SGEfqo058318 for ; Thu, 28 Apr 2005 09:14:41 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3SGEfpR058317 for ietf-usefor-skb; Thu, 28 Apr 2005 09:14:41 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from lon-mail-1.gradwell.net (lon-mail-1.gradwell.net [193.111.201.125]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3SGEdsm058301 for ; Thu, 28 Apr 2005 09:14:40 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from host81-144-74-47.midband.mdip.bt.net ([81.144.74.47]) by lon-mail-1.gradwell.net with esmtp (Gradwell gwh-smtpd 1.181) id 42710bee.165c1.5d for ietf-usefor@imc.org; Thu, 28 Apr 2005 17:14:38 +0100 (envelope-sender ) Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3SGCMJ16999 for ietf-usefor@imc.org; Thu, 28 Apr 2005 17:12:22 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20769 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: Suggested References texts Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: <426E6E66.2070109@epix.net> <426FCA98.3070307@epix.net> Date: Thu, 28 Apr 2005 13:51:26 GMT Lines: 68 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In <426FCA98.3070307@epix.net> "Forrest J. Cavalier III" writes: >Charles Lindsey wrote: > > Inevitably, this changes the information that a reading agent can deduce > > when it sees a References header. It can no longer deduce "this is a > > followup". It can only deduce "the poster wants you to treat this like a > > followup" (by threading it, or by letting you retrieve the precursors by > > clicking on them, or whatever other bells and whistles you provide). >Do you agree we should be concentrating on how the reading agent can >process articles? I think you do. And the text you suggested is flawed >because it declares nothing of use for reading agents. Then what else can be written? If References headers are to be allowed for "pseudo-followups" (and there seems to be agreement on that), then what the reading agent can or cannot deduce is fixed by the structure that we have allowed. Playing around with the definitions cannot alter that. > >> | The References header is used in followups, and other articles with > >> | related precursors (e.g. multipart FAQs and the fragments of a > >> | message/partial), to facilitate the display or retrieval, by reading > >> | and other agents, of threads of related articles. ... >What does "facilitate" mean? What does "other articles" mean? Since "e.g." >is not "i.e.", what other kinds of articles can have References? What stops >someone from using References for other kinds of facilitation of display or >retrieval? (Style sheets!) "Facilitate" means that they can use the information (whose meaning is defined by the semantics further down) in any way they choose in order to provide threading, retrieval, etc. In fact, there are well-known algorithms for threading that work pretty well, but implementors are free to experiment with supposedly better ones, and to some extent different users prefer different presentations. Yes, why not "style sheets"? The important thing is that they need to rely on the References header being present, and being constructed according to the given semantics. >That text is useless. It doesn't even guide posters, much less readers. No, the guidance comes in the semantics paragraph further down, which is supposed to tell them what they can rely on. >If people really want to allow FAQs and multi-part to get threaded, >I wonder if the USEFOR text should be the following single sentence, (which cleverly >does not mention followup or FAQ or multi-part.) > A References header indicates the previously posted context in which > the article was created. Yes, alternative texts to the one I suggested can be considered. >(That "the" means that you can only have at most one context per article.) But people are going to try and followup to more than one article, whether we like it or not. Someday there may even be an agreed way to do it. Hence my semantics was as minimal and future proof as I could make it. -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3SGEenR058308 for ; Thu, 28 Apr 2005 09:14:40 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3SGEePO058307 for ietf-usefor-skb; Thu, 28 Apr 2005 09:14:40 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from lon-mail-1.gradwell.net (lon-mail-1.gradwell.net [193.111.201.125]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3SGEcLf058290 for ; Thu, 28 Apr 2005 09:14:39 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from host81-144-74-47.midband.mdip.bt.net ([81.144.74.47]) by lon-mail-1.gradwell.net with esmtp (Gradwell gwh-smtpd 1.181) id 42710bed.165c1.5c for ietf-usefor@imc.org; Thu, 28 Apr 2005 17:14:37 +0100 (envelope-sender ) Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3SGCHP16973 for ietf-usefor@imc.org; Thu, 28 Apr 2005 17:12:17 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20765 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: References Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: <426E579F.1060809@oceana.com> <426F3B14.54C1@xyzzy.claranet.de> <426F7411.2010903@epix.net> Date: Thu, 28 Apr 2005 11:29:40 GMT Lines: 70 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In <426F7411.2010903@epix.net> "Forrest J. Cavalier III" writes: >I think it is a requirement that a news client be able to determine >"followup" by inspecting the article headers. OK. Then you need to tell us exactly WHY that requirement is needed. As opposed to a requirement "that a news client be able to determine articles that the poster intends to be treated the same way as followups ...". >If it is not a requirement, then the term "followup" is meaningless >and should not appear in any document. The term "followup" has to appear in our standards, if only because we have to explain the purpose of the Followup-To header. And we have to explain the duties of a followup agent. And USEAGE (which is not standards-track) also explains how to contruct attributions and quotations in followups. >In support of this, I point that the References header is for >the use of the READER, not the POSTER. The poster already knows how >all the articles are related. Indeed, but the POSTER gets some say in how the READER gets to view his articles by including (or not including) a References header. >Frank, you allowed that followup => references. If you don't >hold that references => followup, then does the standard allow an >article to reference more than one thread? I think that >allowing References to appear in an article which is not a reply >to a single article (which I thought was the term "followup"), >leads to undecidable actions about threading. What is the READER >agent supposed to do in that situation? Both USEPRO and RFC 2822 state explicitly that they make no provision for followups/replies to multiple precursors, though they do not explicitly forbid them. Some future extension to either or both documents may do so - various suggestions have been made as to how to do it, but there is no agreed method yet. Nevertheless, people will and do do it, and try to construct a References header to suit. In fact, if this is done sensibly, existing threading agents will produce a tolerable behaviour (by which I mean that they will not list an article as available for reading earlier than any of its precursors). In my suggested text for the semantics of the References header, I tried to word it so as to afford the maximum flexibility for such future extensions, whilst always ensuring correct behaviour in the single precursor case. >Is there a term for articles that can have references but are >not followups? If not, why the special term "followup"? What >does the apperance of References header mean, from a READER perspective? I once coined the term "pseudo-followup" just for use in our discussions in this WG (not with any intent of using that term in any document). I was rewarded with the usual multi-page rant from John :-( . -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3SGEc1j058298 for ; Thu, 28 Apr 2005 09:14:38 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3SGEchC058296 for ietf-usefor-skb; Thu, 28 Apr 2005 09:14:38 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from lon-mail-1.gradwell.net (lon-mail-1.gradwell.net [193.111.201.125]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3SGEbFF058278 for ; Thu, 28 Apr 2005 09:14:38 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from host81-144-74-47.midband.mdip.bt.net ([81.144.74.47]) by lon-mail-1.gradwell.net with esmtp (Gradwell gwh-smtpd 1.181) id 42710bec.165c1.5b for ietf-usefor@imc.org; Thu, 28 Apr 2005 17:14:36 +0100 (envelope-sender ) Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3SGCPU17017 for ietf-usefor@imc.org; Thu, 28 Apr 2005 17:12:25 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20770 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: Obsolete vs. obsolescent Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: <200504111212.j3BCChM14516@clerew.man.ac.uk> <87psx1gnjc.fsf@windlord.stanford.edu> <87fyxvx1m6.fsf@windlord.stanford.edu> <87mzs2a1hu.fsf@windlord.stanford.edu> <87wtqxui6s.fsf@windlord.stanford.edu> <87d5so2bpk.fsf@windlord.stanford.edu> <42697BB6.6074@xyzzy.claranet.de> <426BC455.2040108@isode.com> <426E4B45.7010205@oceana.com> <426F4F7C.7385@xyzzy.claranet.de> <87k6moksps.fsf@windlord.stanford.edu> Date: Thu, 28 Apr 2005 14:09:23 GMT Lines: 55 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In <87k6moksps.fsf@windlord.stanford.edu> Russ Allbery writes: >Frank Ellermann writes: >> IMHO all that should be done is to update this statement: >> | As a result, this header is to be regarded as obsolete, and >> | it will likely be removed entirely in a future version of >> | this standard. >> As a result, this header is to be regarded as obsolescent. >> Servers and UAs MUST accept but SHOULD ignore it, and SHOULD >> NOT generate it. >I dislike saying that servers MUST accept a header, since servers are not >required to accept *any* article. I would also leave in the bit about >removing it in a future version of the standard. I agree (though there is a general caveat in USEPRO that the right of any server to drop any article on the floor overrides any "MUST" requirement). But I am happy with the s/obsolete/obsolescent/. >> Subject: cmsg is not mentioned anywhere in usefor-03, that's a >> serious bug. It's mentioned in usepro-03, but it only says >> that this is now obsolete. That's of course FUBAR, it's a >> severe security violation. We cannot jump from RfC 1036 to a >> completely different procedure at date X. AFAIK RfCs must not >> have a flag day. >I think USEFOR does need an explicit mention that subjects starting with >cmsg should not be considered control messages, to document the change >from RFC 1036. No, I think it is a protocol issue, and the mention in USEPRO is sufficient. Any I do not share Frank's worries about a flag day. That usage is already ignored by most servers, so nobody should be relying on its working anymore. If the few remaining servers that honour it suddenly become non-compliant the day our RFC gets published, then so be it. The sky will not immediately fall in, but anyone continuing to use or accept the practice will be severaly sat upon. In any case, I have not seen or heard of anybody actually trying to use the feature in earnest for many years. -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3SGEccR058291 for ; Thu, 28 Apr 2005 09:14:38 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3SGEcrw058288 for ietf-usefor-skb; Thu, 28 Apr 2005 09:14:38 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from lon-mail-1.gradwell.net (lon-mail-1.gradwell.net [193.111.201.125]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3SGEYdL058233 for ; Thu, 28 Apr 2005 09:14:34 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from host81-144-74-47.midband.mdip.bt.net ([81.144.74.47]) by lon-mail-1.gradwell.net with esmtp (Gradwell gwh-smtpd 1.181) id 42710be9.165c1.57 for ietf-usefor@imc.org; Thu, 28 Apr 2005 17:14:33 +0100 (envelope-sender ) Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3SGCI116983 for ietf-usefor@imc.org; Thu, 28 Apr 2005 17:12:18 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20766 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: References Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: <87fyxcksel.fsf@windlord.stanford.edu> <426FDD09.5030102@oceana.com> <426FE5EB.5030302@isode.com> <426FE83E.3040503@isode.com> Date: Thu, 28 Apr 2005 11:40:18 GMT Lines: 28 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In <426FE83E.3040503@isode.com> Alexey Melnikov writes: >Alexey Melnikov wrote: >> I do, but the devil is in details: what exactly should be said and where. >The USEFOR document can say that certain types of messages MAY require >the References header. But I just think this should be obvious. I think we say the minimum necessary to prevent other readers from misinterpreting what the document says in the way that John has been misinterpreting it. A definition of "followup" that clearly distinguishes them from RFC 2822 "replies" is a good start (and I think we already agreed to that). Also, if we intend that References may sometimes be used with non-followups, it would be useful to say so. -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3SGEbtH058276 for ; Thu, 28 Apr 2005 09:14:37 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3SGEbdn058275 for ietf-usefor-skb; Thu, 28 Apr 2005 09:14:37 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from lon-mail-1.gradwell.net (lon-mail-1.gradwell.net [193.111.201.125]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3SGEZPf058248 for ; Thu, 28 Apr 2005 09:14:36 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from host81-144-74-47.midband.mdip.bt.net ([81.144.74.47]) by lon-mail-1.gradwell.net with esmtp (Gradwell gwh-smtpd 1.181) id 42710bea.165c1.59 for ietf-usefor@imc.org; Thu, 28 Apr 2005 17:14:34 +0100 (envelope-sender ) Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3SGCTL17037 for ietf-usefor@imc.org; Thu, 28 Apr 2005 17:12:29 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20773 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: Differences between RFC 2822 and Usefor (fwd) Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: Date: Thu, 28 Apr 2005 15:36:33 GMT Lines: 51 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In Richard Clayton writes: >So it seems to me that this is a genuine inter-operability issue -- not >at the protocol levels : the articles get to the far end just fine and >the servers will serve them perfectly well. The issue occurs at the >level at which the readers view the newsgroups. For that reason I think >that a standardised way of expressing which pile an article belongs in >is of very significant value. Yes at present some software is a bit >flaky, but that doesn't mean we should fail to say what should be done. I agree entirely. RFC 2119 is very badly written, and is wide open to those who claim that "interoperability" means only that "it must be delivered correctly at the far end". The WG has long striven to use "MUST" wording for cases where "it does not behave as it is intended to behave if you don't say 'MUST'". Anyway, to deal with this present case, you will find some wording in USEPRO 7.7 entitled "Duties of a Reading Agent" where it says, in effect, that the agent MAY take note of the References, Subject and Date headers in order to display list of articles sensibly. This makes such displaying an optional part of the protocol, which in turn means that reading agents are entitled to rely upon the presence and correctness of those headers. Yes, it is pretty meaningless, and it is only there to satisfy the pedants who object to the use of MUST and SHOULD when dealing with the behaviour of followup agents. But if that is the price we have to pay to get this wording in, .......... >sounds like a cock-up :( It WAS a cock-up. >so best fix it then! ASAP!! Exactly. >Signature by unknown keyid: 0xD8D683E2 Eh? I though I knew all your public keys. -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3SGEbXJ058285 for ; Thu, 28 Apr 2005 09:14:37 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3SGEbYb058284 for ietf-usefor-skb; Thu, 28 Apr 2005 09:14:37 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from lon-mail-1.gradwell.net (lon-mail-1.gradwell.net [193.111.201.125]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3SGEagN058261 for ; Thu, 28 Apr 2005 09:14:37 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from host81-144-74-47.midband.mdip.bt.net ([81.144.74.47]) by lon-mail-1.gradwell.net with esmtp (Gradwell gwh-smtpd 1.181) id 42710beb.165c1.5a for ietf-usefor@imc.org; Thu, 28 Apr 2005 17:14:35 +0100 (envelope-sender ) Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3SGCKQ16989 for ietf-usefor@imc.org; Thu, 28 Apr 2005 17:12:20 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20767 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: News headers (was References) Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: Date: Thu, 28 Apr 2005 12:06:46 GMT Lines: 58 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In John Stanley writes: >RFC2822 does not say "email replies". Since USEFOR defers to RFC2822, we >have chosen to expand any "email" limitation of RFC2822 into USENET. If we >are so pedantic that we say "RFC2822 is EMAIL only" even in the face of >explicit references to RFC2822 by draft news standards, then exactly what >headers DOES news have and how are they defined? Many of them are defined >as "as found in RFC2822". But since those are EMAIL headers, then they >don't apply to news. Sigh. You do have a point there regarding which headers are officially part of "News". RFC 2822 defines "Optional fields", so that the following header is perfectly legal in any email: Foobar: bar foo However, the clear intent of that field was to allow other standards to define further email headers, such as Content-type: text/plain and have them automatically be acceptable to RFC 2822. It was also intended to allow any header beginning with "X-". But those intents were not clearly stated, which is a pity IMO. The situation was somewhat restored by RFC 3864, which establishes an IANA registry for email (and also news and HTTP) headers. What I would like to see in USEFOR is a clear statement that the headers it explicitly mentions (From, Subject, Newsgroups, etc) are "official Netnews headers" and that ones it does not mention are "passengers" which are generally to be ignored by news software, but which should nevertheless be passed on untouched by transports. Not that I am proposing to define "passenger" as a technical term for the final document. Examples of such "passengers" are the Received and In-Reply-To headers, which are not, and never have been, a part of Netnews, but which do show up occasionally, especially in articles gatewayed from email. Indeed, I would go further, and say that any header appearing in a News article SHOULD either be one blessed by the RFC 3864 registry, or else should begin with "X-". In fact, the old draft-13 contained wording of this nature, but it seems not to have made its way into USEFOR yet. -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3SGEZ4E058258 for ; Thu, 28 Apr 2005 09:14:35 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3SGEZIi058257 for ietf-usefor-skb; Thu, 28 Apr 2005 09:14:35 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from lon-mail-1.gradwell.net (lon-mail-1.gradwell.net [193.111.201.125]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3SGEYko058235 for ; Thu, 28 Apr 2005 09:14:35 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from host81-144-74-47.midband.mdip.bt.net ([81.144.74.47]) by lon-mail-1.gradwell.net with esmtp (Gradwell gwh-smtpd 1.181) id 42710be9.165c1.58 for ietf-usefor@imc.org; Thu, 28 Apr 2005 17:14:33 +0100 (envelope-sender ) Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3SGCS817033 for ietf-usefor@imc.org; Thu, 28 Apr 2005 17:12:28 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20772 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: Fixed Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: <4255153E.3080500@isode.com> <42594879.3B92@xyzzy.claranet.de> <425A8A63.268A@xyzzy.claranet.de> <425C607E.1B5A@xyzzy.claranet.de> <425D8031.D57@xyzzy.claranet.de> <426BB63D.3060506@isode.com> <426F6AC8.5045@xyzzy.claranet.de> Date: Thu, 28 Apr 2005 15:18:48 GMT Lines: 27 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In <426F6AC8.5045@xyzzy.claranet.de> Frank Ellermann writes: >Alexey Melnikov wrote: >>> the general principle that we do not change anything in RFC >>> 2822 that we do not need to change. >> I concur with your preference. >I don't. But you asked for someone to come along and provide a casting vote to settle the difference between you (wanting @) and me (wanting to retain and ). Now you have it, and it puts you in a minority (even if it does not amount to a ruling from the Chair). You are the only person wanting @. -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3SGEZgD058252 for ; Thu, 28 Apr 2005 09:14:35 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3SGEZCe058249 for ietf-usefor-skb; Thu, 28 Apr 2005 09:14:35 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from lon-mail-1.gradwell.net (lon-mail-1.gradwell.net [193.111.201.125]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3SGEYAk058232 for ; Thu, 28 Apr 2005 09:14:34 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from host81-144-74-47.midband.mdip.bt.net ([81.144.74.47]) by lon-mail-1.gradwell.net with esmtp (Gradwell gwh-smtpd 1.181) id 42710be8.165c1.56 for ietf-usefor@imc.org; Thu, 28 Apr 2005 17:14:32 +0100 (envelope-sender ) Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3SGCUs17041 for ietf-usefor@imc.org; Thu, 28 Apr 2005 17:12:31 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20774 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: Suggested References texts Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: <87br80ks7i.fsf@windlord.stanford.edu> Date: Thu, 28 Apr 2005 15:42:13 GMT Lines: 29 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In <87br80ks7i.fsf@windlord.stanford.edu> Russ Allbery writes: >Either we define followups by poster intent and some articles with >Reference headers aren't followups, or we define followups by the presence >of a Reference header and we can't make a strong statement about intent. >We can't have it both ways; Reference headers are used in practice for >things other than followups, and I don't expect that practice to change >(nor do I think we really *want* it to change). Exactly. But still John believes that the real world can be changed by smoke and mirrors. I have explained it to him. Seth has explained it to him. Shmuel has explained it to him. And now you have explained it to him. But he still does not see it. I should not waste any more time on it. -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3SD2iKk021418 for ; Thu, 28 Apr 2005 06:02:44 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3SD2ini021417 for ietf-usefor-skb; Thu, 28 Apr 2005 06:02:44 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from ciao.gmane.org (main.gmane.org [80.91.229.2]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3SD2gx6021392 for ; Thu, 28 Apr 2005 06:02:43 -0700 (PDT) (envelope-from usenet-format@gmane.org) Received: from list by ciao.gmane.org with local (Exim 4.43) id 1DR8YZ-0007Ye-4r for ietf-usefor@imc.org; Thu, 28 Apr 2005 14:55:51 +0200 Received: from c-134-88-108.hh.dial.de.ignite.net ([62.134.88.108]) by main.gmane.org with esmtp (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Thu, 28 Apr 2005 14:55:51 +0200 Received: from nobody by c-134-88-108.hh.dial.de.ignite.net with local (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Thu, 28 Apr 2005 14:55:51 +0200 X-Injected-Via-Gmane: http://gmane.org/ To: ietf-usefor@imc.org From: Frank Ellermann Subject: Re: Obsolete vs. obsolescent Date: Thu, 28 Apr 2005 14:59:39 +0200 Organization: Lines: 34 Message-ID: <4270DE3B.4BCC@xyzzy.claranet.de> References: <200504111212.j3BCChM14516@clerew.man.ac.uk> <87psx1gnjc.fsf@windlord.stanford.edu> <87fyxvx1m6.fsf@windlord.stanford.edu> <87mzs2a1hu.fsf@windlord.stanford.edu> <87wtqxui6s.fsf@windlord.stanford.edu> <87d5so2bpk.fsf@windlord.stanford.edu> <42697BB6.6074@xyzzy.claranet.de> <426BC455.2040108@isode.com> <426E4B45.7010205@oceana.com> <426F4F7C.7385@xyzzy.claranet.de> <87k6moksps.fsf@windlord.stanford.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Complaints-To: usenet@sea.gmane.org X-Gmane-NNTP-Posting-Host: c-134-88-108.hh.dial.de.ignite.net X-Mailer: Mozilla 3.0 (OS/2; U) Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Russ Allbery wrote: > I dislike saying that servers MUST accept a header Yes, we have some seriouss case of "Deja-Vu" on this list, I'm not immune, sorry. What I really wanted to say, there's a difference between the "Lines:" and other obsolete header fields in 3.3. "Lines:" is deprecated _now_ by the future Usefor RfC. It was optional and unreliable in s-o-1036. Some other headers mentioned in 3.3 are something I've never seen or heard of: Relay-Version, Posting-Version, Date-Received. s-o-1036 says "thoroughly obsolete" about these three header fields listed in its appendix A.3, that was 1994. Finally there are some headers introduced by s-o-1036 that didn't make it (unlike Supersedes): See-Also, Article-Names, Article-Replacing, and Also-Control. Strange, my copy of s-o-1036 (Jan 17 1994) says "Article-Replacing", Usefor-03 says "Article-Updates". Nothing of this is critical, but Also-Control could be a special case like Subject: cmsg cancel. It would be really easier to build on s-o-1036, pro forma. Even if the future Usefor RfC is published only one second after s-o-1036 it could eliminate some weird logical problems. Bye, Frank Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3SC8Kv0098885 for ; Thu, 28 Apr 2005 05:08:20 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3SC8K1M098884 for ietf-usefor-skb; Thu, 28 Apr 2005 05:08:20 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from ciao.gmane.org (main.gmane.org [80.91.229.2]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3SC8IpU098867 for ; Thu, 28 Apr 2005 05:08:19 -0700 (PDT) (envelope-from usenet-format@gmane.org) Received: from list by ciao.gmane.org with local (Exim 4.43) id 1DR7hx-0000fZ-8B for ietf-usefor@imc.org; Thu, 28 Apr 2005 14:01:29 +0200 Received: from c-134-88-108.hh.dial.de.ignite.net ([62.134.88.108]) by main.gmane.org with esmtp (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Thu, 28 Apr 2005 14:01:29 +0200 Received: from nobody by c-134-88-108.hh.dial.de.ignite.net with local (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Thu, 28 Apr 2005 14:01:29 +0200 X-Injected-Via-Gmane: http://gmane.org/ To: ietf-usefor@imc.org From: Frank Ellermann Subject: Re: References Date: Thu, 28 Apr 2005 14:04:33 +0200 Organization: Lines: 54 Message-ID: <4270D151.776C@xyzzy.claranet.de> References: <426E579F.1060809@oceana.com> <426F3B14.54C1@xyzzy.claranet.de> <426F7411.2010903@epix.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Complaints-To: usenet@sea.gmane.org X-Gmane-NNTP-Posting-Host: c-134-88-108.hh.dial.de.ignite.net X-Mailer: Mozilla 3.0 (OS/2; U) Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Forrest J. Cavalier III wrote: > I think it is a requirement that a news client be able to > determine "followup" by inspecting the article headers. Maybe you like to join Bruce and some others including me in our time travel excursion to fix some RfCs, you could be the advisor to Dave Crocker when he fixes STD 11 4.6.3. We could do radical things like publishing s-o-1036 as RfC. Seriously, this "requirement" is wishful thinking, you find it as See-Also in s-o-1036. All we can do today is to say how a followup is supposed to be created and how References are trimmed. We cannot guarantee that all References were built this way. The very best we can do is to be silent about more creative ways to generate References. > If it is not a requirement, then the term "followup" is > meaningless and should not appear in any document. "Followup" has a perfect meaning defined by Followup-To (or Newsgroups). > What is the READER agent supposed to do in that situation? It's supposed to expect the worst, but to work for the 99.9% ordinary followups. In the same way as it did for the last decade. BTW, I've tested the effect of an article referencing itself on this list some moths ago. Let's say that I won't test this again. > Is there a term for articles that can have references but > are not followups? How about "literal Reference" ? "STD 11 4.6.3 Reference" ? "Multi-2822 quasi-In-Reply-To" ? Or good old "See-Also" ? > why the special term "followup"? Maybe ask Henry or Bruce, it must be very old, "Followup-To" is very old. For me nothing before s-o-1036 is real, all I knew about Usenet before s-o-1036 was "Emily Postnews", and the fascinating differences from FidoNet, because I used to hang out in a gateways NG ("echo" from my POV) for some time. That's also the reason why I consider attempts to devalue Usenet Message-IDs to some as blasphemy. Bye, Frank Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3SBEO4H076547 for ; Thu, 28 Apr 2005 04:14:24 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3SBEMaR076546 for ietf-usefor-skb; Thu, 28 Apr 2005 04:14:22 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from ciao.gmane.org (main.gmane.org [80.91.229.2]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3SBEKc4076512 for ; Thu, 28 Apr 2005 04:14:21 -0700 (PDT) (envelope-from usenet-format@gmane.org) Received: from list by ciao.gmane.org with local (Exim 4.43) id 1DR6sb-0003D4-Si for ietf-usefor@imc.org; Thu, 28 Apr 2005 13:08:25 +0200 Received: from c-134-88-108.hh.dial.de.ignite.net ([62.134.88.108]) by main.gmane.org with esmtp (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Thu, 28 Apr 2005 13:08:25 +0200 Received: from nobody by c-134-88-108.hh.dial.de.ignite.net with local (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Thu, 28 Apr 2005 13:08:25 +0200 X-Injected-Via-Gmane: http://gmane.org/ To: ietf-usefor@imc.org From: Frank Ellermann Subject: Re: References Date: Thu, 28 Apr 2005 13:12:15 +0200 Organization: Lines: 17 Message-ID: <4270C50F.1637@xyzzy.claranet.de> References: <87fyxcksel.fsf@windlord.stanford.edu> <426FDD09.5030102@oceana.com> <426FE5EB.5030302@isode.com> <426FE83E.3040503@isode.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Complaints-To: usenet@sea.gmane.org X-Gmane-NNTP-Posting-Host: c-134-88-108.hh.dial.de.ignite.net X-Mailer: Mozilla 3.0 (OS/2; U) Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Alexey Melnikov wrote: > I just think this should be obvious. Sure. It was discussed here for the major part of 2004. News is not the same as mail and 2822. There's a note in s-o-1036 explaining why some References should be a See-Also, but we know that See-Also didn't fly. So we have References used for followups as defined in Usepro, and we have References as defined in STD 11 4.6, and unfortunately that's not always identical. What else is new ? Is Usefor supposed to document common practice or wishful thinking based on 2822 ? Bye, Frank Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RNZD7n003222 for ; Wed, 27 Apr 2005 16:35:13 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3RNZDi0003221 for ietf-usefor-skb; Wed, 27 Apr 2005 16:35:13 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from b.mail.peak.org (b.mail.peak.org [69.59.192.42]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RNZCoZ003213 for ; Wed, 27 Apr 2005 16:35:13 -0700 (PDT) (envelope-from stanley@peak.org) Received: from a.shell.peak.org ([69.59.192.81]) by b.mail.peak.org (8.12.10/8.12.8) with ESMTP id j3RNZ6wU068498 (version=TLSv1/SSLv3 cipher=EDH-RSA-DES-CBC3-SHA bits=168 verify=NO) for ; Wed, 27 Apr 2005 16:35:07 -0700 (PDT) Date: Wed, 27 Apr 2005 16:35:06 -0700 (PDT) From: John Stanley X-X-Sender: stanley@a.shell.peak.org To: ietf-usefor@imc.org Subject: Re: References Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Spam-Score: 0 () X-Scanned-By: MIMEDefang 2.39 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Alexey Melnikov : >The USEFOR document can say that certain types of messages MAY require the >References header. But I just think this should be obvious. It is interesting to note that just this morning in the office we had a discussion about the use of the terms "obvious" and "clearly", and that when someone says something is either "obvious" or "clearly", it usually isn't. In this case, USEFOR defers to RFC2822, which clearly and obviously says that References is optional -- specifically, it uses the RFC2119 term "SHOULD", and in some places only "may", and further talks about how to process replies that do not contain a References header. It is hard to imagine that this is a requirement. Saying that USEFOR "obviously" says that some articles require a References header is, well, contrary to what USEFOR actually says. That's fine, it has been changed and we can live with that, as long as the change is propogated to all relevant documents, which includes USEPRO. We could change it back, but so far only a very small handful of people have been arguing that RFC2119 language is justified, and even then they haven't been providing any examples of interoperability issues, just things like "ensure that definitions are obeyed" and they think it should say "MUST". Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RLbWu6092074 for ; Wed, 27 Apr 2005 14:37:32 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3RLbWBo092073 for ietf-usefor-skb; Wed, 27 Apr 2005 14:37:32 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from a.mail.peak.org (a.mail.peak.org [69.59.192.41]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RLbVux092055 for ; Wed, 27 Apr 2005 14:37:31 -0700 (PDT) (envelope-from stanley@peak.org) Received: from a.shell.peak.org ([69.59.192.81]) by a.mail.peak.org (8.12.10/8.12.8) with ESMTP id j3RLbPgU086379 (version=TLSv1/SSLv3 cipher=EDH-RSA-DES-CBC3-SHA bits=168 verify=NO) for ; Wed, 27 Apr 2005 14:37:25 -0700 (PDT) Date: Wed, 27 Apr 2005 14:37:25 -0700 (PDT) From: John Stanley X-X-Sender: stanley@a.shell.peak.org To: ietf-usefor@imc.org Subject: Re: References Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Spam-Score: 0 () X-Scanned-By: MIMEDefang 2.39 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Russ Allbery : Me: Yes. It is pretty clear that all replies are followups. >What about, er, replies? In the trn sense. An e-mail reply to a Usenet >message. Those are what RFC 2822 more obviously applies to; RFC2822 does not say "email replies". Since USEFOR defers to RFC2822, we have chosen to expand any "email" limitation of RFC2822 into USENET. If we are so pedantic that we say "RFC2822 is EMAIL only" even in the face of explicit references to RFC2822 by draft news standards, then exactly what headers DOES news have and how are they defined? Many of them are defined as "as found in RFC2822". But since those are EMAIL headers, then they don't apply to news. Sigh. >This thread is convincing me that it's a bit confusing to defer to RFC >2822 entirely for References when RFC 2822 only talks about replies and >clearly is intended to focus on mail messages. Usenet followups are a bit >of a different beast with historically stronger restrictions on the >presence of the References header that e-mail is only now starting to >adopt. "Historically stronger" is not found in RFC2119 as a reason for mandates, and "historically stronger" is apparently not strong enough to have resulted in the mandates being continued into draft-03 of USEFOR. >I still feel like this could be dealt with definitionally, but >followups in the Usenet sense have References headers, period. It's not >just a SHOULD. Currently, it is just a SHOULD. If we choose to return it to MUST, we can do so, but what do we use to justify that change, and then, do we have the nerve to do it right or do we continue to contradict ourselves? >Either we define followups by poster intent and some articles with >Reference headers aren't followups, The latter does not follow from the former. We can easily (and have, in the past) prohibit References headers in non-followups. > or we define followups by the presence >of a Reference header and we can't make a strong statement about intent. So I can change an article from being a followup to not being a followup simply by removing one header? It seems odd that I can change the origination of an article by simply removing a header when his article reaches me halfway around the world from where he posted it. >We can't have it both ways; Reference headers are used in practice for >things other than followups, and I don't expect that practice to change >(nor do I think we really *want* it to change). We can solve this problem by simply expanding trivially the definition of followup. We don't have to dick with the mandates for when a References header is or is not to appear in an article. But "we" chose the latter course, to the point that they are now optional and followups are indetectable. Ken Murchison : >...USEFOR discusses semantics elsewhere so we can and probably should >discuss the use of the References header in a more Usenet-centric way. And exactly what shall our excuse be for ignoring RFC2119 after removing mandates regarding a header whose mandates have always been a point of debate? Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RLHTX5089026 for ; Wed, 27 Apr 2005 14:17:29 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3RLHTZT089025 for ietf-usefor-skb; Wed, 27 Apr 2005 14:17:29 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from b.mail.peak.org (b.mail.peak.org [69.59.192.42]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RLHSGU089007 for ; Wed, 27 Apr 2005 14:17:29 -0700 (PDT) (envelope-from stanley@peak.org) Received: from a.shell.peak.org ([69.59.192.81]) by b.mail.peak.org (8.12.10/8.12.8) with ESMTP id j3RLHMwU009175 (version=TLSv1/SSLv3 cipher=EDH-RSA-DES-CBC3-SHA bits=168 verify=NO) for ; Wed, 27 Apr 2005 14:17:23 -0700 (PDT) Date: Wed, 27 Apr 2005 14:17:21 -0700 (PDT) From: John Stanley X-X-Sender: stanley@a.shell.peak.org To: ietf-usefor@imc.org Subject: Re: Differences between RFC 2822 and Usefor (fwd) Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Spam-Score: 0 () X-Scanned-By: MIMEDefang 2.39 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Richard Clayton : > one article they will cope with -- Many articles they will cope with, I hope. Many articles have no References header. >-- but if no articles >have References then they will fail to function as their users currently >expect. I call that broken, and so will they. Once again we are stuck with the sometimes unusual expectations of USENET users. And how did we get to "no articles have References headers?" Are you imagining some prohibition on the use of References headers? >I believe that References: should be compulsory, RFC2119 defines when MUST is appropriate. This is not such a time. >however, >now it's clear that there's the prospect of (a) the effort finishing >(one way or another) and (b) posters being permitted to break >"threading" (and if there is no language to compel them, then they will) >then it is time to speak up Posters are always permitted to "break threading". They will always be permitted to "break threading". Seth Breidbart : >Define "break". If you mean "crash and burn and let the magic smoke >out of the computer" then no. Well, let's see. "Break" as in "fail to process articles", ok? No, since we are talking about interoperability, that would have to be "fail to accept articles or to pass them to other news system software." I read news today. I found articles (gasp!) without References headers. My newsreader did not break. It simply did not care. >If you mean "Do the wrong thing (from >the viewpoints of both the poster and reader)" then yes. Since this is a display issue, what is "the wrong thing" and how do you know when it happens? "Wrong" is completely subjective. Subjective is not how one writes standards. In this specific case, some people think that "right" is "sort by subject content". Some think that "sort by precursor" is "right" and "sort by subject" is wrong. Who is "right" and who is "wrong"? >If a random set of Subject header content was changed to "3" the >system wouldn't collapse, either; so why is specifying that some >headers MUST NOT be changed by transit servers an interoperability >issue? Because transport agents are DEFINED not to change the message they are transporing, and the header you are referring to is a USER controlled header, not a news system controlled one. >Answer: That's doing _the wrong thing_. That's your opinion. I could imagine a system where the Subject header IS changed in transport that is quite correct. In fact, some email spam systems do exactly that. Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RJU8mR074222 for ; Wed, 27 Apr 2005 12:30:08 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3RJU8Fa074221 for ietf-usefor-skb; Wed, 27 Apr 2005 12:30:08 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from rufus.isode.com (rufus.isode.com [62.3.217.251]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RJU792074207 for ; Wed, 27 Apr 2005 12:30:07 -0700 (PDT) (envelope-from alexey.melnikov-usefor@isode.com) Received: from [172.16.2.180] (shiny.isode.com [62.3.217.250]) by rufus.isode.com via TCP (internal) with ESMTPA; Wed, 27 Apr 2005 20:30:06 +0100 Message-ID: <426FE83E.3040503@isode.com> Date: Wed, 27 Apr 2005 20:30:06 +0100 From: Alexey Melnikov User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.7.2) Gecko/20040804 Netscape/7.2 (ax) X-Accept-Language: en-us, en To: ietf-usefor@imc.org Subject: Re: References References: <87fyxcksel.fsf@windlord.stanford.edu> <426FDD09.5030102@oceana.com> <426FE5EB.5030302@isode.com> In-Reply-To: <426FE5EB.5030302@isode.com> MIME-version: 1.0 Content-type: text/plain; charset="ISO-8859-1"; format="flowed" Content-transfer-encoding: 7bit Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Alexey Melnikov wrote: > Ken Murchison wrote: > >> Russ Allbery wrote: >> >>> This thread is convincing me that it's a bit confusing to defer to RFC >>> 2822 entirely for References when RFC 2822 only talks about replies and >>> clearly is intended to focus on mail messages. Usenet followups are >>> a bit >>> of a different beast with historically stronger restrictions on the >>> presence of the References header that e-mail is only now starting to >>> adopt. I still feel like this could be dealt with definitionally, but >>> followups in the Usenet sense have References headers, period. It's >>> not >>> just a SHOULD. And I don't feel like that's a divergence from RFC >>> 2822, >>> since RFC 2822 doesn't deal with Usenet followups and the semantics >>> of the >>> References header haven't changed. >> >> I'm leaning towards the same conclusion. Deferring to RFC 2822 in >> draft-03 is looking like bad idea. As Charles stated, USEFOR >> discusses semantics elsewhere so we can and probably should discuss >> the use of the References header in a more Usenet-centric way. >> >> Does our Chair agree? > > I do, but the devil is in details: what exactly should be said and where. The USEFOR document can say that certain types of messages MAY require the References header. But I just think this should be obvious. Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RJKGYN072811 for ; Wed, 27 Apr 2005 12:20:16 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3RJKGww072810 for ietf-usefor-skb; Wed, 27 Apr 2005 12:20:16 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from rufus.isode.com (rufus.isode.com [62.3.217.251]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RJKFi0072804 for ; Wed, 27 Apr 2005 12:20:15 -0700 (PDT) (envelope-from alexey.melnikov-usefor@isode.com) Received: from [172.16.2.180] (shiny.isode.com [62.3.217.250]) by rufus.isode.com via TCP (internal) with ESMTPA; Wed, 27 Apr 2005 20:20:11 +0100 Message-ID: <426FE5EB.5030302@isode.com> Date: Wed, 27 Apr 2005 20:20:11 +0100 From: Alexey Melnikov User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.7.2) Gecko/20040804 Netscape/7.2 (ax) X-Accept-Language: en-us, en To: Ken Murchison CC: ietf-usefor@imc.org Subject: Re: References References: <87fyxcksel.fsf@windlord.stanford.edu> <426FDD09.5030102@oceana.com> In-Reply-To: <426FDD09.5030102@oceana.com> MIME-version: 1.0 Content-type: text/plain; charset="ISO-8859-1"; format="flowed" Content-transfer-encoding: 7bit Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Ken Murchison wrote: > Russ Allbery wrote: > >> This thread is convincing me that it's a bit confusing to defer to RFC >> 2822 entirely for References when RFC 2822 only talks about replies and >> clearly is intended to focus on mail messages. Usenet followups are >> a bit >> of a different beast with historically stronger restrictions on the >> presence of the References header that e-mail is only now starting to >> adopt. I still feel like this could be dealt with definitionally, but >> followups in the Usenet sense have References headers, period. It's not >> just a SHOULD. And I don't feel like that's a divergence from RFC 2822, >> since RFC 2822 doesn't deal with Usenet followups and the semantics >> of the >> References header haven't changed. > > I'm leaning towards the same conclusion. Deferring to RFC 2822 in > draft-03 is looking like bad idea. As Charles stated, USEFOR > discusses semantics elsewhere so we can and probably should discuss > the use of the References header in a more Usenet-centric way. > > Does our Chair agree? I do, but the devil is in details: what exactly should be said and where. Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RIhEc8067300 for ; Wed, 27 Apr 2005 11:43:14 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3RIhEEY067299 for ietf-usefor-skb; Wed, 27 Apr 2005 11:43:14 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from eagle.oceana.com (eagle.oceana.com [208.17.123.12]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RIhDu9067275 for ; Wed, 27 Apr 2005 11:43:13 -0700 (PDT) (envelope-from ken@oceana.com) Received: from [192.168.10.26] (KEN.oceana.com [192.168.10.26]) by eagle.oceana.com (8.13.2/8.13.2) with ESMTP id j3RIh6wi005248; Wed, 27 Apr 2005 14:43:06 -0400 (EDT) Message-ID: <426FDD09.5030102@oceana.com> Date: Wed, 27 Apr 2005 14:42:17 -0400 From: Ken Murchison Organization: Oceana Matrix Ltd. User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.7) Gecko/20040514 X-Accept-Language: en,pdf MIME-Version: 1.0 To: ietf-usefor@imc.org CC: Alexey Melnikov Subject: Re: References References: <87fyxcksel.fsf@windlord.stanford.edu> In-Reply-To: <87fyxcksel.fsf@windlord.stanford.edu> Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit X-Spam-Score: 0 (BAYES_00) X-Scanned-By: MIMEDefang 2.39 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Russ Allbery wrote: > This thread is convincing me that it's a bit confusing to defer to RFC > 2822 entirely for References when RFC 2822 only talks about replies and > clearly is intended to focus on mail messages. Usenet followups are a bit > of a different beast with historically stronger restrictions on the > presence of the References header that e-mail is only now starting to > adopt. I still feel like this could be dealt with definitionally, but > followups in the Usenet sense have References headers, period. It's not > just a SHOULD. And I don't feel like that's a divergence from RFC 2822, > since RFC 2822 doesn't deal with Usenet followups and the semantics of the > References header haven't changed. I'm leaning towards the same conclusion. Deferring to RFC 2822 in draft-03 is looking like bad idea. As Charles stated, USEFOR discusses semantics elsewhere so we can and probably should discuss the use of the References header in a more Usenet-centric way. Does our Chair agree? -- Kenneth Murchison Oceana Matrix Ltd. Software Engineer 21 Princeton Place 716-662-8973 x26 Orchard Park, NY 14127 --PGP Public Key-- http://www.oceana.com/~ken/ksm.pgp Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RI8q2g062183 for ; Wed, 27 Apr 2005 11:08:52 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3RI8qtJ062182 for ietf-usefor-skb; Wed, 27 Apr 2005 11:08:52 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtp1.Stanford.EDU (smtp1.Stanford.EDU [171.67.16.123]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RI8q9I062175 for ; Wed, 27 Apr 2005 11:08:52 -0700 (PDT) (envelope-from rra@stanford.edu) Received: from windlord.stanford.edu (windlord.Stanford.EDU [171.64.19.147]) by smtp1.Stanford.EDU (8.12.11/8.12.11) with SMTP id j3RI8ne4005898 for ; Wed, 27 Apr 2005 11:08:49 -0700 Received: (qmail 10732 invoked by uid 1000); 27 Apr 2005 18:08:49 -0000 To: ietf-usefor@imc.org Subject: Re: Suggested References texts In-Reply-To: (John Stanley's message of "Wed, 27 Apr 2005 09:59:15 -0700 (PDT)") References: From: Russ Allbery Organization: The Eyrie Date: Wed, 27 Apr 2005 11:08:49 -0700 Message-ID: <87br80ks7i.fsf@windlord.stanford.edu> User-Agent: Gnus/5.1007 (Gnus v5.10.7) XEmacs/21.4 (Jumbo Shrimp, linux) MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: John Stanley writes: > Russ Allbery : >> If this is a requirement, then we have to define everything that >> contains a References header as a followup, > You have it backwards. We define a followup by the poster's intent; we > DETECT followups with the References header. There is a difference. Your > definition leaves us with a circle: a followup contains a References > header, and a References header must be used in a followup. Ok, what's a > followup? Either we define followups by poster intent and some articles with Reference headers aren't followups, or we define followups by the presence of a Reference header and we can't make a strong statement about intent. We can't have it both ways; Reference headers are used in practice for things other than followups, and I don't expect that practice to change (nor do I think we really *want* it to change). -- Russ Allbery (rra@stanford.edu) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RI4aJs061123 for ; Wed, 27 Apr 2005 11:04:36 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3RI4arC061122 for ietf-usefor-skb; Wed, 27 Apr 2005 11:04:36 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtp1.Stanford.EDU (smtp1.Stanford.EDU [171.67.16.123]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RI4ZhF061114 for ; Wed, 27 Apr 2005 11:04:35 -0700 (PDT) (envelope-from rra@stanford.edu) Received: from windlord.stanford.edu (windlord.Stanford.EDU [171.64.19.147]) by smtp1.Stanford.EDU (8.12.11/8.12.11) with SMTP id j3RI4YtV004185 for ; Wed, 27 Apr 2005 11:04:35 -0700 Received: (qmail 10674 invoked by uid 1000); 27 Apr 2005 18:04:34 -0000 To: ietf-usefor@imc.org Subject: Re: References In-Reply-To: (John Stanley's message of "Wed, 27 Apr 2005 10:42:19 -0700 (PDT)") References: From: Russ Allbery Organization: The Eyrie Date: Wed, 27 Apr 2005 11:04:34 -0700 Message-ID: <87fyxcksel.fsf@windlord.stanford.edu> User-Agent: Gnus/5.1007 (Gnus v5.10.7) XEmacs/21.4 (Jumbo Shrimp, linux) MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: John Stanley writes: > "Forrest J. Cavalier III" : >> Did anyone ever advocate that 2822-reply => followup? > Yes. It is pretty clear that all replies are followups. What about, er, replies? In the trn sense. An e-mail reply to a Usenet message. Those are what RFC 2822 more obviously applies to; followups in the Usenet sense have additional restrictions that RFC 2822 has no comment on and aren't really something RFC 2822 was trying to deal with. This thread is convincing me that it's a bit confusing to defer to RFC 2822 entirely for References when RFC 2822 only talks about replies and clearly is intended to focus on mail messages. Usenet followups are a bit of a different beast with historically stronger restrictions on the presence of the References header that e-mail is only now starting to adopt. I still feel like this could be dealt with definitionally, but followups in the Usenet sense have References headers, period. It's not just a SHOULD. And I don't feel like that's a divergence from RFC 2822, since RFC 2822 doesn't deal with Usenet followups and the semantics of the References header haven't changed. -- Russ Allbery (rra@stanford.edu) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RI0fww060281 for ; Wed, 27 Apr 2005 11:00:41 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3RI0fZD060280 for ietf-usefor-skb; Wed, 27 Apr 2005 11:00:41 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from a.mail.peak.org (a.mail.peak.org [69.59.192.41]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RI0eZI060268 for ; Wed, 27 Apr 2005 11:00:40 -0700 (PDT) (envelope-from stanley@peak.org) Received: from a.shell.peak.org ([69.59.192.81]) by a.mail.peak.org (8.12.10/8.12.8) with ESMTP id j3RI0XgU077455 (version=TLSv1/SSLv3 cipher=EDH-RSA-DES-CBC3-SHA bits=168 verify=NO) for ; Wed, 27 Apr 2005 11:00:35 -0700 (PDT) Date: Wed, 27 Apr 2005 11:00:33 -0700 (PDT) From: John Stanley X-X-Sender: stanley@a.shell.peak.org To: ietf-usefor@imc.org Subject: Re: Suggested References texts Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Spam-Score: 0 () X-Scanned-By: MIMEDefang 2.39 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: "Charles Lindsey" : >Let us recap the history here. It is accepted, I believe, (using Frank >Ellermann's notation) Frank's notation is confusing. Does '=>' mean "is a member of", "implies", or something else? Simple english works so much better here, I suggest it be used. > followup -> references header A followup requires a References header? (Yes under RFC1036, no under RFC2822). A followup contains a References header? A followup is defined by a References header? What? > no references header -> not a followup Ahhh, perhaps '=>' means "means"? (Is '->' the same as '=>'?) So, no References means an article is not a followup? Well, yes, according to RFC1036, not according to USEFOR draft 03 and RFC2822. >and at one time we had > references header -> followup And then you changed it. >But then, some months back, it was pointed out that some writers of multipart >FAQs regularly used References headers to join the later parts to the >earlier ones, and RFC 2046 suggests the use of References headers in >message/partial. And there were other examples too. None of which required the change you made. >Did we want to forbid these practices, and would they actually disappear >even if we did? So the choice was to evicerate the References header instead of expand the definition of followup trivially. >Inevitably, this changes the information that a reading agent can deduce >when it sees a References header. Wrong. Making References legal in non-followups changed the information that could be deduced. "Is this a followup" cannot be answered using your changes. It could before you made them. >It can no longer deduce "this is a followup". I keep telling you this, and yet, you keep telling me that this is not a change. >So you can no longer assert "references header -> followup". The question >is 'Does this matter?'. I don't see why it should. That is your value judgement, not a fact. In fact, many people do think it matters, and many people want that ability. Richard Clayton was just here arguing that it is important, and that he even considers it an interoperability issue. Do you understand the difference between saying "I don't think it is important and thus shouldn't be possible" and "I don't think it is important enough to merit RFC2119 mandates?" But now that it is gone, we need to unify on one position. Either it is or it isn't important. One document written as if it isn't and one written as if it is is just ridiculous. Pick one and stick with it for more than a week. Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RHvsqN060105 for ; Wed, 27 Apr 2005 10:57:54 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3RHvsUC060104 for ietf-usefor-skb; Wed, 27 Apr 2005 10:57:54 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtp3.Stanford.EDU (smtp3.Stanford.EDU [171.67.16.138]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RHvrKA060098 for ; Wed, 27 Apr 2005 10:57:54 -0700 (PDT) (envelope-from rra@stanford.edu) Received: from windlord.stanford.edu (windlord.Stanford.EDU [171.64.19.147]) by smtp3.Stanford.EDU (8.12.11/8.12.11) with SMTP id j3RHvpWe023780 for ; Wed, 27 Apr 2005 10:57:51 -0700 Received: (qmail 9748 invoked by uid 1000); 27 Apr 2005 17:57:51 -0000 To: ietf-usefor@imc.org Subject: Re: Obsolete vs. obsolescent In-Reply-To: <426F4F7C.7385@xyzzy.claranet.de> (Frank Ellermann's message of "Wed, 27 Apr 2005 10:38:20 +0200") References: <200504111212.j3BCChM14516@clerew.man.ac.uk> <87psx1gnjc.fsf@windlord.stanford.edu> <87fyxvx1m6.fsf@windlord.stanford.edu> <87mzs2a1hu.fsf@windlord.stanford.edu> <87wtqxui6s.fsf@windlord.stanford.edu> <87d5so2bpk.fsf@windlord.stanford.edu> <42697BB6.6074@xyzzy.claranet.de> <426BC455.2040108@isode.com> <426E4B45.7010205@oceana.com> <426F4F7C.7385@xyzzy.claranet.de> From: Russ Allbery Organization: The Eyrie Date: Wed, 27 Apr 2005 10:57:51 -0700 Message-ID: <87k6moksps.fsf@windlord.stanford.edu> User-Agent: Gnus/5.1007 (Gnus v5.10.7) XEmacs/21.4 (Jumbo Shrimp, linux) MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Frank Ellermann writes: > IMHO all that should be done is to update this statement: > | As a result, this header is to be regarded as obsolete, and > | it will likely be removed entirely in a future version of > | this standard. > As a result, this header is to be regarded as obsolescent. > Servers and UAs MUST accept but SHOULD ignore it, and SHOULD > NOT generate it. I dislike saying that servers MUST accept a header, since servers are not required to accept *any* article. I would also leave in the bit about removing it in a future version of the standard. > Subject: cmsg is not mentioned anywhere in usefor-03, that's a > serious bug. It's mentioned in usepro-03, but it only says > that this is now obsolete. That's of course FUBAR, it's a > severe security violation. We cannot jump from RfC 1036 to a > completely different procedure at date X. AFAIK RfCs must not > have a flag day. I think USEFOR does need an explicit mention that subjects starting with cmsg should not be considered control messages, to document the change from RFC 1036. -- Russ Allbery (rra@stanford.edu) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RHoTxC059708 for ; Wed, 27 Apr 2005 10:50:29 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3RHoTe6059707 for ietf-usefor-skb; Wed, 27 Apr 2005 10:50:29 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from mail1.panix.com (mail1.panix.com [166.84.1.72]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RHoS3g059694 for ; Wed, 27 Apr 2005 10:50:28 -0700 (PDT) (envelope-from sethb@panix.com) Received: from panix5.panix.com (panix5.panix.com [166.84.1.5]) by mail1.panix.com (Postfix) with ESMTP id CA10558ACC for ; Wed, 27 Apr 2005 13:50:27 -0400 (EDT) Received: (from sethb@localhost) by panix5.panix.com (8.11.6p3/8.8.8/PanixN1.1) id j3RHoR604675; Wed, 27 Apr 2005 13:50:27 -0400 (EDT) Date: Wed, 27 Apr 2005 13:50:27 -0400 (EDT) Message-Id: <200504271750.j3RHoR604675@panix5.panix.com> From: Seth Breidbart To: ietf-usefor@imc.org In-reply-to: (message from John Stanley on Wed, 27 Apr 2005 10:35:02 -0700 (PDT)) Subject: Re: Differences between RFC 2822 and Usefor (fwd) References: Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: John Stanley wrote: > However, if you are trying to argue that "followup" is an important > concept, and detecting such is an important action, then ok, maybe > it is. Does it reach the level of importance that it is an > interoperability issue? Will newsreaders break if they get an > article without a References header? Does it truly merit a MUST, or > is it only worth a SHOULD? Define "break". If you mean "crash and burn and let the magic smoke out of the computer" then no. If you mean "Do the wrong thing (from the viewpoints of both the poster and reader)" then yes. I'd prefer a MUST but would settle for a SHOULD. > And yet, not a single thing you mentioned is interoperability. In > fact, right now, news readers MUST deal with articles that do not > have a References header. An article without a References header > CANNOT be an interoperability issue, since the system would be > collapsing today. If the system continues to act, but sometimes acts wrongly, it isn't collapsing. But whatever is making it act wrongly _is_ an interoperability issue. If a random set of Subject header content was changed to "3" the system wouldn't collapse, either; so why is specifying that some headers MUST NOT be changed by transit servers an interoperability issue? Answer: That's doing _the wrong thing_. Seth Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RHlfL6059544 for ; Wed, 27 Apr 2005 10:47:41 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3RHlfDP059543 for ietf-usefor-skb; Wed, 27 Apr 2005 10:47:41 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from anchor-post-36.mail.demon.net (anchor-post-36.mail.demon.net [194.217.242.86]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RHleEK059537 for ; Wed, 27 Apr 2005 10:47:40 -0700 (PDT) (envelope-from richard@highwayman.com) Received: from gti.noc.demon.net ([195.11.55.101] helo=happyday.al.cl.cam.ac.uk) by anchor-post-36.mail.demon.net with esmtp (Exim 4.42) id 1DQqdP-000Kz9-Ji for ietf-usefor@imc.org; Wed, 27 Apr 2005 17:47:39 +0000 Message-ID: <2PvOFUX9$8bCFA94@highwayman.com> Date: Wed, 27 Apr 2005 18:46:37 +0100 To: ietf-usefor@imc.org From: Richard Clayton Subject: Re: Differences between RFC 2822 and Usefor (fwd) References: In-Reply-To: MIME-Version: 1.0 X-Mailer: Turnpike Integrated Version 5.02 M Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 In message , John Stanley writes >However, if you are trying to argue that "followup" is an important >concept, and detecting such is an important action, then ok, maybe it is. >Does it reach the level of importance that it is an interoperability >issue? in my view yes >Will newsreaders break if they get an article without a References >header? one article they will cope with -- and their heuristics may mean that they function as if the header had been present -- but if no articles have References then they will fail to function as their users currently expect. I call that broken, and so will they. >Does it truly merit a MUST, or is it only worth a SHOULD? it should be a MUST >>So it seems to me that this is a genuine inter-operability issue -- > >And yet, not a single thing you mentioned is interoperability. I suspect that's because you're thinking about servers and not about the people who read Usenet and the software that they use for that. I gave some examples of how that causes problems earlier >We already say what should be done. What change do you want made? I believe that References: should be compulsory, and that all the other clues we currently have like "Re:", the unmangled subject, a later datestamp, and "In article xxx so-and-so said" can be made optional (or in the interests of tidyness abandoned altogether) >>sounds like a cock-up :( > >Sounds like a decision and the natural consequences of that decision. > >>so best fix it then! ASAP!! > >Best determine that it is broken before demanding it be fixed. This group >has been working for a long time; have you read the history of this >discussion before telling us something we have done is broken? oh yes, I've read a lot ... all the way down the years :( ... however, now it's clear that there's the prospect of (a) the effort finishing (one way or another) and (b) posters being permitted to break "threading" (and if there is no language to compel them, then they will) then it is time to speak up - -- richard Richard Clayton They that can give up essential liberty to obtain a little temporary safety deserve neither liberty nor safety. Benjamin Franklin -----BEGIN PGP SIGNATURE----- Version: PGPsdk version 1.7.1 iQA/AwUBQm/P/ZoAxkTY1oPiEQKq8ACgkMoFxn0uNpSEhbUFVTzVnhHVjbcAnjyZ v40IGNzGWVFSpus4Jt8cEJjL =JMS3 -----END PGP SIGNATURE----- Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RHgQW1059003 for ; Wed, 27 Apr 2005 10:42:26 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3RHgQFq059002 for ietf-usefor-skb; Wed, 27 Apr 2005 10:42:26 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from a.mail.peak.org (a.mail.peak.org [69.59.192.41]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RHgP8R058984 for ; Wed, 27 Apr 2005 10:42:25 -0700 (PDT) (envelope-from stanley@peak.org) Received: from a.shell.peak.org ([69.59.192.81]) by a.mail.peak.org (8.12.10/8.12.8) with ESMTP id j3RHgJgU067732 (version=TLSv1/SSLv3 cipher=EDH-RSA-DES-CBC3-SHA bits=168 verify=NO) for ; Wed, 27 Apr 2005 10:42:19 -0700 (PDT) Date: Wed, 27 Apr 2005 10:42:19 -0700 (PDT) From: John Stanley X-X-Sender: stanley@a.shell.peak.org To: ietf-usefor@imc.org Subject: Re: References Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Spam-Score: 0 () X-Scanned-By: MIMEDefang 2.39 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: "Forrest J. Cavalier III" : >Did anyone ever advocate that 2822-reply => followup? Yes. It is pretty clear that all replies are followups. Look at the definitions we have come up with for "followup", and each one has included the set of articles that are "a response to another". Sometimes the definition also includes other relationships, but "response to another" has always been included. And that is a "reply". Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RHZ8tU058515 for ; Wed, 27 Apr 2005 10:35:08 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3RHZ8gF058514 for ietf-usefor-skb; Wed, 27 Apr 2005 10:35:08 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from b.mail.peak.org (b.mail.peak.org [69.59.192.42]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RHZ7pK058504 for ; Wed, 27 Apr 2005 10:35:08 -0700 (PDT) (envelope-from stanley@peak.org) Received: from a.shell.peak.org ([69.59.192.81]) by b.mail.peak.org (8.12.10/8.12.8) with ESMTP id j3RHZ2wU012976 (version=TLSv1/SSLv3 cipher=EDH-RSA-DES-CBC3-SHA bits=168 verify=NO) for ; Wed, 27 Apr 2005 10:35:02 -0700 (PDT) Date: Wed, 27 Apr 2005 10:35:02 -0700 (PDT) From: John Stanley X-X-Sender: stanley@a.shell.peak.org To: ietf-usefor@imc.org Subject: Re: Differences between RFC 2822 and Usefor (fwd) Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Spam-Score: 0 () X-Scanned-By: MIMEDefang 2.39 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Richard Clayton : >the readers of Usenet believe in things like threads and followups; they >like the ability to split a series of rambling conversations into >disparate piles of articles and understand each of those piles one at a >time... If you read USENET for any length of time, you will find that readers (at least posters) of USENET believe in all kinds of oddball things. Does their belief in something elevate that thing into an interoperability issue, or even something that we need to deal with? >However, there seems to be an argument being presented that when you >look deeply at the specifications the notion of followups and threads is >mainly an illusion fostered by the reading tools and one can lawfully >miss out all the headers that allow the illusion to be constructed. I'm trying to understand what you are saying here, but it is couched in such odd language that I'm not certain. "Lawfully miss out?" If I get what you are saying, then yes, of course we can refuse to mandate a header that allows illusions to be propogated. In fact, propogating illusions is not the job of a technical standard. However, if you are trying to argue that "followup" is an important concept, and detecting such is an important action, then ok, maybe it is. Does it reach the level of importance that it is an interoperability issue? Will newsreaders break if they get an article without a References header? Does it truly merit a MUST, or is it only worth a SHOULD? >So it seems to me that this is a genuine inter-operability issue -- And yet, not a single thing you mentioned is interoperability. In fact, right now, news readers MUST deal with articles that do not have a References header. An article without a References header CANNOT be an interoperability issue, since the system would be collapsing today. >The issue occurs at the level at which the readers view the newsgroups. One newsreader dealing with one article for one reader is hardly an interoperability issue. What you've just described is a display issue, and we do not deal with display issues. Readers are free to display articles as they see fit. >For that reason I think >that a standardised way of expressing which pile an article belongs in >is of very significant value. That may be true. Since we are not undefining the References header, we are not removing the 'standardized way of expressing' anything. It's just been made optional. >Yes at present some software is a bit >flaky, but that doesn't mean we should fail to say what should be done. We already say what should be done. What change do you want made? >sounds like a cock-up :( Sounds like a decision and the natural consequences of that decision. >so best fix it then! ASAP!! Best determine that it is broken before demanding it be fixed. This group has been working for a long time; have you read the history of this discussion before telling us something we have done is broken? >ah.... perhaps I could find it earlier in the "thread" ? It isn't hard. It doesn't even require a References header, which is, by the way, optional in mail, and requires knowledge of the precursor message ids in any case. When the precursor message id is not known, it is improper to fabricate one. Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RHNhMe057860 for ; Wed, 27 Apr 2005 10:23:43 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3RHNhAj057859 for ietf-usefor-skb; Wed, 27 Apr 2005 10:23:43 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from soy.epix.net (soy.epix.net [199.224.64.64]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RHNfoS057853 for ; Wed, 27 Apr 2005 10:23:42 -0700 (PDT) (envelope-from mibsoft@epix.net) Received: from [192.168.2.11] (hrbg-199-224-121-129-pppoe.dsl.hrbg.epix.net [199.224.121.129]) by soy.epix.net (8.12.10/2004120601/PL) with ESMTP id j3RHNZgE028084; Wed, 27 Apr 2005 13:23:36 -0400 (EDT) Message-ID: <426FCA98.3070307@epix.net> Date: Wed, 27 Apr 2005 13:23:36 -0400 From: "Forrest J. Cavalier III" User-Agent: Mozilla Thunderbird 0.7 (Windows/20040616) X-Accept-Language: en-us, en MIME-Version: 1.0 To: Charles Lindsey CC: ietf-usefor@imc.org Subject: Re: Suggested References texts References: <426E6E66.2070109@epix.net> In-Reply-To: Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit X-Scanned-By: MIMEDefang 2.51 on 199.224.89.133 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Charles Lindsey wrote: > Inevitably, this changes the information that a reading agent can deduce > when it sees a References header. It can no longer deduce "this is a > followup". It can only deduce "the poster wants you to treat this like a > followup" (by threading it, or by letting you retrieve the precursors by > clicking on them, or whatever other bells and whistles you provide). Do you agree we should be concentrating on how the reading agent can process articles? I think you do. And the text you suggested is flawed because it declares nothing of use for reading agents. (Here is the text:) >> | The References header is used in followups, and other articles with >> | related precursors (e.g. multipart FAQs and the fragments of a >> | message/partial), to facilitate the display or retrieval, by reading >> | and other agents, of threads of related articles. ... What does "facilitate" mean? What does "other articles" mean? Since "e.g." is not "i.e.", what other kinds of articles can have References? What stops someone from using References for other kinds of facilitation of display or retrieval? (Style sheets!) That text is useless. It doesn't even guide posters, much less readers. If people really want to allow FAQs and multi-part to get threaded, I wonder if the USEFOR text should be the following single sentence, (which cleverly does not mention followup or FAQ or multi-part.) A References header indicates the previously posted context in which the article was created. (That "the" means that you can only have at most one context per article.) I would then say to leave "followup" out of USEFOR, (and maybe out of USEPRO too, appearing in USEAGE, but I am not sure how all the documents are supposed to share duties.) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RHCLDU056981 for ; Wed, 27 Apr 2005 10:12:21 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3RHCLIw056980 for ietf-usefor-skb; Wed, 27 Apr 2005 10:12:21 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from b.mail.peak.org (b.mail.peak.org [69.59.192.42]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RHCK6f056961 for ; Wed, 27 Apr 2005 10:12:20 -0700 (PDT) (envelope-from stanley@peak.org) Received: from a.shell.peak.org ([69.59.192.81]) by b.mail.peak.org (8.12.10/8.12.8) with ESMTP id j3RHCEtO001331 (version=TLSv1/SSLv3 cipher=EDH-RSA-DES-CBC3-SHA bits=168 verify=NO) for ; Wed, 27 Apr 2005 10:12:15 -0700 (PDT) Date: Wed, 27 Apr 2005 10:12:14 -0700 (PDT) From: John Stanley X-X-Sender: stanley@a.shell.peak.org To: ietf-usefor@imc.org Subject: Re: References Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Spam-Score: 0 () X-Scanned-By: MIMEDefang 2.39 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: I said: > It cannot be mandatory in USEPRO when it is already defined to be > optional in both USEFOR and RFC2822. Alexey Melnikov : >Surely it can, because USEFOR equally applies to *all* types of USEFOR >articles and USEPRO may add additional restrictions for particular types >of articles. While USEFOR applies to all types of articles, the section of USEFOR being discussed applies to the References header, and the section of RFC2822 that USEFOR defers to talks specifically about "replies". USEFOR says "see RFC2822", and RFC2822 says References is optional in replies. In addition, since USEFOR applies to ALL articles, a definition therein that makes a header optional makes it optional for ALL articles. >See my comment above. Stop being insulting, as if I would ignore your "comment above" until I saw your command to "see it" here. Frank Ellermann : >One WG member apparently wants an equivalences in the form >of followup <=> references and followup <=> 2822-reply, but >that's obviously wrong, Yes, that's obviously wrong. Can you tell us who it is who has said this? >and the proposed alternative of no >relation at all between followup and references is also wrong. It is not wrong if that is how the standards make it, and that is the direction that we are heading by making References optional. Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RGxMmm055889 for ; Wed, 27 Apr 2005 09:59:22 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3RGxMAj055888 for ietf-usefor-skb; Wed, 27 Apr 2005 09:59:22 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from a.mail.peak.org (a.mail.peak.org [69.59.192.41]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RGxL0x055874 for ; Wed, 27 Apr 2005 09:59:21 -0700 (PDT) (envelope-from stanley@peak.org) Received: from a.shell.peak.org ([69.59.192.81]) by a.mail.peak.org (8.12.10/8.12.8) with ESMTP id j3RGxFZ1044836 (version=TLSv1/SSLv3 cipher=EDH-RSA-DES-CBC3-SHA bits=168 verify=NO) for ; Wed, 27 Apr 2005 09:59:16 -0700 (PDT) Date: Wed, 27 Apr 2005 09:59:15 -0700 (PDT) From: John Stanley X-X-Sender: stanley@a.shell.peak.org To: ietf-usefor@imc.org Subject: Re: Suggested References texts Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Spam-Score: 0 () X-Scanned-By: MIMEDefang 2.39 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Russ Allbery : >If this is a requirement, then we have to >define everything that contains a References header as a followup, You have it backwards. We define a followup by the poster's intent; we DETECT followups with the References header. There is a difference. Your definition leaves us with a circle: a followup contains a References header, and a References header must be used in a followup. Ok, what's a followup? Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RGFJTA051782 for ; Wed, 27 Apr 2005 09:15:19 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3RGFJYd051781 for ietf-usefor-skb; Wed, 27 Apr 2005 09:15:19 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from lon-mail-3.gradwell.net (lon-mail-3.gradwell.net [193.111.201.127]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RGFHeX051773 for ; Wed, 27 Apr 2005 09:15:18 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from host81-144-72-27.midband.mdip.bt.net ([81.144.72.27]) by lon-mail-3.gradwell.net with esmtp (Gradwell gwh-smtpd 1.181) id 426fba94.12d8a.216 for ietf-usefor@imc.org; Wed, 27 Apr 2005 17:15:16 +0100 (envelope-sender ) Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3RGCHU09009 for ietf-usefor@imc.org; Wed, 27 Apr 2005 17:12:17 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20740 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: References Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: <4266EF35.21A5@xyzzy.claranet.de> <87mzrtt1fh.fsf@windlord.stanford.edu> <426E4F27.2020607@oceana.com> Date: Wed, 27 Apr 2005 13:04:25 GMT Lines: 31 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In <426E4F27.2020607@oceana.com> Ken Murchison writes: >I disagree. It is my opinion (and I believe our Chair's) that USEFOR >only document the syntax of an article. The use or non-use of any >particular "optional" header such as References belongs in either USEPRO > or USEAGE. It also documents "semantics" - that is material that explains the "meaning" of the header contents, and other explanatory/introductory material, as in RFC 2822. See my reply to Forrest. For sure, USEPRO tells you how to construct a "followup", but if there is to be mention of "other" uses of References, then there is no obvious place in USEPRO to mention them (unless you can suggest one). The essential semantics of the References header, of course, is that it contains a list of "precursors", and the special significance of the order in which they are listed. But that in turn relies on a careful definition of "precursor" (which is another possible place to point out that they apply to more than "followups"). -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RGFJqY051793 for ; Wed, 27 Apr 2005 09:15:19 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3RGFJv7051792 for ietf-usefor-skb; Wed, 27 Apr 2005 09:15:19 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from lon-mail-3.gradwell.net (lon-mail-3.gradwell.net [193.111.201.127]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RGFIJn051775 for ; Wed, 27 Apr 2005 09:15:19 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from host81-144-72-27.midband.mdip.bt.net ([81.144.72.27]) by lon-mail-3.gradwell.net with esmtp (Gradwell gwh-smtpd 1.181) id 426fba95.12d8a.217 for ietf-usefor@imc.org; Wed, 27 Apr 2005 17:15:17 +0100 (envelope-sender ) Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3RGCGK09003 for ietf-usefor@imc.org; Wed, 27 Apr 2005 17:12:16 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20739 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: Suggested References texts Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: <426E6E66.2070109@epix.net> Date: Wed, 27 Apr 2005 12:04:07 GMT Lines: 125 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In <426E6E66.2070109@epix.net> "Forrest J. Cavalier III" writes: >Charles Lindsey wrote: >> I think we are now sufficiently agreed about the References header (except >> perhaps for John) to write some text. >Point of order. The chair must declare consensus. Since when has the Chair's permission been required before a member of this WG can suggest some text? >If USEFOR is going to be syntax only, then there is no place for "purpose of >the References" header that can be vanilla enough to be in USEFOR. There are two issues here. A. Should USEFOR including brief explanatory wording for each header to say what its purpose is? Remember that USEFOR is semantics as well as syntax. B. If so, then what should be said when introducing the References header? Dealing with A first, RFC 2822 regularly provides such explanatory wording. The "Subject:" field ... contains a short string identifying the topic of the message. These three fields are intended to have only human-readable content ... The originator fields indicate the mailbox(es) of the source of the message. The "From:" field specifies the author(s) of the message, ... And for the References field it says: The "In-Reply-To:" and "References:" fields are used when creating a reply to a message. Note: Some implementations parse the "References:" field to display the "thread of the discussion". ... And USEFOR also contains similar descriptive material (though often it simply inherits what RFC 2822 had already said): The Path header indicates the route taken by an article since its injection into the Netnews system. The Injection-Date header contains ... Its purpose is to prevent the reinjection into the news stream of "stale" articles which have already expired by the time they arrive at some relaying or serving agent. So there is plenty of precedent for incorporating such material. So now to issue B: >> | The References header is used in followups, and other articles with >> | related precursors (e.g. multipart FAQs and the fragments of a >> | message/partial), to facilitate the display or retrieval, by reading >> | and other agents, of threads of related articles. ... >This isn't acceptable wording at all, especially not for USEFOR. A >news client must be able to determine "followup" by looking at an article, >or the term is meaningless and should not appear in any standards document. Let us recap the history here. It is accepted, I believe, (using Frank Ellermann's notation) followup -> references header no references header -> not a followup and at one time we had references header -> followup But then, some months back, it was pointed out that some writers of multipart FAQs regularly used References headers to join the later parts to the earlier ones, and RFC 2046 suggests the use of References headers in message/partial. And there were other examples too. Did we want to forbid these practices, and would they actually disappear even if we did? Far from forbidding them, the view expressed (without dissent, so far as I recall) was that these practices were useful, and should perhaps even be encouraged. Essentially, anyone incorporating a References header for such reasons was in effect inviting reading agents (those that did threading anyway) to present the various parts of the FAQ together and in the correct order (irrespective of their order of arrival, differences in Subject, or order of their Date headers). Inevitably, this changes the information that a reading agent can deduce when it sees a References header. It can no longer deduce "this is a followup". It can only deduce "the poster wants you to treat this like a followup" (by threading it, or by letting you retrieve the precursors by clicking on them, or whatever other bells and whistles you provide). So you can no longer assert "references header -> followup". The question is 'Does this matter?'. I don't see why it should. Why is is necessary for a reading agent to be able to distinguish between (conventional) followups and multipart FAQs (note that humans can still distinguish them easily)? In any case, if this practice continues (as it surely will), there is nothing that can be done about it, since the References header is the only tool available. Now there have been suggestions to "clean it up" by artificially redefining the term "followup" to include any article that has a References header. But that is all "smoke and mirrors" - it makes it appear that you can detect "followups", but in reality it makes no difference at all as to what reading agents can or cannot actually do. Yes, if the WG wants that revised definition, it can be done, but I do not see the point since it would make no technical difference to what actually happens. -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RBIQmj000179 for ; Wed, 27 Apr 2005 04:18:26 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3RBIQNx000178 for ietf-usefor-skb; Wed, 27 Apr 2005 04:18:26 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from gundel.de.clara.net (gundel.de.clara.net [212.82.225.86]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RBIOh4000156 for ; Wed, 27 Apr 2005 04:18:25 -0700 (PDT) (envelope-from nobody@xyzzy.claranet.de) Received: from [212.82.251.220] (helo=xyzzy) by gundel.de.clara.net with smtp (Exim 4.30; FreeBSD) id 1DQklm-000PmB-Rf for ietf-usefor@imc.org; Wed, 27 Apr 2005 13:31:55 +0200 To: ietf-usefor@imc.org Message-ID: <426F6107.666C@xyzzy.claranet.de> Date: Wed, 27 Apr 2005 11:53:11 +0200 From: Frank Ellermann Organization: X-Mailer: Mozilla 3.0 (OS/2; U) MIME-Version: 1.0 Subject: Again broken (was: Fixed) References: <4255153E.3080500@isode.com> <42594879.3B92@xyzzy.claranet.de> <425A8A63.268A@xyzzy.claranet.de> <425C607E.1B5A@xyzzy.claranet.de> <425D8031.D57@xyzzy.claranet.de> <426BB63D.3060506@isode.com> Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Charles Lindsey wrote: ~~~ begin quoted ABNF ~~~ message-id = "Message-ID:" SP [FWS] msg-id [FWS] CRLF msg-id = "<" id-left "@" id-right ">" ; maximum length is 250 octets id-left = dot-atom-text / no-fold-quote id-right = dot-atom-text / no-fold-literal no-fold-quote = DQUOTE ( "." *mqtext / *mqtext "." / *mqtext mqspecial *mqtext ) DQUOTE mqtext = atext / "." / mqspecial mqspecial = "(" / ")" / ; same as specials except "<" / ; "\" and DQUOTE quoted "[" / "]" / ; "." doubled and ">" omitted ":" / ";" / "@" / "," / ".." / "\\" / "\" DQUOTE no-fold-literal = "[" *( mdtext / "\[" / "\]" / "\\" ) "]" mdtext = %d33-61 / ; The rest of the US-ASCII %d63-90 / ; characters not including %d94-126 ; ">", "[", "]", or "\" ~~~ end of quoted sytax ~~~ Is that as you want it ? I'm not sure about Alexey's external expert for 2822bis, so I think the most simple solution to get an expert review for the two proposals is to post them on the rfc-822 mailing list. Further proposed text: > The msg-id-core MUST NOT be more than 250 octets in length. s/msg-id-core/msg-id/g >| Also note that this updated ABNF applies wherever is used, >| including the References header discussed in Section 3.2.1 and the >| Supersedes header discussed in Section 3.2.5. IMHO that's obvious and needs no "note", excl. References: the two other cases (Supersedes: and Control: cancel) don't exist in mail. BTW, in Control: you have the same syntax problem as in the References:, you separate the arguments by an _optional_ [FWS]: Control: cancel Control: newgroupthis.is.new Control: rmgroupthis.is.old You need a mandatory FWS in 3.2.4. "value" is apparently not defined in usefor-03 or 2822. "token" is apparently also not defined in usefor-03 or 2822. Did anybody try an ABNF checker on usefor-03 ? (I didn't) >| NOTE: It in RECOMMENDED in [RFC 2822] that, for ensuring >| global uniqueness, the be some domain identifier It's unnecessary to quote that RECOMMENDATION, least of all the "some domain identifier". Out of its original very convoluted RfC 2822 context this quote is in fact _worse_ than RfC 2822. RfC 1036 or s-o-1036 is _much_ better than this sad parody of a Message-ID. Usenet is better off without this idea, it is harmful, it should never become an RfC. IIRC we had a "pseudo-consensus" (= you and me) about domain- literals in addresses, Message-IDs, and Path-identifiers, i.e. say something like STD 11 without shouting: | Note: THE USE OF DOMAIN-LITERALS IS STRONGLY DISCOURAGED. Apparently that didn't make it yet into usefor-03 or usepro-03, it's roughly equivalent to a SHOULD NOT mentioning the "domain" in all "addr-spec" found in any From: / Sender: / Reply-To: / Approved:, the "id-right" (your 2822-terminology) of a msg-id, the "path-identity", do we need it anywhere else ? Bye, Frank Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RBEUog098132 for ; Wed, 27 Apr 2005 04:14:30 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3RBEUKR098131 for ietf-usefor-skb; Wed, 27 Apr 2005 04:14:30 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from lentil.epix.net (lentil.epix.net [199.224.64.67]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RBET1h098097 for ; Wed, 27 Apr 2005 04:14:29 -0700 (PDT) (envelope-from mibsoft@epix.net) Received: from [192.168.2.11] (hrbg-199-224-121-129-pppoe.dsl.hrbg.epix.net [199.224.121.129]) by lentil.epix.net (8.12.10/2004120601/PL) with ESMTP id j3RBEPl8015331 for ; Wed, 27 Apr 2005 07:14:26 -0400 (EDT) Message-ID: <426F7411.2010903@epix.net> Date: Wed, 27 Apr 2005 07:14:25 -0400 From: "Forrest J. Cavalier III" User-Agent: Mozilla Thunderbird 0.7 (Windows/20040616) X-Accept-Language: en-us, en MIME-Version: 1.0 To: ietf-usefor@imc.org Subject: Re: References References: <426E579F.1060809@oceana.com> <426F3B14.54C1@xyzzy.claranet.de> In-Reply-To: <426F3B14.54C1@xyzzy.claranet.de> Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit X-Scanned-By: MIMEDefang 2.51 on 199.224.89.153 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Frank Ellermann wrote: > One WG member apparently wants an equivalences in the form > of followup <=> references and followup <=> 2822-reply, but > that's obviously wrong, and the proposed alternative of no > relation at all between followup and references is also wrong. Did anyone ever advocate that 2822-reply => followup? It isn't one WG member that wants this: followup <=> references I think it is a requirement that a news client be able to determine "followup" by inspecting the article headers. If it is not a requirement, then the term "followup" is meaningless and should not appear in any document. In support of this, I point that the References header is for the use of the READER, not the POSTER. The poster already knows how all the articles are related. Also, if not mandating anything required by the READER for interoperability, how can a document merit anything stronger than a "MAY add a References header"? (And if it is this weak, what's the point?) Frank, you allowed that followup => references. If you don't hold that references => followup, then does the standard allow an article to reference more than one thread? I think that allowing References to appear in an article which is not a reply to a single article (which I thought was the term "followup"), leads to undecidable actions about threading. What is the READER agent supposed to do in that situation? Is there a term for articles that can have references but are not followups? If not, why the special term "followup"? What does the apperance of References header mean, from a READER perspective? Sure it is POSSIBLE to define a standard which describes unambiguous construction of an article, but not unambiguous interpretation of that article. But I think that's ludicrous. Is there any other IETF standard that does this? Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RAi8kx086346 for ; Wed, 27 Apr 2005 03:44:09 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3RAi8Ra086344 for ietf-usefor-skb; Wed, 27 Apr 2005 03:44:08 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from anchor-post-35.mail.demon.net (anchor-post-35.mail.demon.net [194.217.242.85]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RAi78D086331 for ; Wed, 27 Apr 2005 03:44:08 -0700 (PDT) (envelope-from richard@highwayman.com) Received: from gti.noc.demon.net ([195.11.55.101] helo=happyday.al.cl.cam.ac.uk) by anchor-post-35.mail.demon.net with esmtp (Exim 4.42) id 1DQjyd-000EcH-Hq for ietf-usefor@imc.org; Wed, 27 Apr 2005 10:41:07 +0000 Message-ID: Date: Wed, 27 Apr 2005 11:43:33 +0100 To: ietf-usefor@imc.org From: Richard Clayton Subject: Re: Differences between RFC 2822 and Usefor (fwd) References: In-Reply-To: MIME-Version: 1.0 X-Mailer: Turnpike Integrated Version 5.02 M Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 In message , John Stanley writes >We have two (sane) options: I come at this as a newsreading client designer, author and implementer >1. RFC1036 was correct. Being able to identify "followups" is so important >that the header that allows this merits RFC2119 mandates. the readers of Usenet believe in things like threads and followups; they like the ability to split a series of rambling conversations into disparate piles of articles and understand each of those piles one at a time... of course we normally call the piles "threads" but a thread is always an artificial construct because it cannot be reliably constructed without a complete set of articles to hand and Usenet never guarantees that you have this -- so "piles" is rather closer to reality :) there are flaws with a divide and conquer approach to newsgroup reading, not least if you check a newsgroup for new material faster than new articles flow in -- but in general it has been found to work; and we can see the concept being copied into other media such as web-based forums. viz: it may be flawed, but it's the best UI model we currently have However, there seems to be an argument being presented that when you look deeply at the specifications the notion of followups and threads is mainly an illusion fostered by the reading tools and one can lawfully miss out all the headers that allow the illusion to be constructed. But even if the illusion is built on sand (and my experience tells me how many heuristics you need to stabilise the sand), it still has considerable practical value >2. RFC1036 was incorrect. Being able to identify "followups" is not >important enough to merit RFC2119 language and the header that allows this >is thus optional. if there is no standardised way of making the piles of articles in a consistent manner then as time goes on and posting software diverges from traditional working (and there is no document to guide them otherwise) those heuristics I mentioned above will have less and less to work with. ie: the piles will be different in different software. This will lead to considerable confusion amongst the readers of the newsgroup. Some will believe that a particular article belongs in one pile, some in another. Some will miss the article altogether because they have used "kill" features of their software to exclude some of the piles as uninteresting. They will express this confusion when they write further articles. Most know not to say "the thread above this one" to express article relationships -- but they often say "earlier in this thread" to say where to find other material... So it seems to me that this is a genuine inter-operability issue -- not at the protocol levels : the articles get to the far end just fine and the servers will serve them perfectly well. The issue occurs at the level at which the readers view the newsgroups. For that reason I think that a standardised way of expressing which pile an article belongs in is of very significant value. Yes at present some software is a bit flaky, but that doesn't mean we should fail to say what should be done. >Unfortunately for position 1, we've already removed the mandate. sounds like a cock-up :( so best fix it then! ASAP!! >Other people can read it the first time I send it; you >ought to be able to, too. ah.... perhaps I could find it earlier in the "thread" ? :) - -- richard @ highwayman . com "Nothing seems the same Still you never see the change from day to day And no-one notices the customs slip away" -----BEGIN PGP SIGNATURE----- Version: PGPsdk version 1.7.1 iQA/AwUBQm9s1ZoAxkTY1oPiEQKVqwCeNf1Jz07FuhsHwNLToDuA/9n1+K0AniMv tuxgFCozhLxFHv8oL/P/Lmz3 =2oPP -----END PGP SIGNATURE----- Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RAh80x085887 for ; Wed, 27 Apr 2005 03:43:08 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3RAh8op085886 for ietf-usefor-skb; Wed, 27 Apr 2005 03:43:08 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from ciao.gmane.org (main.gmane.org [80.91.229.2]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3RAh50T085859 for ; Wed, 27 Apr 2005 03:43:06 -0700 (PDT) (envelope-from usenet-format@gmane.org) Received: from list by ciao.gmane.org with local (Exim 4.43) id 1DQju9-0002J9-Ve for ietf-usefor@imc.org; Wed, 27 Apr 2005 12:36:29 +0200 Received: from 212.82.251.220 ([212.82.251.220]) by main.gmane.org with esmtp (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Wed, 27 Apr 2005 12:36:29 +0200 Received: from nobody by 212.82.251.220 with local (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Wed, 27 Apr 2005 12:36:29 +0200 X-Injected-Via-Gmane: http://gmane.org/ To: ietf-usefor@imc.org From: Frank Ellermann Subject: Re: Fixed Date: Wed, 27 Apr 2005 12:34:48 +0200 Organization: Lines: 44 Message-ID: <426F6AC8.5045@xyzzy.claranet.de> References: <4255153E.3080500@isode.com> <42594879.3B92@xyzzy.claranet.de> <425A8A63.268A@xyzzy.claranet.de> <425C607E.1B5A@xyzzy.claranet.de> <425D8031.D57@xyzzy.claranet.de> <426BB63D.3060506@isode.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Complaints-To: usenet@sea.gmane.org X-Gmane-NNTP-Posting-Host: 212.82.251.220 X-Mailer: Mozilla 3.0 (OS/2; U) Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Alexey Melnikov wrote: >> the general principle that we do not change anything in RFC >> 2822 that we do not need to change. > I concur with your preference. I don't. The RfC 2822 concept of Message-IDs is "wrong", and the RfC 1036 concept is better. Of course at least I believe that the RfC 2822 concept is in fact still the same. But as far as I can tell it at least Charles thinks that it's not. I'm not sure what he believes what an "id-right" is. > I suggest they talk to Pete Resnick first. As far as I > remember he has already agreed to publish 2822bis. I don't recall any discussion about an 2822bis in the past 12 months on the 822-list, but I could post the two proposals for the _equivalent_ msg-id syntax there (and the reason for this controversy, "id-right" vs. "id-domain"). I've added an "id-fication" of my proposal below, bye, Frank ~~~ begin ~~~ msg-id = "<" id-local "@" id-domain ">" id-local = dot-atom-text / ( DQUOTE id-quote DQUOTE ) id-quote = ( "." [id-text] ) / ( [id-text] "." ) / ( [id-text] id-special [id-text] ) id-text = 1*( atext / "." / id-special ) id-special = "(" / ")" / "," / ; all specials, minus ">", "[" / "]" / "@" / ; minus DQUOTE, minus "\", ":" / ";" / "<" / ; minus single ".", plus: ".." / "\\" / ( "\" DQUOTE ) id-domain = dot-atom-text / ("[" id-literal "]") id-literal = 1*( %d33-61 / ; printable ASCII minus %d63-90 / ; ">", "[", "\", "]" %d94-126 / ; plus "\[", "\\, "\]" "\[" / "\\" / "\]" ) ~~~ end ~~~ Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3R8eZTh032765 for ; Wed, 27 Apr 2005 01:40:35 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3R8eZLJ032764 for ietf-usefor-skb; Wed, 27 Apr 2005 01:40:35 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from ciao.gmane.org (main.gmane.org [80.91.229.2]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3R8eXoZ032744 for ; Wed, 27 Apr 2005 01:40:33 -0700 (PDT) (envelope-from usenet-format@gmane.org) Received: from list by ciao.gmane.org with local (Exim 4.43) id 1DQi0K-0005dk-HY for ietf-usefor@imc.org; Wed, 27 Apr 2005 10:34:44 +0200 Received: from 212.82.251.220 ([212.82.251.220]) by main.gmane.org with esmtp (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Wed, 27 Apr 2005 10:34:44 +0200 Received: from nobody by 212.82.251.220 with local (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Wed, 27 Apr 2005 10:34:44 +0200 X-Injected-Via-Gmane: http://gmane.org/ To: ietf-usefor@imc.org From: Frank Ellermann Subject: Re: Obsolete vs. obsolescent Date: Wed, 27 Apr 2005 10:38:20 +0200 Organization: Lines: 46 Message-ID: <426F4F7C.7385@xyzzy.claranet.de> References: <200504111212.j3BCChM14516@clerew.man.ac.uk> <87psx1gnjc.fsf@windlord.stanford.edu> <87fyxvx1m6.fsf@windlord.stanford.edu> <87mzs2a1hu.fsf@windlord.stanford.edu> <87wtqxui6s.fsf@windlord.stanford.edu> <87d5so2bpk.fsf@windlord.stanford.edu> <42697BB6.6074@xyzzy.claranet.de> <426BC455.2040108@isode.com> <426E4B45.7010205@oceana.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Complaints-To: usenet@sea.gmane.org X-Gmane-NNTP-Posting-Host: 212.82.251.220 X-Mailer: Mozilla 3.0 (OS/2; U) Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Ken Murchison wrote: > Care to craft some text and tell me where you'd like to see > it? IMHO all that should be done is to update this statement: | As a result, this header is to be regarded as obsolete, and | it will likely be removed entirely in a future version of | this standard. As a result, this header is to be regarded as obsolescent. Servers and UAs MUST accept but SHOULD ignore it, and SHOULD NOT generate it. For the obsolete headers in 3.3 you already have a definition: | Articles containing these headers MUST NOT be generated. | Persons writing new agents SHOULD ignore any former meanings | these headers. That should be "of these headers", but the sentence makes no sense, the handling depends on the header field. Probably new agents MUST ignore the former meaning of the obsolete header fields. But for Also-Control I'd opt for the same procedure as for a Subject: cmsg without a corresponding Control: Oops, "new" problem: Subject: cmsg is not mentioned anywhere in usefor-03, that's a serious bug. It's mentioned in usepro-03, but it only says that this is now obsolete. That's of course FUBAR, it's a severe security violation. We cannot jump from RfC 1036 to a completely different procedure at date X. AFAIK RfCs must not have a flag day. Even if it's only in theory (which it's not, Subject: cmsg has serious difficulties to propagate, the last time I checked it Google ignored a Subject: Re: cmsg, and that's not the only oddity with a Subject: cmsg) Maybe publishing Henry's s-o-1036 first could formally save this situation, otherwise Subject: cmsg is IMO a showstopper. Bye, Frank Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3R7uikl015581 for ; Wed, 27 Apr 2005 00:56:44 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3R7uiog015580 for ietf-usefor-skb; Wed, 27 Apr 2005 00:56:44 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from ciao.gmane.org (main.gmane.org [80.91.229.2]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3R7ugGJ015568 for ; Wed, 27 Apr 2005 00:56:43 -0700 (PDT) (envelope-from usenet-format@gmane.org) Received: from list by ciao.gmane.org with local (Exim 4.43) id 1DQhJT-0000Y4-RO for ietf-usefor@imc.org; Wed, 27 Apr 2005 09:50:27 +0200 Received: from 212.82.251.220 ([212.82.251.220]) by main.gmane.org with esmtp (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Wed, 27 Apr 2005 09:50:27 +0200 Received: from nobody by 212.82.251.220 with local (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Wed, 27 Apr 2005 09:50:27 +0200 X-Injected-Via-Gmane: http://gmane.org/ To: ietf-usefor@imc.org From: Frank Ellermann Subject: Re: Suggested References texts Date: Wed, 27 Apr 2005 09:50:20 +0200 Organization: Lines: 65 Message-ID: <426F443C.6561@xyzzy.claranet.de> References: Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Complaints-To: usenet@sea.gmane.org X-Gmane-NNTP-Posting-Host: 212.82.251.220 X-Mailer: Mozilla 3.0 (OS/2; U) Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Charles Lindsey wrote: > I have also fixed the syntax error (the CFWS between the > s is obligatory), and have continued the liberty of > using rather than . Good (but see below). > what RFC 2046 suggests is woefully inadequate > A "followup" is an article containing a response to the > contents of an earlier article. It will always include a > References header pointing to that earlier article and any > other "precursors". Strong NAK to any change of <426E4FC9.7020903@oceana.com> We know that References do _not_ always include 'any other "precursors"', in fact there's a complete chapter explaining how to trim the References. > An article is a "precursor" of some later article which is a > followup to it Strong NAK to any change of <426E4FC9.7020903@oceana.com> > or which is otherwise intended to be grouped with it for > purposes of display (e.g. as a part of a multipart posting > such as a FAQ). Strong NAK, the different parts of a multipart FAQ might be related by References, but this relation is not necessarily that of a "precursor". A "precursor" is simply the opposite of a "followup", so if A is the precursor of B, then B is a followup of A. There's no case where A is the precursor of B, but B is no followup of A, that would be madness. > Within USEFOR: Strong NAK, after <426E4FC9.7020903@oceana.com> this issue is settled in Usefor, we only have to fix the syntax. >| references = "References:" SP [CFWS] 1*(msg-id [CFWS]) CRLF Strong NAK to anything below this line. And the syntax is not correct, we wanted at least one WSP to separate the msg-id. references = "References:" SP msg-id-list CRLF msg-id-list = [CFWS] msg-id *( CFWS msg-id ) [CFWS] That's the simple variant where a comment is enough to separate the Message-IDs as in (c) but adjacent Message-IDs as in aren't allowed. [Usepro] > If the resulting References header is excessively long, it > MAY be trimmed, but the first and the last two message > identifiers MUST NOT be removed. I'd go for a SHOULD here, because some news servers and tools really hate it if the References are too long, I've seen cases where the References were truncated Bye, Frank Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3R7DsBj095446 for ; Wed, 27 Apr 2005 00:13:54 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3R7DsdU095445 for ietf-usefor-skb; Wed, 27 Apr 2005 00:13:54 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from ciao.gmane.org (main.gmane.org [80.91.229.2]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3R7DpKk095422 for ; Wed, 27 Apr 2005 00:13:52 -0700 (PDT) (envelope-from usenet-format@gmane.org) Received: from list by ciao.gmane.org with local (Exim 4.43) id 1DQgdd-0004JE-D1 for ietf-usefor@imc.org; Wed, 27 Apr 2005 09:07:13 +0200 Received: from 212.82.251.220 ([212.82.251.220]) by main.gmane.org with esmtp (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Wed, 27 Apr 2005 09:07:13 +0200 Received: from nobody by 212.82.251.220 with local (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Wed, 27 Apr 2005 09:07:13 +0200 X-Injected-Via-Gmane: http://gmane.org/ To: ietf-usefor@imc.org From: Frank Ellermann Subject: Re: References Date: Wed, 27 Apr 2005 09:11:16 +0200 Organization: Lines: 20 Message-ID: <426F3B14.54C1@xyzzy.claranet.de> References: <426E579F.1060809@oceana.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Complaints-To: usenet@sea.gmane.org X-Gmane-NNTP-Posting-Host: 212.82.251.220 X-Mailer: Mozilla 3.0 (OS/2; U) Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Ken Murchison wrote: > I don't see a contradiction here There's no contradiction. We have followup => references, and after you've modified the definition as indicated in... <426E4FC9.7020903@oceana.com> | Done. ...this issue is settled for Usefor-04 and the future RfC, as far as I'm concerned. One WG member apparently wants an equivalences in the form of followup <=> references and followup <=> 2822-reply, but that's obviously wrong, and the proposed alternative of no relation at all between followup and references is also wrong. Bye, Frank Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3QKSHQK049791 for ; Tue, 26 Apr 2005 13:28:17 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3QKSHwR049790 for ietf-usefor-skb; Tue, 26 Apr 2005 13:28:17 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from a.mail.peak.org (a.mail.peak.org [69.59.192.41]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3QKSGwt049763 for ; Tue, 26 Apr 2005 13:28:16 -0700 (PDT) (envelope-from stanley@peak.org) Received: from a.shell.peak.org ([69.59.192.81]) by a.mail.peak.org (8.12.10/8.12.8) with ESMTP id j3QKS8Z1023583 (version=TLSv1/SSLv3 cipher=EDH-RSA-DES-CBC3-SHA bits=168 verify=NO) for ; Tue, 26 Apr 2005 13:28:10 -0700 (PDT) Date: Tue, 26 Apr 2005 13:28:08 -0700 (PDT) From: John Stanley X-X-Sender: stanley@a.shell.peak.org To: ietf-usefor@imc.org Subject: Re: Differences between RFC 2822 and Usefor (fwd) Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Spam-Score: 0 () X-Scanned-By: MIMEDefang 2.39 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Ken Murchison : >You have misquoted me here. The above was written by Charles. You are correct. My mistake, I'm sorry. I'm used to the standard methods of quoting material (using '>') and not this blue-line crap that doesn't cut-and-paste. I'm also used to people eliding things they are NOT responding to and not including large volumes of text just to add one line at the bottom. > Thank you for your pleasant and constructive tone. You are welcome. I'm sorry you are unhappy that you got caught trying to reword the problem to make it go away. "Should/should not" is not the issue and pretending it is is pretty insulting to the person you tried pulling that on. Excuse me for pointing out that I caught you. >Is your argument that References shouldn't be required for followups >(regardless of what USEFOR/USEPRO currently says) or that the current text >is incorrect/ambiguous/contradictory/shitty? My argument is exactly as I stated: you cannot say "this is optional" and "this is mandatory" at the same time. This is a problem; at least, I think most people view contradictions in technical standards as a problem. Perhaps you do not. The solution is as I stated: pick one and stick with it, including all the implications that go with it. >If the former, please state so. If the latter, please suggest some text >for one or more of the documents that you feel is technically sound. Before we can write text we must make the choice. To do otherwise is a complete waste of time. Writing text and THEN justifying it is backwards, and it is why we are spending so much time now. The mandate has been removed from the draft and people are busy trying to pretend it is still there or should simply be reinstated without justification. And yet, there must have been a reason it was removed. We have two (sane) options: 1. RFC1036 was correct. Being able to identify "followups" is so important that the header that allows this merits RFC2119 mandates. The implication from this is that References MUST appear in followups and MUST NOT appear elsewhere, since any other combination of RFC2119 language would not accomplish the goal of allowing identification of followups. A further implication of this choice is that we MUST document the difference from RFC2822 where the header and article format are defined. That's in USEFOR. Another implication is that we have to define followups. That's not really that hard; I've provided a proposed definition that meets current practice. 2. RFC1036 was incorrect. Being able to identify "followups" is not important enough to merit RFC2119 language and the header that allows this is thus optional. Since the only method of identifying a followup has been made optional, the concept of followup is thus not worth defining since nobody can detect them when they appear anyway. At least, not detect them with any certainty. (Do you have a mole on your left shoulder? Your messages don't contain any way of telling; fortunately, I don't care, although someone else might.) Once we have chosen the path we wish to take, we can look at how to accomplish it. Unfortunately for position 1, we've already removed the mandate. USEFOR defers to RFC2822, RFC2822 says "SHOULD" and "may contain". Putting a mandate BACK after deciding to remove it should, in a correctly operating system, require justification. We were getting by with "historical reasons" when it was a mandate copied from RFC1036; if that wasn't sufficient to keep it in place, then it won't be enough to put it back now that it is gone. I've seen no interoperability issues; do you have something to report? So, looking at RFC2119, it appears the only remaining option is 2. USEFOR already makes the header optional. The best that USEPRO can do is say "IF you include the header, it must contain X". Most of the text Charles just proposed meets that criterion. Remove the parts that refer to the specious mandate. How do we deal with "duties of a followup agent"? Well, RFC2822 talks about "replies". That's a pretty common english word, why is it unavailable to us? USEPRO can simply talk about what headers are available for a reply and what the headers that DO get used in the reply contain. E.g., "If a References header is included, it MUST contain ...". That's fine. "You MUST include" this optional header is not. And when you talk about headers like Newsgroups and Path, you can say "This MUST be present and MUST contain..." because USEFOR already says they are mandatory. Now, I've said it all again. Anyone up for a third pass? And, if you are keeping score on such things, your demand that I repeat myself or provide text, with the implications of that demand, is pretty insulting, too. Other people can read it the first time I send it; you ought to be able to, too. Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3QKENtm047613 for ; Tue, 26 Apr 2005 13:14:23 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3QKENRI047612 for ietf-usefor-skb; Tue, 26 Apr 2005 13:14:23 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from rufus.isode.com (rufus.isode.com [62.3.217.251]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3QKELBR047589 for ; Tue, 26 Apr 2005 13:14:22 -0700 (PDT) (envelope-from alexey.melnikov-usefor@isode.com) Received: from [192.168.0.3] ([62.3.217.253]) by rufus.isode.com via TCP (internal) with ESMTPA; Tue, 26 Apr 2005 21:11:51 +0100 Message-ID: <426EA086.6020400@isode.com> Date: Tue, 26 Apr 2005 21:11:50 +0100 From: Alexey Melnikov User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.7.2) Gecko/20040804 Netscape/7.2 (ax) X-Accept-Language: en-us, en MIME-Version: 1.0 To: John Stanley CC: ietf-usefor@imc.org Subject: Re: References References: In-Reply-To: Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: John Stanley wrote: >Ken Murchison : > > >>It is mandatory as stated in USEPRO. >> >> >It cannot be mandatory in USEPRO when it is already defined to be optional >in both USEFOR and RFC2822. > Surely it can, because USEFOR equally applies to *all* types of USEFOR articles and USEPRO may add additional restrictions for particular types of articles. >>I'll reiterate what I stated elsewhere in the thread. >> >> >>USEFOR documents the proper syntax of the set of headers typically used in >>news articles. >> >> >That is not what you said. You said: > > >>It is my opinion (and I believe our Chair's) that USEFOR only document the >>syntax of an article. >> >> > >The "syntax of an article" includes what headers it is required to >contain. The "syntax of a header" defines the contents of a header; the >syntax of an article defines the contents of an article. USEFOR is USENET >Format -- which is clearly not just the headers but the contents of >articles that are valid USENET articles. > >USEFOR does not just tell us the correct syntax of the From header, it >tells us that the From header is MANDATORY. It does the same for several >other headers. Thus, it is clear that USEFOR is not just "syntax of >headers", it includes "syntax of articles". > >In that job of "syntax of articles", USEFOR tells us that References is >optional. References is optional. > > See my comment above. Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3QJT3he039778 for ; Tue, 26 Apr 2005 12:29:03 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3QJT3IV039777 for ietf-usefor-skb; Tue, 26 Apr 2005 12:29:03 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from b.mail.peak.org (b.mail.peak.org [69.59.192.42]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3QJT2se039759 for ; Tue, 26 Apr 2005 12:29:02 -0700 (PDT) (envelope-from stanley@peak.org) Received: from a.shell.peak.org ([69.59.192.81]) by b.mail.peak.org (8.12.10/8.12.8) with ESMTP id j3QJSptO024401 (version=TLSv1/SSLv3 cipher=EDH-RSA-DES-CBC3-SHA bits=168 verify=NO) for ; Tue, 26 Apr 2005 12:28:57 -0700 (PDT) Date: Tue, 26 Apr 2005 12:28:51 -0700 (PDT) From: John Stanley X-X-Sender: stanley@a.shell.peak.org To: ietf-usefor@imc.org Subject: Re: References Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Spam-Score: 0 () X-Scanned-By: MIMEDefang 2.39 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Ken Murchison : > It is mandatory as stated in USEPRO. It cannot be mandatory in USEPRO when it is already defined to be optional in both USEFOR and RFC2822. > I'll reiterate what I stated elsewhere in the thread. > USEFOR documents the proper syntax of the set of headers typically used in > news articles. That is not what you said. You said: > It is my opinion (and I believe our Chair's) that USEFOR only document the > syntax of an article. The "syntax of an article" includes what headers it is required to contain. The "syntax of a header" defines the contents of a header; the syntax of an article defines the contents of an article. USEFOR is USENET Format -- which is clearly not just the headers but the contents of articles that are valid USENET articles. USEFOR does not just tell us the correct syntax of the From header, it tells us that the From header is MANDATORY. It does the same for several other headers. Thus, it is clear that USEFOR is not just "syntax of headers", it includes "syntax of articles". In that job of "syntax of articles", USEFOR tells us that References is optional. References is optional. USEFOR and USEPRO are supposed to complement each other, not contradict. Our Chair has said USEFOR is right not to talk about a mandate, although he's said that because he thinks it can be slipped in after the fact in USEPRO. I agree USEFOR is right not to talk about a mandate -- for a different reason. If you want a mandate, do it honestly and in conformance with RFC2119. Justify it and you can have it. So far, you haven't. "Easier" and "play nice" don't even come close. Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3QJC8vJ036852 for ; Tue, 26 Apr 2005 12:12:08 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3QJC8vN036851 for ietf-usefor-skb; Tue, 26 Apr 2005 12:12:08 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from a.mail.peak.org (a.mail.peak.org [69.59.192.41]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3QJC3ni036815 for ; Tue, 26 Apr 2005 12:12:07 -0700 (PDT) (envelope-from stanley@peak.org) Received: from a.shell.peak.org ([69.59.192.81]) by a.mail.peak.org (8.12.10/8.12.8) with ESMTP id j3QJBtZ1074485 (version=TLSv1/SSLv3 cipher=EDH-RSA-DES-CBC3-SHA bits=168 verify=NO) for ; Tue, 26 Apr 2005 12:11:58 -0700 (PDT) Date: Tue, 26 Apr 2005 12:11:54 -0700 (PDT) From: John Stanley X-X-Sender: stanley@a.shell.peak.org To: ietf-usefor@imc.org Subject: Suggested References texts Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Spam-Score: 0 () X-Scanned-By: MIMEDefang 2.39 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: "Charles Lindsey" : > A "followup" is an article containing a response to the contents of > an earlier article. It will always include a References header > pointing to that earlier article and any other "precursors". USEFOR and RFC2822 contradict this. Since we've already decided to follow RFC2822 in this matter, your text is wrong. RFC2822 is explicit in providing guidance for processing of replies to replies that have no References header; clearly a reply sans References header is allowed by RFC2822. > 4. If the precursor did not have a References header (F-3.2.1), the > content of the followup's References header MUST be inherited from > that of the Message-ID header of the precursor. A followup to an > article which already had a References header MUST have a > References header comprising the precursor's References header > (subject to trimming as described below) followed by CFWS and the > Message-ID header content of the precursor. RFC2822 contradicts this, and we have already decided that RFC2822 is the definition we are using. Further, these mandates have not been justified wrt RFC2119, which prohibits their use except for good cause. > NOTE: The two "MUST"s above ensure compliance with the > definition of the term "followup". This is absolutely ridiculous. RFC2119 does not allow mandates to be used just to "ensure compliance with a definition". There has to be an interoperatbility issue, and no such issue has been shown. This "MUST" is specious and violates existing standards, and contradicts our own draft regarding article format. > with the weaker recomendation using "SHOULD" applied, in [RFC > 2822], to the generation of "replies" in email. RFC2822 is not limited to "email", and it is dishonest to try to pretend that it is, especially when we defer to RFC2822 for news headers. >I think we are now sufficiently agreed about the References header (except >perhaps for John) to write some text. It is dishonest to ignore people who disagree with you, Charles. Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3QJ8Q2M036280 for ; Tue, 26 Apr 2005 12:08:26 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3QJ8Q7V036279 for ietf-usefor-skb; Tue, 26 Apr 2005 12:08:26 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtp2.Stanford.EDU (smtp2.Stanford.EDU [171.67.16.125]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3QJ8PAN036272 for ; Tue, 26 Apr 2005 12:08:25 -0700 (PDT) (envelope-from rra@stanford.edu) Received: from windlord.stanford.edu (windlord.Stanford.EDU [171.64.19.147]) by smtp2.Stanford.EDU (8.12.11/8.12.11) with SMTP id j3QJ8ODB018791 for ; Tue, 26 Apr 2005 12:08:24 -0700 Received: (qmail 23475 invoked by uid 1000); 26 Apr 2005 19:08:24 -0000 To: ietf-usefor@imc.org Subject: Re: Suggested References texts In-Reply-To: <426E6E66.2070109@epix.net> (Forrest J. Cavalier, III's message of "Tue, 26 Apr 2005 12:37:58 -0400") References: <426E6E66.2070109@epix.net> From: Russ Allbery Organization: The Eyrie Date: Tue, 26 Apr 2005 12:08:24 -0700 Message-ID: <87ekcxnyon.fsf@windlord.stanford.edu> User-Agent: Gnus/5.1007 (Gnus v5.10.7) XEmacs/21.4 (Jumbo Shrimp, linux) MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Forrest J Cavalier writes: > This isn't acceptable wording at all, especially not for USEFOR. A news > client must be able to determine "followup" by looking at an article, or > the term is meaningless and should not appear in any standards document. Well, John's right about this. If this is a requirement, then we have to define everything that contains a References header as a followup, whether somewhat artificially or not, since there isn't any other way to detect a followup. I'm not sure where this requirement comes from, though. It seems to me that certain concepts that cannot be detected in software are still useful to talk about in a standard when they affect what software does in response to common user interface commands (like "post followup"). While I can't tell in software if a random article on the wire is a followup or is some other case where References is useful, I *can* tell if a particular news client does the right thing by looking at it, which to me makes it standards territory. -- Russ Allbery (rra@stanford.edu) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3QHgsiG020130 for ; Tue, 26 Apr 2005 10:42:54 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3QHgsu5020129 for ietf-usefor-skb; Tue, 26 Apr 2005 10:42:54 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from eagle.oceana.com (eagle.oceana.com [208.17.123.12]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3QHgsnM020107 for ; Tue, 26 Apr 2005 10:42:54 -0700 (PDT) (envelope-from ken@oceana.com) Received: from [192.168.10.26] (KEN.oceana.com [192.168.10.26]) by eagle.oceana.com (8.13.2/8.13.2) with ESMTP id j3QHgkJW012123 for ; Tue, 26 Apr 2005 13:42:46 -0400 (EDT) Message-ID: <426E7D66.40008@oceana.com> Date: Tue, 26 Apr 2005 13:41:58 -0400 From: Ken Murchison Organization: Oceana Matrix Ltd. User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.7) Gecko/20040514 X-Accept-Language: en,pdf MIME-Version: 1.0 To: ietf-usefor@imc.org Subject: Re: Differences between RFC 2822 and Usefor (fwd) References: In-Reply-To: Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit X-Spam-Score: 0 (BAYES_00) X-Scanned-By: MIMEDefang 2.39 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: John Stanley wrote: > > I said: > > >>The fact that I am the only one who even commented on it being gone is a >>pretty good sign that nobody cares that it is gone. > > > Ken Murchison responds: > > >>No, I knew that it wasn't there, You have misquoted me here. The above was written by Charles. > > What are you saying "no" to? You did not comment on it being gone, at > least not in any public forum I know of designed for this discussion. > Don't tell me I'm wrong when clearly I am not. The fact that you knew it > wasn't there and did not comment is even stronger evidence that it not > being there is ok by you. If you cared, and knew, you could have > commented. > > In fact, since your name is on the draft as an editor, you could have put > it back had you thought it necessary; the fact you had access to the > document and did not act to change it is an even stronger indication that > it is ok by you. And that it came OUT under your name is proof that it is > ok with you. > > >>and there are a lot of other things that >>should be in the Usefor draft and that have not made it there yet. Again a misquote. If you're going to quote text, please attribute it to the correct party. > Since you conveniently reworded the problem into one of "should and should > not" instead of the actual "MUST and MUST NOT", you can easily say you see > no contradiction and then pretend that this lack of contradiction applies > to the real issue at hand. Nice trick. Didn't work this time. > > You cannot say "this is optional" and "this is mandatory" at the same > time. Pick one, stick with it, and accept the implications from your > choice. I've covered the implications, and "mandatory option" is not one > of them. Thank you for your pleasant and constructive tone. Is your argument that References shouldn't be required for followups (regardless of what USEFOR/USEPRO currently says) or that the current text is incorrect/ambiguous/contradictory/shitty? If the former, please state so. If the latter, please suggest some text for one or more of the documents that you feel is technically sound. -- Kenneth Murchison Oceana Matrix Ltd. Software Engineer 21 Princeton Place 716-662-8973 x26 Orchard Park, NY 14127 --PGP Public Key-- http://www.oceana.com/~ken/ksm.pgp Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3QGnOgP012487 for ; Tue, 26 Apr 2005 09:49:24 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3QGnOJk012486 for ietf-usefor-skb; Tue, 26 Apr 2005 09:49:24 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from a.mail.peak.org (a.mail.peak.org [69.59.192.41]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3QGnN1J012463 for ; Tue, 26 Apr 2005 09:49:23 -0700 (PDT) (envelope-from stanley@peak.org) Received: from a.shell.peak.org ([69.59.192.81]) by a.mail.peak.org (8.12.10/8.12.8) with ESMTP id j3QGnFfO080426 (version=TLSv1/SSLv3 cipher=EDH-RSA-DES-CBC3-SHA bits=168 verify=NO) for ; Tue, 26 Apr 2005 09:49:17 -0700 (PDT) Date: Tue, 26 Apr 2005 09:49:15 -0700 (PDT) From: John Stanley X-X-Sender: stanley@a.shell.peak.org To: ietf-usefor@imc.org Subject: Re: Differences between RFC 2822 and Usefor (fwd) Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Spam-Score: 0 () X-Scanned-By: MIMEDefang 2.39 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: I said: >The fact that I am the only one who even commented on it being gone is a >pretty good sign that nobody cares that it is gone. Ken Murchison responds: >No, I knew that it wasn't there, What are you saying "no" to? You did not comment on it being gone, at least not in any public forum I know of designed for this discussion. Don't tell me I'm wrong when clearly I am not. The fact that you knew it wasn't there and did not comment is even stronger evidence that it not being there is ok by you. If you cared, and knew, you could have commented. In fact, since your name is on the draft as an editor, you could have put it back had you thought it necessary; the fact you had access to the document and did not act to change it is an even stronger indication that it is ok by you. And that it came OUT under your name is proof that it is ok with you. >and there are a lot of other things that >should be in the Usefor draft and that have not made it there yet. This group has been working on this stuff for YEARS. "Not made it there, yet"? If it hasn't "made it there, yet", it's way too late. How many "last calls" has this group been through, already? >I disagree. It is my opinion (and I believe our Chair's) that USEFOR only >document the syntax of an article. The use or non-use of any particular >"optional" header such as References belongs in either USEPRO or USEAGE. If a header is MANDATORY in an article, then it is part of the syntax of that article. >I think you're taking the fact that neither RFC2822 nor USEFOR mandate >References too literally. Excuse me? RFC2822 uses specific terms that say that the References header is NOT mandatory, and even gives an example of how to deal with a reply that does NOT contain a References header, and I'm being "too literal"? This is a technical standard. Either it says what it means or it is broken. >USEFOR only documents what is necessary to make an article syntactically >correct, not what makes it useful in the Netnews world. If a followup REQUIRES a References header, then a References header is REQUIRED to make that article "syntactically correct". Screw "useful", "useful" doesn't get you RFC2119 mandates. >In this regard, References is optional. Yep. According to RFC2822 and draft-usefor-03, References is optional. In any article, even replies, and since replies are followups in our terminology, optional even in followups. >A "followup" article which doesn't have a References header is still >syntactically correct. Not if you think it MUST contain a References header. If you think it MUST contain a References header, a "followup that doesn't have a References header" is broken. Unfortunately, RFC2822 tells us how to process replies (one kind of followup) that don't contain a References header, so trying to pretend there is a mandate is just silly. >USEPRO documents how to construct an article >("followup" or otherwise) which will play nice in the Netnews world. Can we drop this stupid anthropomorphism? Articles don't "play", they are either correct or they are not. Transport agents either transport them or drop them. Reading agents either display them or do not. Articles without References headers exist and they are transported and displayed by the news system every day. They are not mandatory, and our draft now makes that clear. I'm fine with that. I'm not fine with this pretense that we can somehow sneak a mandate back into the standard after explicitely removing it. >In this regard, USEPRO can and should mandate that References be present >in "followup" articles. Show me in RFC2119 where it uses the words "play nice" in regards to MUST/MUST NOT language. In fact, from what you've said about the References header, I see that RFC2119 says that we must not use RFC2119 imperatives, since there is no requirement for interoperability in anything you've said. "Play nice" and "useful" are not reasons for RFC2119 language. >USEPRO (and possibly USEAGE) will dictate when the optional headers >documented in USEFOR should and should not be used. >I don't see a contradiction here, but maybe I'm wrong. Since you conveniently reworded the problem into one of "should and should not" instead of the actual "MUST and MUST NOT", you can easily say you see no contradiction and then pretend that this lack of contradiction applies to the real issue at hand. Nice trick. Didn't work this time. You cannot say "this is optional" and "this is mandatory" at the same time. Pick one, stick with it, and accept the implications from your choice. I've covered the implications, and "mandatory option" is not one of them. Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3QGcB4t010478 for ; Tue, 26 Apr 2005 09:38:11 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3QGcB5N010477 for ietf-usefor-skb; Tue, 26 Apr 2005 09:38:11 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from soy.epix.net (soy.epix.net [199.224.64.64]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3QGcAaX010470 for ; Tue, 26 Apr 2005 09:38:10 -0700 (PDT) (envelope-from mibsoft@epix.net) Received: from [192.168.2.11] (hrbg-216-108-206-228-pppoe.dsl.hrbg.epix.net [216.108.206.228]) by soy.epix.net (8.12.10/2004120601/PL) with ESMTP id j3QGbrgE004420; Tue, 26 Apr 2005 12:38:04 -0400 (EDT) Message-ID: <426E6E66.2070109@epix.net> Date: Tue, 26 Apr 2005 12:37:58 -0400 From: "Forrest J. Cavalier III" User-Agent: Mozilla Thunderbird 0.7 (Windows/20040616) X-Accept-Language: en-us, en MIME-Version: 1.0 To: chl@clerew.man.ac.uk CC: ietf-usefor@imc.org Subject: Re: Suggested References texts References: In-Reply-To: Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit X-Scanned-By: MIMEDefang 2.51 on 199.224.89.153 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Charles Lindsey wrote: > I think we are now sufficiently agreed about the References header (except > perhaps for John) to write some text. Point of order. The chair must declare consensus. There are problems with what you wrote, and I think that is precisely because you misunderstand the consensus..... > > Here is what I suggest. Note that, in addition to the USEFOR text written > by Ken, I have added some preliminary wording to outline the purpose of > the References header, and some semantics text to explain the "meaning" > of the list of message-ids. If USEFOR is going to be syntax only, then there is no place for "purpose of the References" header that can be vanilla enough to be in USEFOR. So, the following is wrong: > Within USEFOR: > > 3.2.1 References > > | The References header is used in followups, and other articles with > | related precursors (e.g. multipart FAQs and the fragments of a > | message/partial), to facilitate the display or retrieval, by reading > | and other agents, of threads of related articles. It is the same as > | that specified in Section 3.6.4 of [RFC2822] with the added > | restrictions detailed in Section 2.2 and those listed below: This isn't acceptable wording at all, especially not for USEFOR. A news client must be able to determine "followup" by looking at an article, or the term is meaningless and should not appear in any standards document. Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3QGFA7t005548 for ; Tue, 26 Apr 2005 09:15:10 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3QGFA7d005547 for ietf-usefor-skb; Tue, 26 Apr 2005 09:15:10 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from lon-mail-4.gradwell.net (lon-mail-4.gradwell.net [193.111.201.130]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3QGF8rS005537 for ; Tue, 26 Apr 2005 09:15:09 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from host81-144-72-114.midband.mdip.bt.net ([81.144.72.114]) by lon-mail-4.gradwell.net with esmtp (Gradwell gwh-smtpd 1.182) id 426e690a.6ead.47 for ietf-usefor@imc.org; Tue, 26 Apr 2005 17:15:06 +0100 (envelope-sender ) Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3QGCGl00858 for ietf-usefor@imc.org; Tue, 26 Apr 2005 17:12:16 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20722 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Suggested References texts Message-ID: X-Newsreader: NN version 6.5.2 (NOV) Date: Tue, 26 Apr 2005 15:12:10 GMT Lines: 116 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: I think we are now sufficiently agreed about the References header (except perhaps for John) to write some text. Here is what I suggest. Note that, in addition to the USEFOR text written by Ken, I have added some preliminary wording to outline the purpose of the References header, and some semantics text to explain the "meaning" of the list of message-ids. I have also fixed the syntax error (the CFWS between the s is obligatory), and have continued the liberty of using rather than . There is also a tentative NOTE at the end regarding the use of the References header in message/partial. This is a place where what RFC 2046 suggests is woefully inadequate, and that needs to be pointed out somewhere. But maybe this is not the right place. Share and Enjoy! Definitions (for inclusion in USEFOR or USEPRO as appropriate). A "followup" is an article containing a response to the contents of an earlier article. It will always include a References header pointing to that earlier article and any other "precursors". An article is a "precursor" of some later article which is a followup to it, or which is otherwise intended to be grouped with it for purposes of display (e.g. as a part of a multipart posting such as a FAQ). A "followup agent" is a combination of reading agent and posting agent that aids in the preparation and posting of a followup. Within USEFOR: 3.2.1 References | The References header is used in followups, and other articles with | related precursors (e.g. multipart FAQs and the fragments of a | message/partial), to facilitate the display or retrieval, by reading | and other agents, of threads of related articles. It is the same as | that specified in Section 3.6.4 of [RFC2822] with the added | restrictions detailed in Section 2.2 and those listed below: o The updated construct defined in Section 3.1.3 MUST be used. o Message identifiers MUST be separated with CFWS. o Comments in CFWS between message identifiers can cause interoperability problems, so comments SHOULD NOT be generated, but MUST be accepted. | references = "References:" SP [CFWS] 1*( msg-id [CFWS] ) | CRLF | The list is composed of message identifiers of precursors of the | current article, sorted so that no article precedes any of its own | precursors. It SHOULD include both the earliest and the immediate | precursors on the current article, even if some of the intermediate | ones are omitted. A given message identifier MUST NOT appear more | than once. [That actually defines a partial list. Even though USEPRO makes no provsision for followups to several articles, people are bound to try to do it. That is a minimal wording which reduces to the usual linear list in the case of followups to a single precursor.] | | The process of generating a References header by a followup agent is | set out in [USEPRO]. | | NOTE: The suggestion in [RFC 2046] that each fragment of a | message/partial should contain a References header referring only | to the immediatelyt preceding fragment does not fulfil the | recommendation above. It would therefore be better for each | fragment to refer to all of the preceding ones. Within USEPRO: 7.6. Duties of a Followup Agent ....... 4. If the precursor did not have a References header (F-3.2.1), the content of the followup's References header MUST be inherited from that of the Message-ID header of the precursor. A followup to an article which already had a References header MUST have a References header comprising the precursor's References header (subject to trimming as described below) followed by CFWS and the Message-ID header content of the precursor. | NOTE: The two "MUST"s above ensure compliance with the | definition of the term "followup". They are to be contrasted | with the weaker recomendation using "SHOULD" applied, in [RFC | 2822], to the generation of "replies" in email. Moreover, in | Netnews, there is no expectation of any In-Reply-To header in a | followup. If the resulting References header is excessively long, it MAY be trimmed, but the first and the last two message identifiers MUST NOT be removed. -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3QFFodk094300 for ; Tue, 26 Apr 2005 08:15:50 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3QFFoE7094299 for ietf-usefor-skb; Tue, 26 Apr 2005 08:15:50 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from eagle.oceana.com (eagle.oceana.com [208.17.123.12]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3QFFnSQ094282 for ; Tue, 26 Apr 2005 08:15:49 -0700 (PDT) (envelope-from ken@oceana.com) Received: from [192.168.10.26] (KEN.oceana.com [192.168.10.26]) by eagle.oceana.com (8.13.2/8.13.2) with ESMTP id j3QFFgDe009529 for ; Tue, 26 Apr 2005 11:15:42 -0400 (EDT) Message-ID: <426E5AEF.3040407@oceana.com> Date: Tue, 26 Apr 2005 11:14:55 -0400 From: Ken Murchison Organization: Oceana Matrix Ltd. User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.7) Gecko/20040514 X-Accept-Language: en,pdf MIME-Version: 1.0 To: ietf-usefor@imc.org Subject: Re: References References: In-Reply-To: Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit X-Spam-Score: 0 (BAYES_00) X-Scanned-By: MIMEDefang 2.39 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: John Stanley wrote: > > "Charles Lindsey" : > > >>Our Chair has ruled that the correct place to say such things is in USEPRO >>rather than USEFOR. > > > Then the Chair has effectively ruled that References is optional as > defined in RFC2822 and that "followup" is a meaningless term, since nobody > past the poster can identify what is and is not a followup using the > optional definition of References. If the Chair thinks that we can say in > one news standard that References is an optional header (by referring to > the SHOULD appear and "may appear" definition in RFC2822) and then change > our minds in mid-stream by saying that it is MANDATORY in another news > standard, then he is demonstrating schizophrenia. Either it IS mandatory > in followups or it is not. It is mandatory as stated in USEPRO. I'll reiterate what I stated elsewhere in the thread. USEFOR documents the proper syntax of the set of headers typically used in news articles. USEPRO documents how to properly construct an article from this set per the Netnews "protocol" or "etiquette" -- Kenneth Murchison Oceana Matrix Ltd. Software Engineer 21 Princeton Place 716-662-8973 x26 Orchard Park, NY 14127 --PGP Public Key-- http://www.oceana.com/~ken/ksm.pgp Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3QF1gil091742 for ; Tue, 26 Apr 2005 08:01:42 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3QF1gmq091741 for ietf-usefor-skb; Tue, 26 Apr 2005 08:01:42 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from eagle.oceana.com (eagle.oceana.com [208.17.123.12]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3QF1fu2091697 for ; Tue, 26 Apr 2005 08:01:42 -0700 (PDT) (envelope-from ken@oceana.com) Received: from [192.168.10.26] (KEN.oceana.com [192.168.10.26]) by eagle.oceana.com (8.13.2/8.13.2) with ESMTP id j3QF1YoV009273 for ; Tue, 26 Apr 2005 11:01:34 -0400 (EDT) Message-ID: <426E579F.1060809@oceana.com> Date: Tue, 26 Apr 2005 11:00:47 -0400 From: Ken Murchison Organization: Oceana Matrix Ltd. User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.7) Gecko/20040514 X-Accept-Language: en,pdf MIME-Version: 1.0 To: ietf-usefor@imc.org Subject: Re: References References: In-Reply-To: Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit X-Spam-Score: 0 (BAYES_00) X-Scanned-By: MIMEDefang 2.39 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: John Stanley wrote: > > "Charles Lindsey" : > > >>Frank Ellermann writes: > > >>>No References => no follow-up. > > >>Which indeed follows from the fact that USEPRO says that a followup agent >>MUST incorporate a References header (and do it correctly, of course). > > > Unfortunately, RFC2822 says only SHOULD regarding the References header > and followups, and USEFOR says that the References header is defined as > found in RFC2822 (and for the pedants who cannot identify the context of > this discussion, I'll add "other than a few syntax changes that don't > change the requirements for usage, which is the context of this discussion > here.) So References in a followup is only a SHOULD, and if USEPRO says > otherwise, it contradicts both RFC2822 and USEFOR. While one of those we > can modify for good reason, the other is the result of our modifications, > and we did NOT modify the References header to make it mandatory in any > use. I think you're taking the fact that neither RFC2822 nor USEFOR mandate References too literally. USEFOR only documents what is necessary to make an article syntactically correct, not what makes it useful in the Netnews world. In this regard, References is optional. A "followup" article which doesn't have a References header is still syntactically correct. This doesn't mean that its recommended or useful. USEPRO documents how to construct an article ("followup" or otherwise) which will play nice in the Netnews world. In this regard, USEPRO can and should mandate that References be present in "followup" articles. USEPRO (and possibly USEAGE) will dictate when the optional headers documented in USEFOR should and should not be used. I don't see a contradiction here, but maybe I'm wrong. -- Kenneth Murchison Oceana Matrix Ltd. Software Engineer 21 Princeton Place 716-662-8973 x26 Orchard Park, NY 14127 --PGP Public Key-- http://www.oceana.com/~ken/ksm.pgp Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3QESG9c086177 for ; Tue, 26 Apr 2005 07:28:16 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3QESGqX086176 for ietf-usefor-skb; Tue, 26 Apr 2005 07:28:16 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from eagle.oceana.com (eagle.oceana.com [208.17.123.12]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3QESFDT086157 for ; Tue, 26 Apr 2005 07:28:16 -0700 (PDT) (envelope-from ken@oceana.com) Received: from [192.168.10.26] (KEN.oceana.com [192.168.10.26]) by eagle.oceana.com (8.13.2/8.13.2) with ESMTP id j3QES9ZJ008552 for ; Tue, 26 Apr 2005 10:28:09 -0400 (EDT) Message-ID: <426E4FC9.7020903@oceana.com> Date: Tue, 26 Apr 2005 10:27:21 -0400 From: Ken Murchison Organization: Oceana Matrix Ltd. User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.7) Gecko/20040514 X-Accept-Language: en,pdf MIME-Version: 1.0 To: ietf-usefor@imc.org Subject: Re: References References: <4266EF35.21A5@xyzzy.claranet.de> <42686E91.42E2@xyzzy.claranet.de> In-Reply-To: Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit X-Spam-Score: 0 (BAYES_00) X-Scanned-By: MIMEDefang 2.39 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Charles Lindsey wrote: > In <42686E91.42E2@xyzzy.claranet.de> Frank Ellermann writes: > >> A "followup" is an article containing a response to the >> contents of an earlier article (the followup's "precursor" >> indicated in the "References" header field). > > > Yes, I had already concluded that something like that would be helpful in > the definition. Russ seems to agree. Done. -- Kenneth Murchison Oceana Matrix Ltd. Software Engineer 21 Princeton Place 716-662-8973 x26 Orchard Park, NY 14127 --PGP Public Key-- http://www.oceana.com/~ken/ksm.pgp Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3QEPYje085824 for ; Tue, 26 Apr 2005 07:25:34 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3QEPYVo085823 for ietf-usefor-skb; Tue, 26 Apr 2005 07:25:34 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from eagle.oceana.com (eagle.oceana.com [208.17.123.12]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3QEPXJo085813 for ; Tue, 26 Apr 2005 07:25:34 -0700 (PDT) (envelope-from ken@oceana.com) Received: from [192.168.10.26] (KEN.oceana.com [192.168.10.26]) by eagle.oceana.com (8.13.2/8.13.2) with ESMTP id j3QEPRw1008512 for ; Tue, 26 Apr 2005 10:25:27 -0400 (EDT) Message-ID: <426E4F27.2020607@oceana.com> Date: Tue, 26 Apr 2005 10:24:39 -0400 From: Ken Murchison Organization: Oceana Matrix Ltd. User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.7) Gecko/20040514 X-Accept-Language: en,pdf MIME-Version: 1.0 To: ietf-usefor@imc.org Subject: Re: References References: <4266EF35.21A5@xyzzy.claranet.de> <87mzrtt1fh.fsf@windlord.stanford.edu> In-Reply-To: Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit X-Spam-Score: 0 (BAYES_00) X-Scanned-By: MIMEDefang 2.39 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Charles Lindsey wrote: > In <87mzrtt1fh.fsf@windlord.stanford.edu> Russ Allbery writes: > > >>Sounds right to me. > > >>If it doesn't have References, it's not a follow-up, but in practice >>people use References for things other than follow-ups, and I don't see >>any reason to ban those uses. So the question is whether we want to >>artificially define all those things as follow-ups as well, or not worry >>about it and just say that References is required for follow-ups but may >>also be used for other things. > > > Exactly. Those other uses for the References header deserve a brief > mention somewhere, probably in USEFOR, but it would indeed be somewhat > artificial to define those cases as "followups", and I personally do not > want to do that. I disagree. It is my opinion (and I believe our Chair's) that USEFOR only document the syntax of an article. The use or non-use of any particular "optional" header such as References belongs in either USEPRO or USEAGE. -- Kenneth Murchison Oceana Matrix Ltd. Software Engineer 21 Princeton Place 716-662-8973 x26 Orchard Park, NY 14127 --PGP Public Key-- http://www.oceana.com/~ken/ksm.pgp Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3QEHG3v084678 for ; Tue, 26 Apr 2005 07:17:16 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3QEHGlD084677 for ietf-usefor-skb; Tue, 26 Apr 2005 07:17:16 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from eagle.oceana.com (eagle.oceana.com [208.17.123.12]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3QEHFgk084662 for ; Tue, 26 Apr 2005 07:17:15 -0700 (PDT) (envelope-from ken@oceana.com) Received: from [192.168.10.26] (KEN.oceana.com [192.168.10.26]) by eagle.oceana.com (8.13.2/8.13.2) with ESMTP id j3QEH8Le008349 for ; Tue, 26 Apr 2005 10:17:08 -0400 (EDT) Message-ID: <426E4D35.6080409@oceana.com> Date: Tue, 26 Apr 2005 10:16:21 -0400 From: Ken Murchison Organization: Oceana Matrix Ltd. User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.7) Gecko/20040514 X-Accept-Language: en,pdf MIME-Version: 1.0 To: ietf-usefor@imc.org Subject: Re: Differences between RFC 2822 and Usefor (fwd) References: In-Reply-To: Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit X-Spam-Score: 0 (BAYES_00) X-Scanned-By: MIMEDefang 2.39 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Charles Lindsey wrote: > In John Stanley writes: > > >>"Charles Lindsey" : > > >>>The "MUST" in question arises from the long held consensus in this Working >>>Group that the References header in News is NOT an "optional extra". > > >>Not according to draft-usefor-03. According to draft-usefor-03, >>References is defined to be the same as in RFC2822, with the addition of a >>few bits that are not mandates for using it. This consensus you claim >>exists has a very odd way of being expressed in our products. Almost as if >>it doesn't really exist. > > >>The fact that I am the only one who even commented on it being gone is a >>pretty good sign that nobody cares that it is gone. > > > No, I knew that it wasn't there, and there are a lot of other things that > should be in the Usefor draft and that have not made it there yet. But I > do not want to be seen arguing with my co-editor in public if it can be > avoided. Especially on issues that do not appear to be the subject of > controversy. > > > >>>That clear requirement has been in all our drafts up to article-13. > > >>It is no longer a requirement at all in our draft. > > > Indeed, but a draft is only a draft. There has been no WG decision to > change the requirement for a References header in followups. Our Chair has > taken the view that Usefor is not the place to say it, but that Usepro IS > the proper place (and indeed there is already a corresponding MUST in > Usepro, though it may need a little more amplification). I do not agree > with the Chair on this one, but if nobody else speaks up and if his > ruling stands, then Usepro is where it will be said (which also, > incidentally, removes most of the cause of the disagreement there has been > about how to word it all). I am in agreement with the Chair that this discussion belongs in USEPRO (I thought we had discussed this offline, but perhaps not). -- Kenneth Murchison Oceana Matrix Ltd. Software Engineer 21 Princeton Place 716-662-8973 x26 Orchard Park, NY 14127 --PGP Public Key-- http://www.oceana.com/~ken/ksm.pgp Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3QECeWq083990 for ; Tue, 26 Apr 2005 07:12:40 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3QECexc083989 for ietf-usefor-skb; Tue, 26 Apr 2005 07:12:40 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from eagle.oceana.com (eagle.oceana.com [208.17.123.12]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3QECdim083972 for ; Tue, 26 Apr 2005 07:12:40 -0700 (PDT) (envelope-from ken@oceana.com) Received: from [192.168.10.26] (KEN.oceana.com [192.168.10.26]) by eagle.oceana.com (8.13.2/8.13.2) with ESMTP id j3QECXs1008268 for ; Tue, 26 Apr 2005 10:12:33 -0400 (EDT) Message-ID: <426E4C22.1040808@oceana.com> Date: Tue, 26 Apr 2005 10:11:46 -0400 From: Ken Murchison Organization: Oceana Matrix Ltd. User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.7) Gecko/20040514 X-Accept-Language: en,pdf MIME-Version: 1.0 To: ietf-usefor@imc.org Subject: Re: Differences between RFC 2822 and Usefor References: In-Reply-To: Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit X-Spam-Score: 0 (BAYES_00) X-Scanned-By: MIMEDefang 2.39 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: John Stanley wrote: > > "Charles Lindsey" : > > >>And why does the Usefor draft contain the words > > >> None of the headers appearing in this section is required to appear >> in every article but some of them are required in certain types of >> article, such as followups. > > > Because someone (an editor, perhaps?) overlooked this statement when > removing the mandate for References? Remove the "such as" clause and the > problem is fixed. Do it and lets move on. Problem solved. Done. -- Kenneth Murchison Oceana Matrix Ltd. Software Engineer 21 Princeton Place 716-662-8973 x26 Orchard Park, NY 14127 --PGP Public Key-- http://www.oceana.com/~ken/ksm.pgp Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3QE90HC083430 for ; Tue, 26 Apr 2005 07:09:00 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3QE905H083429 for ietf-usefor-skb; Tue, 26 Apr 2005 07:09:00 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from eagle.oceana.com (eagle.oceana.com [208.17.123.12]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3QE8xII083408 for ; Tue, 26 Apr 2005 07:09:00 -0700 (PDT) (envelope-from ken@oceana.com) Received: from [192.168.10.26] (KEN.oceana.com [192.168.10.26]) by eagle.oceana.com (8.13.2/8.13.2) with ESMTP id j3QE8r4j008207; Tue, 26 Apr 2005 10:08:53 -0400 (EDT) Message-ID: <426E4B45.7010205@oceana.com> Date: Tue, 26 Apr 2005 10:08:05 -0400 From: Ken Murchison Organization: Oceana Matrix Ltd. User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.7) Gecko/20040514 X-Accept-Language: en,pdf MIME-Version: 1.0 To: Alexey Melnikov CC: ietf-usefor@imc.org Subject: Re: Obsolete vs. obsolescent References: <200504111212.j3BCChM14516@clerew.man.ac.uk> <87psx1gnjc.fsf@windlord.stanford.edu> <87fyxvx1m6.fsf@windlord.stanford.edu> <87mzs2a1hu.fsf@windlord.stanford.edu> <87wtqxui6s.fsf@windlord.stanford.edu> <87d5so2bpk.fsf@windlord.stanford.edu> <42697BB6.6074@xyzzy.claranet.de> <426BC455.2040108@isode.com> In-Reply-To: <426BC455.2040108@isode.com> Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit X-Spam-Score: 0 (BAYES_00) X-Scanned-By: MIMEDefang 2.39 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Alexey Melnikov wrote: >> > In order to close this issue I suggest that the USEFOR document can > include few sentences describing what do we mean when calling a > particular header "obsolete". Care to craft some text and tell me where you'd like to see it? -- Kenneth Murchison Oceana Matrix Ltd. Software Engineer 21 Princeton Place 716-662-8973 x26 Orchard Park, NY 14127 --PGP Public Key-- http://www.oceana.com/~ken/ksm.pgp Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3QE3EH2082286 for ; Tue, 26 Apr 2005 07:03:14 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3QE3Ecc082285 for ietf-usefor-skb; Tue, 26 Apr 2005 07:03:14 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from eagle.oceana.com (eagle.oceana.com [208.17.123.12]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3QE3DCm082276 for ; Tue, 26 Apr 2005 07:03:13 -0700 (PDT) (envelope-from ken@oceana.com) Received: from [192.168.10.26] (KEN.oceana.com [192.168.10.26]) by eagle.oceana.com (8.13.2/8.13.2) with ESMTP id j3QE36Gc008111 for ; Tue, 26 Apr 2005 10:03:06 -0400 (EDT) Message-ID: <426E49EB.3010709@oceana.com> Date: Tue, 26 Apr 2005 10:02:19 -0400 From: Ken Murchison Organization: Oceana Matrix Ltd. User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.7) Gecko/20040514 X-Accept-Language: en,pdf MIME-Version: 1.0 To: ietf-usefor@imc.org Subject: Re: Fwd: http://www.ietf.org/internet-drafts/draft-ietf-usefor-usefor-03.txt References: <200504111212.j3BCChM14516@clerew.man.ac.uk> In-Reply-To: <200504111212.j3BCChM14516@clerew.man.ac.uk> Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit X-Spam-Score: 0 (BAYES_00) X-Scanned-By: MIMEDefang 2.39 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Charles Lindsey wrote: > In local.usefor you write: > > >>------- Forwarded message ------- >>From: "Dr John Stockton" >>To: "Charles. H. Lindsey" >>Subject: >>http://www.ietf.org/internet-drafts/draft-ietf-usefor-usefor-03.txt >>Date: Sun, 10 Apr 2005 19:44:55 +0100 > > > >>"1.1 Basic Concepts >> "Netnews" is a set of protocols for generating, storing and >> retrieving news "articles" (which are a subset of Email messages)" > > >>They are surely not a subset of E-mail messages. They might be >>described as similar. But E-mail is now commonly HTML & News should not >>be; I'd omit the parenthetic analogy. > > > Yes, it might be better to say "(whose format is a subset of that for > Email messages)". Ken? Yes, I'll make this change. -- Kenneth Murchison Oceana Matrix Ltd. Software Engineer 21 Princeton Place 716-662-8973 x26 Orchard Park, NY 14127 --PGP Public Key-- http://www.oceana.com/~ken/ksm.pgp Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3PJFA0w065996 for ; Mon, 25 Apr 2005 12:15:10 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3PJFA05065995 for ietf-usefor-skb; Mon, 25 Apr 2005 12:15:10 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from a.mail.peak.org (a.mail.peak.org [69.59.192.41]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3PJF9u2065955 for ; Mon, 25 Apr 2005 12:15:09 -0700 (PDT) (envelope-from stanley@peak.org) Received: from a.shell.peak.org ([69.59.192.81]) by a.mail.peak.org (8.12.10/8.12.8) with ESMTP id j3PJF0fO064514 (version=TLSv1/SSLv3 cipher=EDH-RSA-DES-CBC3-SHA bits=168 verify=NO) for ; Mon, 25 Apr 2005 12:15:02 -0700 (PDT) Date: Mon, 25 Apr 2005 12:15:00 -0700 (PDT) From: John Stanley X-X-Sender: stanley@a.shell.peak.org To: ietf-usefor@imc.org Subject: Re: References Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Spam-Score: 0 () X-Scanned-By: MIMEDefang 2.39 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: "Charles Lindsey" : >Please show me where the word "followup" is to be found in RFC 2822. I won't bother doing so, because we've already covered RFC2822 using the term "reply" in the relevant section, and that all replies are followups, using any common-sense or even news-based definition of the word followup. Not all followups are necessarily replies, but every "reply" is "an article posted in response to another". Since RFC2822 clearly does NOT mandate References headers in all replies, then it just as clearly exempts some followups -- the same replies that have no mandate for its use are the same followups that have no mandate. You cannot say that all followups MUST have a References header without also saying all replies MUST, since the latter is a proper subset of the former. > USEFOR incorparates, by reference, many concepts from RFC 2822, Including, explicitely, the References header. Thus RFC2822 not only tells us how to construct the "proper Email [sic]" that you grant in your previous paragraph, but a "proper News" article, because USEFOR says that RFC2822 applies to news, too. >But nowhere does USEFOR >equate the concepts of "Email Reply" with "Netnews Followup". It doesn't need to. It is quite disengenuous of you to pretend that RFC2822 says "Email Reply" in the section about References headers, and to pretend that there has to be some equating of the two phrases for RFC2822 to apply. I've quoted the specific parts of RFC2822 that talk about "replies", and they do not limit themselves to email, and USEFOR-03 explicitely expands any limit of "Email Reply" that you may fictionalize for the References header and RFC2822 to "News reply", as well. And you know very well that the two concepts do not have to be equivalent for one to apply to the other. All replies are followups using standard common-sense and common-practice news definitions for the same. If you say, as RFC2822 does, and draft-usefor-03 brings into the news context, that not all replies require a References header, then you just as clearly say not all followups require a References header. This is now the third time in two days I've covered this ground for you, Charles. >I grant you that the concepts are sufficiently similar that, wherever we >say that News Followups are different in some way from Email Replies (e.g. >MUST rather than SHOULD), Except we DO NOT say this. We EXPLICITELY say that the RFC2822 definition of References applies (with a few small syntax differences) but we do NOT say that one of the differences is that References is a MUST. You can say that you grant that the concepts are similar, even though you have just argued that they are completely different and thus statements about one thing do not apply to the other, but you cannot grant that we say that there is a mandate, because draft-usefor-03 and RFC2822 clearly says otherwise. >Our Chair has ruled that the correct place to say such things is in USEPRO >rather than USEFOR. Then the Chair has effectively ruled that References is optional as defined in RFC2822 and that "followup" is a meaningless term, since nobody past the poster can identify what is and is not a followup using the optional definition of References. If the Chair thinks that we can say in one news standard that References is an optional header (by referring to the SHOULD appear and "may appear" definition in RFC2822) and then change our minds in mid-stream by saying that it is MANDATORY in another news standard, then he is demonstrating schizophrenia. Either it IS mandatory in followups or it is not. I'll be happy with whichever position we take: either RFC1036 had it right and References is MUST/MUST NOT and followups are thus identifiable; or RFC1036 has it wrong, References is SHOULD/MAY and followups are not identifiable and thus a meaningless term. But since it has already been changed, it now requires justification for bringing back the MUST/anything language, since RFC2119 clearly limits such language to serious matters, and we can no longer use the 'it was pre-existing' excuse. It's not existing anymore; can you justify bringing it back? I'll even suggest that if you cannot justify the return of a mandate (and do not just unilaterally reinsert it as is your wont to do with our work products), we can create a specific header to replace References, call it "Thread-Me-Next-To:" or "Display-Near:" or something similar that makes it clear this is a display issue and not a protocol one, and make that optional. That's fine. It's being honest then about the value of the information and not pretending it is some critical protocol interoperability thing we're trying to solve. And before you start playing the "show me where it says X" game with RFC1036 and MUST/MUST NOT, I'll point out that RFC1036 predates RFC2119 but it uses clear terms like "required" and "prohibited", which map into RFC2119 MUST/MUST NOT. But we've been through that before, too, so I don't expect this preemptive note to have any effect on your argument. >If a poster intends a followup, then a References header is REQUIRED. No. Refer to RFC2822, which says that References is only SHOULD and "may", and even talks about messages that are clearly replies (contains In-Reply-To) but has no References. >> That is a change, despite your repeated denials. >Exactly. That has never been denied. You've repeatedly denied it. You keep saying we can use your proposed wording because it makes no difference from what was there originally, regarding MUST/MUST NOT and followups. There's a difference from how things are now. There's a change. >You cannot tell from looking at the >References header whether the poster intended to write a followup, or >merely intended that it should be threaded like a followup (with all the >other benefits accruing from a References header). Let's put back some context to this, shall we? If you put back the MUST/MUST NOT dichotomy for the References header, which has been removed from draft-usefor-03 now (but still exists in news today due to RFC1036), then you can tell precisely if the poster intended a followup or not because a followup MUST have the header and a non-followup MUST NOT. (The third option is the poster is stupid/incompetent/ignorant and cannot properly form a news article, but let's stick with properly formattted news articles and not try to guess at the meaning of broken ones, ok?) It is only when you water down the References header as it has now been that you cannot tell what was intended. An article with a References header (under draft-usefor-03) might be a followup or it might not. An article without a References header might not be a followup, or it might be. You have no way of knowing. You can only guess. Technical standards are broken if they involve guessing about things that are supposed to be important, especially when they could easily be written to remove the guessing. >So f*****g what? Why should the reader care? If it is not important to the reader, then the References header absolutely does not merit any RFC2119 language regarding its use, much less MUST. It is not an interoperability issue in any case, and if the information that it conveys is "so fucking what?" for the reader, then lets make it obsolete and be done with it. I'm fine with that. References is obsolete. The term "followup" is removed from the drafts. Let's git-r-done. When can we expect you to produce the next version with these changes we've agreed to? Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3PJ03Ir063414 for ; Mon, 25 Apr 2005 12:00:03 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3PJ03PI063413 for ietf-usefor-skb; Mon, 25 Apr 2005 12:00:03 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from lon-mail-1.gradwell.net (lon-mail-1.gradwell.net [193.111.201.125]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3PJ01hR063403 for ; Mon, 25 Apr 2005 12:00:02 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from host81-144-77-174.midband.mdip.bt.net ([81.144.77.174]) by lon-mail-1.gradwell.net with esmtp (Gradwell gwh-smtpd 1.181) id 426d3e2f.2b99.6b for ietf-usefor@imc.org; Mon, 25 Apr 2005 19:59:59 +0100 (envelope-sender ) Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3PIv6522990 for ietf-usefor@imc.org; Mon, 25 Apr 2005 19:57:06 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20720 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: Fixed (was: Broken Message-ID syntax) Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: <4255153E.3080500@isode.com> <42594879.3B92@xyzzy.claranet.de> <425A8A63.268A@xyzzy.claranet.de> <425C607E.1B5A@xyzzy.claranet.de> <425D8031.D57@xyzzy.claranet.de> <426BB63D.3060506@isode.com> Date: Mon, 25 Apr 2005 16:17:30 GMT Lines: 131 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In <426BB63D.3060506@isode.com> Alexey Melnikov writes: >Charles Lindsey wrote: >>My preference is to retain the RFC 2822 naming, which accords with the >>general principle that we do not change anything in RFC 2822 that we do >>not need to change. >> >I concur with your preference. In that case, we are ready for some syntax. I have given below what I now think should be in the Message-ID section of USEFOR, including the necessary syntax changes and rewordings of the descriptive text. I have also taken the liberty of removing in favour of , which I think we agreed in some earlier discussions on this list (there will be a few consequential changes elsewhere in USEFOR). Also, I have suggested a final paragraph regarding use of domains in the , which closes a possible shortcoming in the wording of RFC 2822, as spotted by Frank. >> >If WG members believe that better names should be used, I suggest they >talk to Pete Resnick first. As far as I remember he has already agreed >to publish 2822bis. I don't think Pete has gone beyond agreeing in principle to 2822bis. Has he actually published any Internet Drafts? If so, I will be needing to comment upon them. Here follows my suggested Message-ID text: 3.1.3 Message-ID The Message-ID header contains a single unique message identifier. This document updates the construct from Section 3.6.4 of [RFC2822] so as to ensure that Internet Message Format Message-IDs are usable in widely deployed news software. The global uniqueness requirement for in [RFC2822] is to be understood as applying across all protocols using such message identifiers, and across both Email and Netnews in particular. A revised syntax for is given below, but the requirements and descriptive text from Section 3.6.4 of [RFC2822] still apply. | message-id = "Message-ID:" SP [FWS] msg-id [FWS] CRLF | | msg-id = "<" id-left "@" id-right ">" | ; maximum length is 250 octets id-left = dot-atom-text / no-fold-quote id-right = dot-atom-text / no-fold-literal | no-fold-quote = DQUOTE | ( "." *mqtext / | *mqtext "." / | *mqtext mqspecial *mqtext ) | DQUOTE | | mqtext = atext / "." / mqspecial | mqspecial = "(" / ")" / ; same as specials except "<" / ; "\" and DQUOTE quoted | "[" / "]" / ; "." doubled and ">" omitted ":" / ";" / "@" / "," / | ".." / "\\" / "\" DQUOTE no-fold-literal = "[" *( mdtext / "\[" / "\]" / "\\" ) "]" | mdtext = %d33-61 / ; The rest of the US-ASCII %d63-90 / ; characters not including %d94-126 ; ">", "[", "]", or "\" The msg-id-core MUST NOT be more than 250 octets in length. NOTE: The length restriction ensures that systems which accept message identifiers as a parameter when retrieving an article (e.g. [NNTP]) can rely on a bounded length. | Observe that includes the < and >. Observe also that in contrast to the corresponding header in | [RFC2822] | o the syntax does not allow comments within the Message-ID | header, | o it ensures that no string of characters is quoted if it was already | a (it must start or end with a ".", or contain at | least one mqspecial), | o it ensures that no single character is prefixed by a "\" in the | form of a unless strictly necessary, | o it excludes all control characters, and | o there is no possibility for ">" or WSP to occur inside a | , whether quoted or not. | | This is to simplify processing by relaying and serving agents, and to | ensure interoperability with existing implementations and compliance | with [NNTP]. Thus, whereas under [RFC2822] the following s | would be considered semantically equivalent, | | <"ab.cd"@example.com> | <"ab.\cd"@example.com> only the first of them is syntactically permitted by this standard, and hence a simple comparison of octets will always suffice to determine the identity of two s. | Also note that this updated ABNF applies wherever is used, | including the References header discussed in Section 3.2.1 and the | Supersedes header discussed in Section 3.2.5. | NOTE: It in RECOMMENDED in [RFC 2822] that, for ensuring global | uniqueness, the be some domain identifier within whose | scope the uniqueness of the can be guaranteed. When | following this recommendation, any or | used for the are to be interpreted as | s as described in section 3.4.1 of [RFC 2822]. -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3PGDsac038040 for ; Mon, 25 Apr 2005 09:13:54 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3PGDsKA038039 for ietf-usefor-skb; Mon, 25 Apr 2005 09:13:54 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from lon-mail-2.gradwell.net (lon-mail-2.gradwell.net [193.111.201.126]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3PGDrMO038024 for ; Mon, 25 Apr 2005 09:13:54 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from host81-144-65-120.midband.mdip.bt.net ([81.144.65.120]) by lon-mail-2.gradwell.net with esmtp (Gradwell gwh-smtpd 1.181) id 426d1740.e3c.3a for ietf-usefor@imc.org; Mon, 25 Apr 2005 17:13:52 +0100 (envelope-sender ) Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3PGCFg21653 for ietf-usefor@imc.org; Mon, 25 Apr 2005 17:12:15 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20719 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: References Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: Date: Mon, 25 Apr 2005 14:16:34 GMT Lines: 69 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In John Stanley writes: >"Charles Lindsey" : >Unfortunately, RFC2822 says only SHOULD regarding the References header >and followups, Please show me where the word "followup" is to be found in RFC 2822. If you use your newsreader to send an Email reply to the author of a News article, then indeed you (only) SHOULD include a References header (and In-Reply-To as well to do the job properly). If you use your newsreader to post a News followup to a News article (usually in the same newsgroups as the precursor) then RFC 2822 is silent on the issue (because the whole purpose of RFC 2822 is to tell you how to construct a proper Email. USEFOR incorparates, by reference, many concepts from RFC 2822, including many headers and their meanings (semantics). But nowhere does USEFOR equate the concepts of "Email Reply" with "Netnews Followup". Therefore they remain different concepts with different rules. I grant you that the concepts are sufficiently similar that, wherever we say that News Followups are different in some way from Email Replies (e.g. MUST rather than SHOULD), then we need to draw attention to this. Our Chair has ruled that the correct place to say such things is in USEPRO rather than USEFOR. >You are apparently incapable of differentiating between "poster intends a >followup in what he sends" and "reader detects a followup in what he >gets." The previous standards and drafts have said that the intent for >something to be a followup by a poster REQUIRES a References header to be >inserted (MUST), and otherwise that header is prohibited. That means that >the READER can then determine what the poster intends to be a followup by >looking for a References header. If a poster intends a followup, then a References header is REQUIRED. If the poster does not intend a follouwp, but nevertheless includes a References header (e.g. in a multi-part FAQ, or to stitch together a message/partial), then what is the effect? >And yes, Charles, when you say "non-followups MAY contain a References >header", you have removed that ability to DETECT at the reader's end of >the pipe what is and is not intended to be a followup. That is a change, >despite your repeated denials. Exactly. That has never been denied. You cannot tell from looking at the References header whether the poster intended to write a followup, or merely intended that it should be threaded like a followup (with all the other benefits accruing from a References header). So f*****g what? Why should the reader care? So long as the reader (or his agent) does the Right Thing (as intended by the poster in both cases), why should anybody get upset? -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3PGDsAZ038032 for ; Mon, 25 Apr 2005 09:13:54 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3PGDs29038031 for ietf-usefor-skb; Mon, 25 Apr 2005 09:13:54 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from lon-mail-2.gradwell.net (lon-mail-2.gradwell.net [193.111.201.126]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3PGDrdc038023 for ; Mon, 25 Apr 2005 09:13:53 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from host81-144-65-120.midband.mdip.bt.net ([81.144.65.120]) by lon-mail-2.gradwell.net with esmtp (Gradwell gwh-smtpd 1.181) id 426d173f.e3c.39 for ietf-usefor@imc.org; Mon, 25 Apr 2005 17:13:51 +0100 (envelope-sender ) Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3PGCFi21649 for ietf-usefor@imc.org; Mon, 25 Apr 2005 17:12:15 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20718 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: References Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: <4266EF35.21A5@xyzzy.claranet.de> <42686E91.42E2@xyzzy.claranet.de> Date: Mon, 25 Apr 2005 13:47:06 GMT Lines: 48 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In <42686E91.42E2@xyzzy.claranet.de> Frank Ellermann writes: >Charles Lindsey wrote: >> How do we define the term "followup". >| A "followup" is an article containing a response to the >| contents of an earlier article (the followup's "precursor"). >Found in usefor-03. Or as Alternative-1 in USEPRO. >> or is it anything that has a References header. Which is Alternative-2 in USEPRO. I prefer Alternative-1, and I think most people are now of that view (modulo minor wording tweaks). >Not sure. If it's the RfC 2822 idea of a reply coming from a >mail2news gateway it probably is some "followup", but behind a >gateway you never know. Maybe it only had an In-Reply-To, the >gateway tried to fix it, but it was a broken pipermail archive >In-ReplyTo (the thread is then correct, but the "followup" is >shown at the wrong place in the thread). And lots of other ways it might have arisen. As Russ has said, referring to some of those cases as followups would be "artificial". > A "followup" is an article containing a response to the > contents of an earlier article (the followup's "precursor" > indicated in the "References" header field). Yes, I had already concluded that something like that would be helpful in the definition. Russ seems to agree. -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3OGoDWQ000986 for ; Sun, 24 Apr 2005 09:50:13 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3OGoDp2000985 for ietf-usefor-skb; Sun, 24 Apr 2005 09:50:13 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from rufus.isode.com (rufus.isode.com [62.3.217.251]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3OGnjcw000937 for ; Sun, 24 Apr 2005 09:50:12 -0700 (PDT) (envelope-from alexey.melnikov-usefor@isode.com) Received: from [192.168.0.3] ([62.3.217.253]) by rufus.isode.com via TCP (internal) with ESMTPA; Sun, 24 Apr 2005 17:49:29 +0100 Message-ID: <426BC455.2040108@isode.com> Date: Sun, 24 Apr 2005 17:07:49 +0100 From: Alexey Melnikov User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.7.2) Gecko/20040804 Netscape/7.2 (ax) X-Accept-Language: en-us, en MIME-Version: 1.0 To: Frank Ellermann CC: ietf-usefor@imc.org Subject: Re: Obsolete vs. obsolescent References: <200504111212.j3BCChM14516@clerew.man.ac.uk> <87psx1gnjc.fsf@windlord.stanford.edu> <87fyxvx1m6.fsf@windlord.stanford.edu> <87mzs2a1hu.fsf@windlord.stanford.edu> <87wtqxui6s.fsf@windlord.stanford.edu> <87d5so2bpk.fsf@windlord.stanford.edu> <42697BB6.6074@xyzzy.claranet.de> In-Reply-To: <42697BB6.6074@xyzzy.claranet.de> Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Frank Ellermann wrote: >Charles Lindsey wrote: > > >>"Obsolete" means this is no longer an official Netnews >>header. It would be labelled as "obsolete" in the IANA >>Registy of headers. >> >> > >Bruce has created a draft with mail headers found in obsolete >RfCs. Among others Fcc: found in RfC 724 obsoleted by RfC 733. > >For news we'd declare See-Also: and Also-Control: as obsolete, >as soon as Henry publishes s-o-1036 for information (obsoleted >by - I'm an optimist here - draft-ietf-usefor-usefor-04.tx) > > In order to close this issue I suggest that the USEFOR document can include few sentences describing what do we mean when calling a particular header "obsolete". >>"Obsolescent" means that is is still an official Netnews >>header (just), but that its continued use is discouraged >>(for reasons that are given), and that it is likely to be >>declared fully obsolete in a future standard. >> >> > >Now as pessimist: There won't be any future standard, UseFor >is the last news standard before its death. So Lines: will be >obsolescent "forever" > I tend to share your pessimism, but maybe for a slightly different reason. > until the last news server goes jabber, >or atom over soap based on beep, or whatever comes next. Bye. > > Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3OGoC0A000969 for ; Sun, 24 Apr 2005 09:50:12 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3OGoC4J000968 for ietf-usefor-skb; Sun, 24 Apr 2005 09:50:12 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from rufus.isode.com (rufus.isode.com [62.3.217.251]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3OGnjcu000937 for ; Sun, 24 Apr 2005 09:50:11 -0700 (PDT) (envelope-from alexey.melnikov-usefor@isode.com) Received: from [192.168.0.3] ([62.3.217.253]) by rufus.isode.com via TCP (internal) with ESMTPA; Sun, 24 Apr 2005 17:49:24 +0100 Message-ID: <426BB63D.3060506@isode.com> Date: Sun, 24 Apr 2005 16:07:41 +0100 From: Alexey Melnikov User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.7.2) Gecko/20040804 Netscape/7.2 (ax) X-Accept-Language: en-us, en MIME-Version: 1.0 To: Charles Lindsey CC: ietf-usefor@imc.org, Pete Resnick Subject: Re: Fixed (was: Broken Message-ID syntax) References: <4255153E.3080500@isode.com> <42594879.3B92@xyzzy.claranet.de> <425A8A63.268A@xyzzy.claranet.de> <425C607E.1B5A@xyzzy.claranet.de> <425D8031.D57@xyzzy.claranet.de> In-Reply-To: Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Charles Lindsey wrote: >In <425D8031.D57@xyzzy.claranet.de> Frank Ellermann writes: > > >>Charles Lindsey wrote: >> >> >>>Yes, you have convinced me now. >>> >>> >>Great, we have a msg-id syntax. >> >> >>>there remains the issue of the naming of the syntax rules, >>> >>> >>We're talking about seven names. So far we agree on msg-id, >>and we disagree on names for LHS, RHS, and address-literal. >> >> >>Maybe we can solve it for the remaining three names without >>bothering Alexey, Henry, or Ken: >> >> > >Essentially, we either use the same names as RFC 2822 uses, saying that >"the following syntax rules replace the corresponding rules in RFC 2822" >(giving us id-left, id-right, no-fold-quote, no-fold literal, etc), or >else we depart from the RFC 2822 names entirely, except for the msg-id at >the top, in which case your names are as good as any. I don't see that >there is anything in-between. > >My preference is to retain the RFC 2822 naming, which accords with the >general principle that we do not change anything in RFC 2822 that we do >not need to change. > I concur with your preference. >If our Chair says that we can change these names, then >let it be so. Alexey? > > If WG members believe that better names should be used, I suggest they talk to Pete Resnick first. As far as I remember he has already agreed to publish 2822bis. Alexey Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3OGoC3Q000977 for ; Sun, 24 Apr 2005 09:50:12 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3OGoCSw000976 for ietf-usefor-skb; Sun, 24 Apr 2005 09:50:12 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from rufus.isode.com (rufus.isode.com [62.3.217.251]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3OGnjcv000937 for ; Sun, 24 Apr 2005 09:50:12 -0700 (PDT) (envelope-from alexey.melnikov-usefor@isode.com) Received: from [192.168.0.3] ([62.3.217.253]) by rufus.isode.com via TCP (internal) with ESMTPA; Sun, 24 Apr 2005 17:49:26 +0100 Message-ID: <426BB84C.8060309@isode.com> Date: Sun, 24 Apr 2005 16:16:28 +0100 From: Alexey Melnikov User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.7.2) Gecko/20040804 Netscape/7.2 (ax) X-Accept-Language: en-us, en MIME-Version: 1.0 To: Russ Allbery CC: ietf-usefor@imc.org Subject: Re: Differences between RFC 2822 and Usefor (fwd) References: <87ekd5t0ae.fsf@windlord.stanford.edu> In-Reply-To: <87ekd5t0ae.fsf@windlord.stanford.edu> Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Russ Allbery wrote: >John Stanley writes: > > >>USEFOR-03 says it is defined as found in RFC2822, and is >>optional. Optional means "not mandatory". There is no MUST. >> >> > >I think it's obvious that the References header has to be optional in that >sense, since you can construct a well-formed Usenet message that doesn't >have a References header. > >I'll repeat my earlier suggestion here, which is to simply define >followups in such a way as to say that if something doesn't have a >References header, it's not a followup. That neatly avoids the whole >issue without needing to worry about MUSTs and SHOULDs. It becomes a >definitional issue. If it doesn't have References, it's not a followup. > > I fully agree. >>Like I already said, once you remove the ability to detect followups by >>making References optional, it doesn't matter when it may or may not be >>used. In fact, it may be used anytime (that's what "optional" means), >>even apparently when there is nothing at all for it to convey. >> >> >>By changing the requirement for References, we've solve a whole passle >>of things that should be be CHANGED BACK to the way they were before you >>decided to take it upon yourself to change them. Like "non-followups MAY >>contain" References headers, which, despite your repeated claims to the >>contrary, is a major change that nobody asked for. >> >> > >Well, to be fair, I think it's a reasonable way of dealing with the issue. >I'm not sure if that means I was asking for it, but I can see the logic >behind it. > >The way I look at this, there are two separate issues that we keep mixing >together: > > * What a news reader should do when constructing a posted reply to a > particular post. > > * How a news reader should display messages with References headers. > >The former is what we mean by a followup, and by definition it has to >build a References header. If it doesn't, it's doing something other than >posting a followup. > >The latter is a separate issue, a UI issue. It is expected that messages >connected by References will be shown in a thread, subject to various user >configuration and choices like whether changing the Subject should result >in a separate thread in the user's display (something that should be a >configuration option, not something the standard should be mandating >either way, as there are people who prefer it both ways and there is no >interoperability issue). > >There are various reasons why sometimes people like to connect messages >that are not posted replies to each other, and that do not, or only barely >do, fit into the notion of a followup. > >I don't think there's really anything else we can say that's more specific >than that. > > I agree. And my guess is that RFC 2822 says so little on the subject exactly for the same reasons. Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3MMdBaU019347 for ; Fri, 22 Apr 2005 15:39:11 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3MMdBfl019346 for ietf-usefor-skb; Fri, 22 Apr 2005 15:39:11 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from ciao.gmane.org (main.gmane.org [80.91.229.2]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3MMd9hm019333 for ; Fri, 22 Apr 2005 15:39:09 -0700 (PDT) (envelope-from usenet-format@gmane.org) Received: from list by ciao.gmane.org with local (Exim 4.43) id 1DP6iK-0001JV-JN for ietf-usefor@imc.org; Sat, 23 Apr 2005 00:33:32 +0200 Received: from 212.82.251.181 ([212.82.251.181]) by main.gmane.org with esmtp (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Sat, 23 Apr 2005 00:33:32 +0200 Received: from nobody by 212.82.251.181 with local (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Sat, 23 Apr 2005 00:33:32 +0200 X-Injected-Via-Gmane: http://gmane.org/ To: ietf-usefor@imc.org From: Frank Ellermann Subject: Obsolete vs. obsolescent (was: Fwd: http://www.ietf.org/internet-drafts/draft-ietf-usefor-usefor-03.txt) Date: Sat, 23 Apr 2005 00:33:26 +0200 Organization: Lines: 23 Message-ID: <42697BB6.6074@xyzzy.claranet.de> References: <200504111212.j3BCChM14516@clerew.man.ac.uk> <87psx1gnjc.fsf@windlord.stanford.edu> <87fyxvx1m6.fsf@windlord.stanford.edu> <87mzs2a1hu.fsf@windlord.stanford.edu> <87wtqxui6s.fsf@windlord.stanford.edu> <87d5so2bpk.fsf@windlord.stanford.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Complaints-To: usenet@sea.gmane.org X-Gmane-NNTP-Posting-Host: 212.82.251.181 X-Mailer: Mozilla 3.0 (OS/2; U) Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Charles Lindsey wrote: > "Obsolete" means this is no longer an official Netnews > header. It would be labelled as "obsolete" in the IANA > Registy of headers. Bruce has created a draft with mail headers found in obsolete RfCs. Among others Fcc: found in RfC 724 obsoleted by RfC 733. For news we'd declare See-Also: and Also-Control: as obsolete, as soon as Henry publishes s-o-1036 for information (obsoleted by - I'm an optimist here - draft-ietf-usefor-usefor-04.tx) > "Obsolescent" means that is is still an official Netnews > header (just), but that its continued use is discouraged > (for reasons that are given), and that it is likely to be > declared fully obsolete in a future standard. Now as pessimist: There won't be any future standard, UseFor is the last news standard before its death. So Lines: will be obsolescent "forever" until the last news server goes jabber, or atom over soap based on beep, or whatever comes next. Bye. Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3MKeJmu096898 for ; Fri, 22 Apr 2005 13:40:19 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3MKeJnl096897 for ietf-usefor-skb; Fri, 22 Apr 2005 13:40:19 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from a.mail.peak.org (a.mail.peak.org [69.59.192.41]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3MKeItM096864 for ; Fri, 22 Apr 2005 13:40:18 -0700 (PDT) (envelope-from stanley@peak.org) Received: from a.shell.peak.org ([69.59.192.81]) by a.mail.peak.org (8.12.10/8.12.8) with ESMTP id j3MKeDfO044114 (version=TLSv1/SSLv3 cipher=EDH-RSA-DES-CBC3-SHA bits=168 verify=NO) for ; Fri, 22 Apr 2005 13:40:13 -0700 (PDT) Date: Fri, 22 Apr 2005 13:40:13 -0700 (PDT) From: John Stanley X-X-Sender: stanley@a.shell.peak.org To: ietf-usefor@imc.org Subject: Re: References Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Spam-Score: 0 () X-Scanned-By: MIMEDefang 2.39 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: "Charles Lindsey" : >>Since References is now optional, yes, >No, they are NOT optional. Read Usepro. Read USEEFOR, where the header is defined, and then read RFC2822, which is what USEFOR refers to. That "SHOULD" in RFC2822 means it is not mandatory, it is just a "strong" recommendation. A recommendation is still just a recommendation. >Everything that comes out of a followup agent MUST have one. Not according to RFC2822, which is where the header is actually defined. >Please stop deliberately distorting what everyone else is saying. Coming from someone who deliberately and repeatedly attempts to put words in my mouth, your command is simply ludicrous. I'm not distorting anything; I'm reading the draft USEFOR standard and RFC2822. I suggest you try doing so before you toss about claims of distortion. >> something without References can be a followup. >No. Of course it can. RFC2822 says, and to avoid more condemnation from you that I am "distorting" something, I'll quote: references 0* 1 SHOULD occur in some replies - see 3.6.4 "Some replies". Doesn't even go as far as to say "SHOULD occur in all replies", just "some". And: Furthermore, reply messages SHOULD have "In-Reply-To:" and "References:" fields as appropriate, as described below. Just SHOULD. And: The "In-Reply-To:" field may be used to identify the message (or messages) to which the new message is a reply, while the "References:" field may be used to identify a "thread" of conversation. "May". Not a requirement. Furthermore, RFC2822 continues: If the parent message does not contain a "References:" field but does have an "In-Reply-To:" field containing a single message identifier, So, RFC2822 explicitely covers an example case of a "reply" (followup, in USENET terminology) that does NOT contain a References header. >No, its primary purpose is to facilitate threading, Threading is a user interface issue, and we do not deal with user interface issues. Furthermore, the "threading" you speak of is due to the followup nature of the articles it identifies as such. That makes its PRIMARY purpose to IDENTIFY followups (and the articles to which this is a followup); after that, the user agent can do what it wants with them when it displays them. >No. If you read USEPRO carefully, If you read RFC2822 carefully, which you must do if you want to understand USEFOR, you will see that References is OPTIONAL, and SHOULD appear in SOME replies. Not mandatory in anything. >you will see that threading (or other >means of presenting lists of articles for display) is an (optional) >feature of the protocol; It is ridiculous to argue that a header that implements an OPTIONAL "feature of the protocol" is in any way MANDATORY. But then, it's not a "feature of the protocol", it is just a user-interface issue, and that makes it even more ridiculous to argue that it is mandatory. Nothing at all cares about that header in the protocol; only the user agent that wants to do real threading cares, and it has to be able to deal with articles sans References headers already. It can deal with them; why can't you? >...they are entitled to assume that the headers they rely on for >the purpose (References, Subject, Date) have been properly constructed, Nobody but YOU is talking about the syntax of the headers. The issue at hand is a mandate for CONTAINING a header. Said mandate does not exist. It is an etirely different issue to say "IF you contain header X, it MUST be constructed this way" than "you MUST contain header X". The former is without question. We do it all the time. The latter requires a justification for interoperability on the EXISTANCE of the header, not just ability to decode it if it there. I'm quite happy if you want to say "the References header MUST contain ...". RFC2822 already deals with the question of when References headers are mandatory, and the answer is "they are not." >No it doesn't. It says they should be used for "Replies", which is not >quite the same thing. Close enough. All replies are followups. When RFC2822 says that the header SHOULD appear in replies, then whatever REPLIES it excludes from a mandate are also FOLLOWUPs that are excluded. >Well you are giving a good impression of not caring :-( . And even if you >don't care, then I still do. No, Charles, you do not. When you promote the concept "non-followups MAY contain References header", you exhibit a complete lack of caring about the ability to detect what is and is not a followup. It is only when the References header is in the current (RFC1036) MUST/MUST NOT version can you use it to identify a followup. MUST/MAY throws that ability away. >> and 2) the existing standard does, >Which is a good reason for keeping it so, One cannot "keep" what one has already let go. draft-usefor-03 removed the mandate. It's gone. If you want it back, justify it. "existing standard says" wasn't sufficient to keep it in usefor-03, so it's not sufficient to put it back. "Charles says" is even less convincing. >... and we should not change it >without a clear agreement in this WG to do so. Given your history of making changes to the drafts without any input from the WG at all, much less a "clear agreement", this sudden religion you have gotten for sticking to the process is, shall I say, questionable. And considering that we are discussing a change to the draft that was made with NO input from the WG in the first place ... >>Currently, you cannot construct a well-formed USENET message that is a >>followup without a References header. That is a direct effect of the >>language in RFC1036 that says "It is required for all follow-up messages". >And it's a direct effect of what USEPRO says. Baloney. USEPRO is a DRAFT standard. It has no effect on existing requirements, only on what might be required some day maybe. When I say "currently ...", USEPRO is irrelevant. IF USEFOR-03 ever makes it to standard, then USEFOR will tell me that I CAN construct a well-formed USENET message that is a followup without a References header, because RFC2822 says I can. POOF goes your caring about what is and is not a followup. >>We are not debating what is and is not a followup here. >Yes we are. Maybe you are, but the topic of THIS discussion is the lack of mandate for a References header, and that has NOTHING to do with what is or is not a followup, only on whether there is a mandate for using that header in any circumstance. RFC2822 says SHOULD. That is not a mandate. End of story. >And the opinion I am hearing (even from you now) is that it is >whatever a followup agent produces. Stop putting words in my mouth, Charles. I'm getting fucking tired of it. This is no longer just a difference of opinion, it's you pretending to speak for me, and you do NOT have the right to do that. Period. Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3MKIXhH091197 for ; Fri, 22 Apr 2005 13:18:33 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3MKIXm4091196 for ietf-usefor-skb; Fri, 22 Apr 2005 13:18:33 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from lentil.epix.net (lentil.epix.net [199.224.64.67]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3MKIWwI091189 for ; Fri, 22 Apr 2005 13:18:32 -0700 (PDT) (envelope-from mibsoft@epix.net) Received: from [192.168.2.11] (hrbg-66-33-227-136-pppoe.dsl.hrbg.epix.net [66.33.227.136]) by lentil.epix.net (8.12.10/2004120601/PL) with ESMTP id j3MKIMl8000281 for ; Fri, 22 Apr 2005 16:18:27 -0400 (EDT) Message-ID: <42695C10.2030408@epix.net> Date: Fri, 22 Apr 2005 16:18:24 -0400 From: "Forrest J. Cavalier III" User-Agent: Mozilla Thunderbird 0.7 (Windows/20040616) X-Accept-Language: en-us, en MIME-Version: 1.0 To: ietf-usefor@imc.org Subject: Re: References References: In-Reply-To: Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit X-Scanned-By: MIMEDefang 2.51 on 199.224.89.154 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: > Either we say "RFC1036 was right" regarding the References header and > followups and propogate that, with all the implications, into the new > standard, or we say "RFC1036 was wrong" and propogate that, with all the > implications, into the new standard. This half of one, half of the other > attitude is ridiculous. If RFC1036 is right, then Rererences is a > MUST/MUST NOT header, and followup is a valuable concept because we can > identify them at the recipient's end. If RFC1036 is wrong, then References > is a MAY/SHOULD/SOMETIMES header and followup is meaningless because we > cannot identify them. > > Me too. John has been pointing out problems with this all along, much past the point of getting ignored. So I want to be on record that I tend to think this paragraph sums it up. There is no reason USEFOR should get away with the result of something like a bad Abbot and Costello routine. Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3MJqjnU086990 for ; Fri, 22 Apr 2005 12:52:45 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3MJqjGV086989 for ietf-usefor-skb; Fri, 22 Apr 2005 12:52:45 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from a.mail.peak.org (a.mail.peak.org [69.59.192.41]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3MJqjkG086971 for ; Fri, 22 Apr 2005 12:52:45 -0700 (PDT) (envelope-from stanley@peak.org) Received: from a.shell.peak.org ([69.59.192.81]) by a.mail.peak.org (8.12.10/8.12.8) with ESMTP id j3MJqdfO024233 (version=TLSv1/SSLv3 cipher=EDH-RSA-DES-CBC3-SHA bits=168 verify=NO) for ; Fri, 22 Apr 2005 12:52:39 -0700 (PDT) Date: Fri, 22 Apr 2005 12:52:39 -0700 (PDT) From: John Stanley X-X-Sender: stanley@a.shell.peak.org To: ietf-usefor@imc.org Subject: Re: References Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Spam-Score: 0 () X-Scanned-By: MIMEDefang 2.39 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Russ Allbery : >I don't want to use any sort of SHOULD or MUST in this area (beyond the >one we inherit from RFC 2822). Ok. I'm happy with that. But we cannot then change our minds in mid-stream and put a SHOULD NOT or MUST into other pieces of the news standards, since that would contradict USEFOR and RFC2822. >I want to say that, in a Usenet context, a >followup is *defined* to be a message posted in reply to another message >with a References header pointing to the message being replied to. So what? Not "so what, you want to say", but "so what if a followup is defined that way?" You can't identify one as such using the RFC2822 headers, so why bother putting a special name on it? It's like defining a "narquark" as "an article posted by a person with a mole on his left shoulder." If you cannot identify "narquarks", you aren't able to treat them specially, so they are still just "articles" and get treated like any other. In other words, if it is important enough to DEFINE what a followup is (presumably so it can be treated specially), then it is important enough to provide a mechanism to identify followups at the point you want to apply that special treatment. The only person who really knows "this is a response to somethign else" is the poster. Without a References MUST/MUST NOT dichotomy, you cannot identify them once they leave his hands. You get "well, this sorta looks like a followup" and MAYBE it is, maybe it isn't. (And in case the argument that "it guesses right often enough", I'll point out that this is a TECHNICAL standard and TECHNICAL standards don't promote guessing, especially when they could promote fact just as easily.) >In other words, if your news posting software doesn't insert a References >header, it's still posting well-formed messages. It's just not posting >*followups*, by definition. Well, I guess we have to disagree here. I don't think you can isolate the concept of a "followup" from the intent of the poster. An article that someone posts using F in trn, where he puts in a new subject and entirely new content, but forgets to remove the References header, is not really a followup of any kind. The recipient would treat it as such based on the References header, but it really isn't. It's really an incorrectly formatted article (today) because it contains a header it is prohibited from containing based on its purpose. >In practice, I don't think there's any need to distinguish, for UI >purposes, between different reasons why an article cites a previous >article with References. Then "followup" is a meaningless concept, since it cannot be identified for special treatment. >The function the References header provides is enabling construction of >threads. This is closely related to the notion of followups but, to me, >is not identical to it. What is a "thread" if there is no "followup"? What is a "thread" when a "thread" may contain any collection of articles at all, not just ones that are somehow related in the real world? So, this looks like a complete turnaround in what I'm saying, huh? No followups, no threads, References meaningless? Nope. Here's the crux: Either we say "RFC1036 was right" regarding the References header and followups and propogate that, with all the implications, into the new standard, or we say "RFC1036 was wrong" and propogate that, with all the implications, into the new standard. This half of one, half of the other attitude is ridiculous. If RFC1036 is right, then Rererences is a MUST/MUST NOT header, and followup is a valuable concept because we can identify them at the recipient's end. If RFC1036 is wrong, then References is a MAY/SHOULD/SOMETIMES header and followup is meaningless because we cannot identify them. Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3MIVthl063985 for ; Fri, 22 Apr 2005 11:31:55 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3MIVtxs063984 for ietf-usefor-skb; Fri, 22 Apr 2005 11:31:55 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from b.mail.peak.org (b.mail.peak.org [69.59.192.42]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3MIVsBT063950 for ; Fri, 22 Apr 2005 11:31:54 -0700 (PDT) (envelope-from stanley@peak.org) Received: from a.shell.peak.org ([69.59.192.81]) by b.mail.peak.org (8.12.10/8.12.8) with ESMTP id j3MIVmI5037154 (version=TLSv1/SSLv3 cipher=EDH-RSA-DES-CBC3-SHA bits=168 verify=NO) for ; Fri, 22 Apr 2005 11:31:49 -0700 (PDT) Date: Fri, 22 Apr 2005 11:31:48 -0700 (PDT) From: John Stanley X-X-Sender: stanley@a.shell.peak.org To: ietf-usefor@imc.org Subject: Re: References Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Spam-Score: 0 () X-Scanned-By: MIMEDefang 2.39 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: "Charles Lindsey" : >Frank Ellermann writes: >>No References => no follow-up. >Which indeed follows from the fact that USEPRO says that a followup agent >MUST incorporate a References header (and do it correctly, of course). Unfortunately, RFC2822 says only SHOULD regarding the References header and followups, and USEFOR says that the References header is defined as found in RFC2822 (and for the pedants who cannot identify the context of this discussion, I'll add "other than a few syntax changes that don't change the requirements for usage, which is the context of this discussion here.) So References in a followup is only a SHOULD, and if USEPRO says otherwise, it contradicts both RFC2822 and USEFOR. While one of those we can modify for good reason, the other is the result of our modifications, and we did NOT modify the References header to make it mandatory in any use. >A pure logical deduction indeed. But for the removal of all doubt, and to >emphasise the difference from the treatment of "replies" in RFC 2822, we >used to say that as well. Now we don't (because our Chair so decrees), but >because it was redundant nothing has actually changed (John's claims to >the contrary notwithstanding). Stop putting words in my mouth, Charles. >Is is the the thing which is produced (as carefully defined in USEPRO) by >a "followup agent", or is it anything that has a References header. I >prefer the former (and you seem to have been using it in that sense), and >John has been arguing for the latter. Technically, it makes no difference >so long as we use the term consistently. Stop putting words in my mouth, Charles. You don't understand what I've said, so you are the last person who ought to be pretending you do. You being the editor and able to make willy-nilly changes to the drafts does NOT give you the right to put words specifically in MY mouth, even though you are putting words regularly into the mouths of the group as a whole. You are apparently incapable of differentiating between "poster intends a followup in what he sends" and "reader detects a followup in what he gets." The previous standards and drafts have said that the intent for something to be a followup by a poster REQUIRES a References header to be inserted (MUST), and otherwise that header is prohibited. That means that the READER can then determine what the poster intends to be a followup by looking for a References header. And yes, Charles, when you say "non-followups MAY contain a References header", you have removed that ability to DETECT at the reader's end of the pipe what is and is not intended to be a followup. That is a change, despite your repeated denials. Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3MIH7HA058335 for ; Fri, 22 Apr 2005 11:17:07 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3MIH7Tb058334 for ietf-usefor-skb; Fri, 22 Apr 2005 11:17:07 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from b.mail.peak.org (b.mail.peak.org [69.59.192.42]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3MIH6Gh058278 for ; Fri, 22 Apr 2005 11:17:06 -0700 (PDT) (envelope-from stanley@peak.org) Received: from a.shell.peak.org ([69.59.192.81]) by b.mail.peak.org (8.12.10/8.12.8) with ESMTP id j3MIH0I5031160 (version=TLSv1/SSLv3 cipher=EDH-RSA-DES-CBC3-SHA bits=168 verify=NO) for ; Fri, 22 Apr 2005 11:17:01 -0700 (PDT) Date: Fri, 22 Apr 2005 11:16:59 -0700 (PDT) From: John Stanley X-X-Sender: stanley@a.shell.peak.org To: ietf-usefor@imc.org Subject: Re: Differences between RFC 2822 and Usefor (fwd) Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Spam-Score: 0 () X-Scanned-By: MIMEDefang 2.39 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: "Charles Lindsey" : > No, USEPRO-03 does not say it is optional. Read what I wrote, Charles, and respond to that. I said: >USEFOR-03 says it is defined as found in RFC2822, and is optional. USEFOR is where the definition appears, and that definition defers to RFC2822 for almost everything. One of the areas it does NOT change RFC2822 is in the requirement for use of the header, which says "SHOULD". There is no MUST, either in RFC2822 or USEFOR. The header is optional, even in followups. Now, I realize that YOU say that RFC2119 "SHOULD" is a requirement of some kind, but RFC2119 actually says that it is a recommendation, and "required" and "recommended" are two very different words with two very different meanings, even when one of them carries an additional adjective "STRONG". Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3MGEPT6019201 for ; Fri, 22 Apr 2005 09:14:25 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3MGEPFQ019195 for ietf-usefor-skb; Fri, 22 Apr 2005 09:14:25 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from lon-mail-6.gradwell.net (lon-mail-6.gradwell.net [193.111.201.132]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3MGENdq019148 for ; Fri, 22 Apr 2005 09:14:24 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from host81-144-64-140.midband.mdip.bt.net ([81.144.64.140]) by lon-mail-6.gradwell.net with esmtp (Gradwell gwh-smtpd 1.181) id 426922de.cea8.b7 for ietf-usefor@imc.org; Fri, 22 Apr 2005 17:14:22 +0100 (envelope-sender ) Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3MGCK102367 for ietf-usefor@imc.org; Fri, 22 Apr 2005 17:12:20 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20706 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: References Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: Date: Fri, 22 Apr 2005 13:29:11 GMT Lines: 106 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In John Stanley writes: >Frank Ellermann : >Since References is now optional, yes, No, they are NOT optional. Read Usepro. Everything that comes out of a followup agent MUST have one. Please stop deliberately distorting what everyone else is saying. > something without References can be >a followup. No. And something WITH a References can be something other than a >followup. Yes. >The only purpose for the References header is to identify followups and to >what article the present one is a followup. No, its primary purpose is to facilitate threading, but it is also useful to identify followups, as you say, and also to enable quick retrieval of precursors by clicking on them. And to stitch message/partials together. And maybe other things. > Previous thought on the matter >was that this function ('identify followups') was important enough to >justify an RFC2119 mandate. It is, after all, how true threading works and >the only way it can be accomplished. But threading is a display issue, >not an interoperability one, so it was an abuse of the language to >pretend it merited an RFC2119 "MUST". No. If you read USEPRO carefully, you will see that threading (or other means of presenting lists of articles for display) is an (optional) feature of the protocol; i.e. implememtors MAY provide it and, if they so choose, then they are entitled to assume that the headers they rely on for the purpose (References, Subject, Date) have been properly constructed, and this merits MUST or SHOULD wording in the appropriate places. >Russ Allbery : >>So the question is whether we want to >>artificially define all those things as follow-ups as well, or not worry >>about it and just say that References is required for follow-ups but may >>also be used for other things. >RFC2822 says that References SHOULD be used in followups. No it doesn't. It says they should be used for "Replies", which is not quite the same thing. However, followups and replies are sufficiently similar that if we put a MUST somewhere regarding followups, we need to draw attention to the difference from the treatment of replies by RFC 2822. >And if you remove the one function that the header performs today >("identify a followup"), then why bother keeping that header? How do you >differentiate between "this article is a followup", and "this article >contains a References header for some other reason"? And if you say "who >cares?", I'll say 1) I do, Well you are giving a good impression of not caring :-( . And even if you don't care, then I still do. > and 2) the existing standard does, Which is a good reason for keeping it so, and we should not change it without a clear agreement in this WG to do so. >>I think it's obvious that the References header has to be optional in that >>sense, since you can construct a well-formed Usenet message that doesn't >>have a References header. >Currently, you cannot construct a well-formed USENET message that is a >followup without a References header. That is a direct effect of the >language in RFC1036 that says "It is required for all follow-up messages". And it's a direct effect of what USEPRO says. >>I'll repeat my earlier suggestion here, which is to simply define >>followups in such a way as to say that if something doesn't have a >>References header, it's not a followup. >We are not debating what is and is not a followup here. Yes we are. And the opinion I am hearing (even from you now) is that it is whatever a followup agent produces. I think, though, I would also want to include the case where a user constructs a followup manually (possibly to more than one precursor, as you seem to do) and generates a suitable References header manually. -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3MGEMLE019145 for ; Fri, 22 Apr 2005 09:14:22 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3MGEMri019144 for ietf-usefor-skb; Fri, 22 Apr 2005 09:14:22 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from lon-mail-6.gradwell.net (lon-mail-6.gradwell.net [193.111.201.132]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3MGELqK019137 for ; Fri, 22 Apr 2005 09:14:22 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from host81-144-64-140.midband.mdip.bt.net ([81.144.64.140]) by lon-mail-6.gradwell.net with esmtp (Gradwell gwh-smtpd 1.181) id 426922dc.cea8.b5 for ietf-usefor@imc.org; Fri, 22 Apr 2005 17:14:20 +0100 (envelope-sender ) Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3MGCIR02359 for ietf-usefor@imc.org; Fri, 22 Apr 2005 17:12:18 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20705 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: Fwd: http://www.ietf.org/internet-drafts/draft-ietf-usefor-usefor-03.txt Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: <200504111212.j3BCChM14516@clerew.man.ac.uk> <87psx1gnjc.fsf@windlord.stanford.edu> <87fyxvx1m6.fsf@windlord.stanford.edu> <87mzs2a1hu.fsf@windlord.stanford.edu> <87wtqxui6s.fsf@windlord.stanford.edu> <87d5so2bpk.fsf@windlord.stanford.edu> Date: Fri, 22 Apr 2005 12:09:06 GMT Lines: 27 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In <87d5so2bpk.fsf@windlord.stanford.edu> Russ Allbery writes: >I got this part, but it doesn't answer my confusion above. What's the >difference between obsolete and obsolescent? "Obsolete" means this is no longer an official Netnews header. It would be labelled as "obsolete" in the IANA Registy of headers. "Obsolescent" means that is is still an official Netnews header (just), but that its continued use is discouraged (for reasons that are given), and that it is likely to be declared fully obsolete in a future standard. Our earlier drafts uses the words "is to be regarded as obsolete", which falls slightly short of "is obsolete" and would IMO be better expressed as "obsolescent", if indeed that is the meaning we wish to convey (it was certainly the meaning I intended to convey in those drafts). -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3M4PZl3079903 for ; Thu, 21 Apr 2005 21:25:35 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3M4PZQo079902 for ietf-usefor-skb; Thu, 21 Apr 2005 21:25:35 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtp3.Stanford.EDU (smtp3.Stanford.EDU [171.67.16.138]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3M4PZnj079896 for ; Thu, 21 Apr 2005 21:25:35 -0700 (PDT) (envelope-from rra@stanford.edu) Received: from windlord.stanford.edu (windlord.Stanford.EDU [171.64.19.147]) by smtp3.Stanford.EDU (8.12.11/8.12.11) with SMTP id j3M4PYKL006950 for ; Thu, 21 Apr 2005 21:25:34 -0700 Received: (qmail 32648 invoked by uid 1000); 22 Apr 2005 04:25:34 -0000 To: ietf-usefor@imc.org Subject: Re: References In-Reply-To: <42686E91.42E2@xyzzy.claranet.de> (Frank Ellermann's message of "Fri, 22 Apr 2005 05:25:05 +0200") References: <4266EF35.21A5@xyzzy.claranet.de> <42686E91.42E2@xyzzy.claranet.de> From: Russ Allbery Organization: The Eyrie Date: Thu, 21 Apr 2005 21:25:34 -0700 Message-ID: <87fyxja14x.fsf@windlord.stanford.edu> User-Agent: Gnus/5.1007 (Gnus v5.10.7) XEmacs/21.4 (Jumbo Shrimp, linux) MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Frank Ellermann writes: > ACK. You could add "it has a References header field" to be > absolutely sure: > A "followup" is an article containing a response to the > contents of an earlier article (the followup's "precursor" > indicated in the "References" header field). This is the definition that I'd go for. -- Russ Allbery (rra@stanford.edu) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3M3ZNKO070504 for ; Thu, 21 Apr 2005 20:35:23 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3M3ZNl3070503 for ietf-usefor-skb; Thu, 21 Apr 2005 20:35:23 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from ciao.gmane.org (main.gmane.org [80.91.229.2]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3M3ZLJq070495 for ; Thu, 21 Apr 2005 20:35:22 -0700 (PDT) (envelope-from usenet-format@gmane.org) Received: from list by ciao.gmane.org with local (Exim 4.43) id 1DOosI-00022L-0N for ietf-usefor@imc.org; Fri, 22 Apr 2005 05:30:38 +0200 Received: from 212.82.251.14 ([212.82.251.14]) by main.gmane.org with esmtp (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Fri, 22 Apr 2005 05:30:37 +0200 Received: from nobody by 212.82.251.14 with local (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Fri, 22 Apr 2005 05:30:37 +0200 X-Injected-Via-Gmane: http://gmane.org/ To: ietf-usefor@imc.org From: Frank Ellermann Subject: Re: References Date: Fri, 22 Apr 2005 05:25:05 +0200 Organization: Lines: 37 Message-ID: <42686E91.42E2@xyzzy.claranet.de> References: <4266EF35.21A5@xyzzy.claranet.de> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Complaints-To: usenet@sea.gmane.org X-Gmane-NNTP-Posting-Host: 212.82.251.14 X-Mailer: Mozilla 3.0 (OS/2; U) Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Charles Lindsey wrote: > How do we define the term "followup". | A "followup" is an article containing a response to the | contents of an earlier article (the followup's "precursor"). Found in usefor-03. It's used for things like "Followup-To: ". Just above the definition of "Followup agent" in chapter 1.5 "Definitions". > Is is the the thing which is produced (as carefully defined > in USEPRO) by a "followup agent" Sure. > or is it anything that has a References header. Not sure. If it's the RfC 2822 idea of a reply coming from a mail2news gateway it probably is some "followup", but behind a gateway you never know. Maybe it only had an In-Reply-To, the gateway tried to fix it, but it was a broken pipermail archive In-ReplyTo (the thread is then correct, but the "followup" is shown at the wrong place in the thread). > I prefer the former (and you seem to have been using it in > that sense) ACK. You could add "it has a References header field" to be absolutely sure: A "followup" is an article containing a response to the contents of an earlier article (the followup's "precursor" indicated in the "References" header field). Bye, Frank Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3M2Nav8051286 for ; Thu, 21 Apr 2005 19:23:36 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3M2NaQN051279 for ietf-usefor-skb; Thu, 21 Apr 2005 19:23:36 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from lon-mail-6.gradwell.net (lon-mail-6.gradwell.net [193.111.201.132]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3M2NXYi051229 for ; Thu, 21 Apr 2005 19:23:34 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from host81-144-73-224.midband.mdip.bt.net ([81.144.73.224]) by lon-mail-6.gradwell.net with esmtp (Gradwell gwh-smtpd 1.181) id 42686024.1595d.ff for ietf-usefor@imc.org; Fri, 22 Apr 2005 03:23:32 +0100 (envelope-sender ) Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3M2CBl28135 for ietf-usefor@imc.org; Fri, 22 Apr 2005 03:12:11 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20700 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: References Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: <4266EF35.21A5@xyzzy.claranet.de> <87mzrtt1fh.fsf@windlord.stanford.edu> Date: Thu, 21 Apr 2005 16:29:29 GMT Lines: 30 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In <87mzrtt1fh.fsf@windlord.stanford.edu> Russ Allbery writes: >Sounds right to me. >If it doesn't have References, it's not a follow-up, but in practice >people use References for things other than follow-ups, and I don't see >any reason to ban those uses. So the question is whether we want to >artificially define all those things as follow-ups as well, or not worry >about it and just say that References is required for follow-ups but may >also be used for other things. Exactly. Those other uses for the References header deserve a brief mention somewhere, probably in USEFOR, but it would indeed be somewhat artificial to define those cases as "followups", and I personally do not want to do that. >The second seems clearer and simpler to me. Yes, that now seems to be where we are headed. -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3M2Na8u051285 for ; Thu, 21 Apr 2005 19:23:36 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3M2NanH051272 for ietf-usefor-skb; Thu, 21 Apr 2005 19:23:36 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from lon-mail-6.gradwell.net (lon-mail-6.gradwell.net [193.111.201.132]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3M2NYnn051234 for ; Thu, 21 Apr 2005 19:23:34 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from host81-144-73-224.midband.mdip.bt.net ([81.144.73.224]) by lon-mail-6.gradwell.net with esmtp (Gradwell gwh-smtpd 1.181) id 42686024.1595d.100 for ietf-usefor@imc.org; Fri, 22 Apr 2005 03:23:32 +0100 (envelope-sender ) Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3M2CAf28127 for ietf-usefor@imc.org; Fri, 22 Apr 2005 03:12:10 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20699 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: References (was: Differences between RFC 2822 and Usefor (fwd)) Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: <4266EF35.21A5@xyzzy.claranet.de> Date: Thu, 21 Apr 2005 16:22:39 GMT Lines: 62 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In <4266EF35.21A5@xyzzy.claranet.de> Frank Ellermann writes: >Charles Lindsey wrote: >> There has been no WG decision to change the requirement for >> a References header in followups. >I'm completely lost with this stuff. What's the precise >problem ? Something without References is never a follow-up. >No References => no follow-up. Which indeed follows from the fact that USEPRO says that a followup agent MUST incorporate a References header (and do it correctly, of course). >If that's the case we obviously also have: >Follow-up => references. A pure logical deduction indeed. But for the removal of all doubt, and to emphasise the difference from the treatment of "replies" in RFC 2822, we used to say that as well. Now we don't (because our Chair so decrees), but because it was redundant nothing has actually changed (John's claims to the contrary notwithstanding). >Based on this premise we don't have (and IMHO don't want): >References => follow-up But there is just one niggle here (yes, I agree we want sometimes to use References for multi-part FAQs and other special cases). How do we define the term "followup". Is is the the thing which is produced (as carefully defined in USEPRO) by a "followup agent", or is it anything that has a References header. I prefer the former (and you seem to have been using it in that sense), and John has been arguing for the latter. Technically, it makes no difference so long as we use the term consistently. >> I do not agree with the Chair on this one >I'm not even sure where you disagree. Is it related to >anything I said above ? It is because I preferred to mention it in both places, for the removal of all doubt as stated above, and also to provide a convenient place to introduce the special cases (FASQs etc) and to exaplain the semantics of the header properly. That still has to be done somewhere, and probably in USEFOR. -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3M1bd2K036542 for ; Thu, 21 Apr 2005 18:37:39 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3M1bdRF036541 for ietf-usefor-skb; Thu, 21 Apr 2005 18:37:39 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from ciao.gmane.org (main.gmane.org [80.91.229.2]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3M1bbL5036530 for ; Thu, 21 Apr 2005 18:37:38 -0700 (PDT) (envelope-from usenet-format@gmane.org) Received: from list by ciao.gmane.org with local (Exim 4.43) id 1DOn2O-0008Lb-Ab for ietf-usefor@imc.org; Fri, 22 Apr 2005 03:32:56 +0200 Received: from 212.82.251.14 ([212.82.251.14]) by main.gmane.org with esmtp (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Fri, 22 Apr 2005 03:32:56 +0200 Received: from nobody by 212.82.251.14 with local (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Fri, 22 Apr 2005 03:32:56 +0200 X-Injected-Via-Gmane: http://gmane.org/ To: ietf-usefor@imc.org From: Frank Ellermann Subject: Re: References Date: Fri, 22 Apr 2005 03:36:29 +0200 Organization: Lines: 10 Message-ID: <4268551D.5434@xyzzy.claranet.de> References: <87y8bb23jd.fsf@windlord.stanford.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Complaints-To: usenet@sea.gmane.org X-Gmane-NNTP-Posting-Host: 212.82.251.14 X-Mailer: Mozilla 3.0 (OS/2; U) Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Russ Allbery wrote: > people normally associate the idea of followup with a reply People: yes. Anybody who knows RfC 2822 and (s-o-) 1036: no. A followup is defined in usepro, a reply is defined in 2822, they are to great parts unrelated and / or different. There is not one string "followup" in the complete 2822 text. Bye. Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3LM0vKe075343 for ; Thu, 21 Apr 2005 15:00:57 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3LM0vXJ075341 for ietf-usefor-skb; Thu, 21 Apr 2005 15:00:57 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtp1.Stanford.EDU (smtp1.Stanford.EDU [171.67.16.123]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3LM0vgX075335 for ; Thu, 21 Apr 2005 15:00:57 -0700 (PDT) (envelope-from rra@stanford.edu) Received: from windlord.stanford.edu (windlord.Stanford.EDU [171.64.19.147]) by smtp1.Stanford.EDU (8.12.11/8.12.11) with SMTP id j3LM0tWn026241 for ; Thu, 21 Apr 2005 15:00:55 -0700 Received: (qmail 16756 invoked by uid 1000); 21 Apr 2005 22:00:54 -0000 To: ietf-usefor@imc.org Subject: Re: References In-Reply-To: (John Stanley's message of "Thu, 21 Apr 2005 10:49:17 -0700 (PDT)") References: From: Russ Allbery Organization: The Eyrie Date: Thu, 21 Apr 2005 15:00:54 -0700 Message-ID: <87y8bb23jd.fsf@windlord.stanford.edu> User-Agent: Gnus/5.1007 (Gnus v5.10.7) XEmacs/21.4 (Jumbo Shrimp, linux) MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: John Stanley writes: > RFC2822 says that References SHOULD be used in followups. This is not a > requirement, it is only a recommendation. What is the justification for > making References a requirement? What interoperability issue exists? > What breaks if a References header is not in an article, and don't say > "news readers", because news readers deal all the time with articles > that don't have References headers. They have to, since currently the > standard says that non-followups must not have them. I've yet to see a > news reader that doesn't deal with non-followups; I suggest that if you > have one, you get a better newsreader. I don't want to use any sort of SHOULD or MUST in this area (beyond the one we inherit from RFC 2822). I want to say that, in a Usenet context, a followup is *defined* to be a message posted in reply to another message with a References header pointing to the message being replied to. In other words, if your news posting software doesn't insert a References header, it's still posting well-formed messages. It's just not posting *followups*, by definition. (So if your software calls such messages followups, it has a serious definitional problem.) I think we're mostly agreeing. I want to take this approach since it avoids the whole question of whether things really break without References headers. > And if you remove the one function that the header performs today > ("identify a followup"), then why bother keeping that header? How do you > differentiate between "this article is a followup", and "this article > contains a References header for some other reason"? In practice, I don't think there's any need to distinguish, for UI purposes, between different reasons why an article cites a previous article with References. I just think it's a definitional stretch to call all such messages followups, since people normally associate the idea of followup with a reply and clearly not all of those articles are replies. > And if you say "who cares?", I'll say 1) I do, and 2) the existing > standard does, and 3) if nobody does, then get rid of the header, since > it performs no useful function. The function the References header provides is enabling construction of threads. This is closely related to the notion of followups but, to me, is not identical to it. > True wrt RFC1036. Untrue wrt draft-usefor-03. SHOULD is a > recommendation. However strong you think it is, it is still only a > recommendation. Right, I agree that what I'm pushing for is not what's in the current language. -- Russ Allbery (rra@stanford.edu) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3LJ4Pp4036310 for ; Thu, 21 Apr 2005 12:04:25 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3LJ4Po5036309 for ietf-usefor-skb; Thu, 21 Apr 2005 12:04:25 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtp3.Stanford.EDU (smtp3.Stanford.EDU [171.67.16.138]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3LJ4Pgn036303 for ; Thu, 21 Apr 2005 12:04:25 -0700 (PDT) (envelope-from rra@stanford.edu) Received: from windlord.stanford.edu (windlord.Stanford.EDU [171.64.19.147]) by smtp3.Stanford.EDU (8.12.11/8.12.11) with SMTP id j3LJ4NSv006253 for ; Thu, 21 Apr 2005 12:04:24 -0700 Received: (qmail 11849 invoked by uid 1000); 21 Apr 2005 19:04:23 -0000 To: ietf-usefor@imc.org Subject: Re: Fwd: http://www.ietf.org/internet-drafts/draft-ietf-usefor-usefor-03.txt In-Reply-To: (Charles Lindsey's message of "Thu, 21 Apr 2005 15:44:40 GMT") References: <200504111212.j3BCChM14516@clerew.man.ac.uk> <87psx1gnjc.fsf@windlord.stanford.edu> <87fyxvx1m6.fsf@windlord.stanford.edu> <87mzs2a1hu.fsf@windlord.stanford.edu> <87wtqxui6s.fsf@windlord.stanford.edu> From: Russ Allbery Organization: The Eyrie Date: Thu, 21 Apr 2005 12:04:23 -0700 Message-ID: <87d5so2bpk.fsf@windlord.stanford.edu> User-Agent: Gnus/5.1007 (Gnus v5.10.7) XEmacs/21.4 (Jumbo Shrimp, linux) MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Charles Lindsey writes: > Russ Allbery writes: >> Then I don't understand the distinction you're drawing between obsolete >> and obsolescent. As near as I could tell, the only difference in >> practice was that servers were rejecting the former headers, but now >> you're saying that they shouldn't even do that. > Generally speaking, our draft has always said (and I think Usefor still > says) that if you see a header that you do not recognize as a news > header, you just leave it be and ignore it. Essentially the same applies > in Mail. I got this part, but it doesn't answer my confusion above. What's the difference between obsolete and obsolescent? -- Russ Allbery (rra@stanford.edu) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3LHnP0x011354 for ; Thu, 21 Apr 2005 10:49:25 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3LHnPfM011353 for ietf-usefor-skb; Thu, 21 Apr 2005 10:49:25 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from a.mail.peak.org (a.mail.peak.org [69.59.192.41]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3LHnODg011333 for ; Thu, 21 Apr 2005 10:49:24 -0700 (PDT) (envelope-from stanley@peak.org) Received: from a.shell.peak.org ([69.59.192.81]) by a.mail.peak.org (8.12.10/8.12.8) with ESMTP id j3LHnHfO031096 (version=TLSv1/SSLv3 cipher=EDH-RSA-DES-CBC3-SHA bits=168 verify=NO) for ; Thu, 21 Apr 2005 10:49:19 -0700 (PDT) Date: Thu, 21 Apr 2005 10:49:17 -0700 (PDT) From: John Stanley X-X-Sender: stanley@a.shell.peak.org To: ietf-usefor@imc.org Subject: References Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Spam-Score: 0 () X-Scanned-By: MIMEDefang 2.39 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Frank Ellermann : >I'm completely lost with this stuff. What's the precise >problem ? Something without References is never a follow-up. Since References is now optional, yes, something without References can be a followup. And something WITH a References can be something other than a followup. The current language says that followups SHOULD have that header, it does not prohibit it in anything else. >If that's the case we obviously also have: >Follow-up => references. Not obvious. The header is optional. See RFC2822. >Based on this premise we don't have (and IMHO don't want): >References => follow-up The only purpose for the References header is to identify followups and to what article the present one is a followup. Previous thought on the matter was that this function ('identify followups') was important enough to justify an RFC2119 mandate. It is, after all, how true threading works and the only way it can be accomplished. But threading is a display issue, not an interoperability one, so it was an abuse of the language to pretend it merited an RFC2119 "MUST". If you have some other function in mind, please enlighten us, and then justify any mandates you want to apply. What interoperability problem does making References mandatory solve? >Now to the claim that it's exactly the same as in RfC 2822: >That's not exactly true. Of course it isn't true. Nobody has said it is. Read draft-usefor-03 to see the differences. None of those differences deals with a mandate for use. Notice that "we" were able to insert the differences that "we" thought were important from RFC2822. "We" apparently didn't think a MUST was important; I agree. Let's move on. Russ Allbery : >So the question is whether we want to >artificially define all those things as follow-ups as well, or not worry >about it and just say that References is required for follow-ups but may >also be used for other things. RFC2822 says that References SHOULD be used in followups. This is not a requirement, it is only a recommendation. What is the justification for making References a requirement? What interoperability issue exists? What breaks if a References header is not in an article, and don't say "news readers", because news readers deal all the time with articles that don't have References headers. They have to, since currently the standard says that non-followups must not have them. I've yet to see a news reader that doesn't deal with non-followups; I suggest that if you have one, you get a better newsreader. And if you remove the one function that the header performs today ("identify a followup"), then why bother keeping that header? How do you differentiate between "this article is a followup", and "this article contains a References header for some other reason"? And if you say "who cares?", I'll say 1) I do, and 2) the existing standard does, and 3) if nobody does, then get rid of the header, since it performs no useful function. >> So what's the problem ? It's an optional header field as >> in RfC 2822, and its presence does not automatically mean >> "follow-up", it could be also something else. >Exactly. So the purpose of References headers is now deprecated. That's an argument for making them optional (as we have already done), not for making them a requirement. >I think it's obvious that the References header has to be optional in that >sense, since you can construct a well-formed Usenet message that doesn't >have a References header. Currently, you cannot construct a well-formed USENET message that is a followup without a References header. That is a direct effect of the language in RFC1036 that says "It is required for all follow-up messages". While the article may appear to be well-formed, it is technically in violation of the standards and thus is not valid. This is exactly the reason why I argued that a specific proposed moderated newsgroup that was intended to contain only followups should enforce that requirement, and, as I recall, you agreed with the technical argument but disagreed with doing that for social/political reasons. Correct me if I am wrong in my recollection. Once you say it is optional, you cannot also say it is mandatory. If USEFOR defines valid USENET format messages, and USEFOR says that a header is optional, then any message that does not contain that header can be valid. Another standard for the same medium that says "this article is not valid", when USEFOR says it is, is contradictory and broken. >I'll repeat my earlier suggestion here, which is to simply define >followups in such a way as to say that if something doesn't have a >References header, it's not a followup. We are not debating what is and is not a followup here. We are discussing the fact that our draft now defers the definition of References to RFC2822 with certain limited differences, none of which deal with changing the RFC2119 status of that header. References is now a SHOULD, not a MUST. If it is to be changed to a MUST, then justify the change. If you cannot justify the change, it ought not be made. This group has suffered too many times from unilateral changes made at the whim of the person who controls the text to allow it to continue. Charles says it needs to be changed back; Charles ought to justify that "need" or drop it. > * What a news reader should do when constructing a posted reply to a > particular post. >[That] is what we mean by a followup, and by definition it has to >build a References header. If it doesn't, it's doing something other than >posting a followup. True wrt RFC1036. Untrue wrt draft-usefor-03. SHOULD is a recommendation. However strong you think it is, it is still only a recommendation. Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3LGDZO7084549 for ; Thu, 21 Apr 2005 09:13:35 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3LGDZst084548 for ietf-usefor-skb; Thu, 21 Apr 2005 09:13:35 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from lon-mail-2.gradwell.net (lon-mail-2.gradwell.net [193.111.201.126]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3LGDYhS084542 for ; Thu, 21 Apr 2005 09:13:35 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from host81-144-72-42.midband.mdip.bt.net ([81.144.72.42]) by lon-mail-2.gradwell.net with esmtp (Gradwell gwh-smtpd 1.181) id 4267d12d.858f.b for ietf-usefor@imc.org; Thu, 21 Apr 2005 17:13:33 +0100 (envelope-sender ) Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3LGCXh23009 for ietf-usefor@imc.org; Thu, 21 Apr 2005 17:12:33 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20697 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: Fwd: http://www.ietf.org/internet-drafts/draft-ietf-usefor-usefor-03.txt Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: <200504111212.j3BCChM14516@clerew.man.ac.uk> <87psx1gnjc.fsf@windlord.stanford.edu> <87fyxvx1m6.fsf@windlord.stanford.edu> <87mzs2a1hu.fsf@windlord.stanford.edu> <87wtqxui6s.fsf@windlord.stanford.edu> Date: Thu, 21 Apr 2005 15:44:40 GMT Lines: 34 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In <87wtqxui6s.fsf@windlord.stanford.edu> Russ Allbery writes: >Charles Lindsey writes: >> Ah! But in that case I don't think servers should be rejecting such >> headers. If the article has been wandering in any out of both email and >> news systems, then it is sometimes useful to have the full history >> recorded (though I might agree that if there was a mailing list involved >> I might not want to know what happened before it reached the mailing >> list). >Then I don't understand the distinction you're drawing between obsolete >and obsolescent. As near as I could tell, the only difference in practice >was that servers were rejecting the former headers, but now you're saying >that they shouldn't even do that. Generally speaking, our draft has always said (and I think Usefor still says) that if you see a header that you do not recognize as a news header, you just leave it be and ignore it. Essentially the same applies in Mail. The only time you should be worrying is if you DO recognize it as a news header, and can see that it is malformed and likely to cause trouble elsewhere (and especially so if you are an injecting agent). -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3LGDYT7084539 for ; Thu, 21 Apr 2005 09:13:34 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3LGDYri084537 for ietf-usefor-skb; Thu, 21 Apr 2005 09:13:34 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from lon-mail-2.gradwell.net (lon-mail-2.gradwell.net [193.111.201.126]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3LGDXsK084531 for ; Thu, 21 Apr 2005 09:13:33 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from host81-144-72-42.midband.mdip.bt.net ([81.144.72.42]) by lon-mail-2.gradwell.net with esmtp (Gradwell gwh-smtpd 1.181) id 4267d12b.858f.9 for ietf-usefor@imc.org; Thu, 21 Apr 2005 17:13:31 +0100 (envelope-sender ) Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3LGCYU23014 for ietf-usefor@imc.org; Thu, 21 Apr 2005 17:12:34 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20698 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: Differences between RFC 2822 and Usefor (fwd) Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: Date: Thu, 21 Apr 2005 15:55:38 GMT Lines: 29 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In John Stanley writes: >"Charles Lindsey" : >>Our Chair has >>taken the view that Usefor is not the place to say it, but that Usepro IS >>the proper place (and indeed there is already a corresponding MUST in >>Usepro, though it may need a little more amplification). >USEFOR-03 says it is defined as found in RFC2822, and is optional. >Optional means "not mandatory". There is no MUST. No, USEPRO-03 does not say it is optional. It is silent on the issue (and our Chair confirms this by saying that RFC 2822 contains both format and protocol material, and so if you want to see whether some protocol material in RFC 2822 still applies, then you have to look at USEPRO). I am not sure I like that way of looking at it, but it is what our Chair says he wants. -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3L9jQv7023930 for ; Thu, 21 Apr 2005 02:45:26 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3L9jQY4023929 for ietf-usefor-skb; Thu, 21 Apr 2005 02:45:26 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from mail.gmx.net (pop.gmx.de [213.165.64.20]) by above.proper.com (8.12.11/8.12.9) with SMTP id j3L9jOXT023851 for ; Thu, 21 Apr 2005 02:45:25 -0700 (PDT) (envelope-from mail@sebastian-brocks.de) Received: (qmail invoked by alias); 21 Apr 2005 09:45:18 -0000 Received: from xdsl-213-196-200-53.netcologne.de (EHLO [192.168.1.2]) [213.196.200.53] by mail.gmx.net (mp029) with SMTP; 21 Apr 2005 11:45:18 +0200 X-Authenticated: #1840277 Message-ID: <4267762D.6020902@sebastian-brocks.de> Date: Thu, 21 Apr 2005 11:45:17 +0200 From: Sebastian Brocks User-Agent: Mozilla/5.0 (X11; U; Linux i686; de-DE; rv:1.7.6) Gecko/20050404 Thunderbird/1.0.2 Mnenhy/0.7.2.0 X-Accept-Language: de-DE, de, en-us, en MIME-Version: 1.0 To: ietf-usefor@imc.org Subject: Re: Differences between RFC 2822 and Usefor (fwd) References: <87ekd5t0ae.fsf@windlord.stanford.edu> In-Reply-To: <87ekd5t0ae.fsf@windlord.stanford.edu> X-Enigmail-Version: 0.90.0.0 X-Enigmail-Supports: pgp-inline, pgp-mime Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit X-Y-GMX-Trusted: 0 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Russ Allbery schrieb: > I don't think there's really anything else we can say that's more specific > than that. I agree. greetings, Sebastian Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3L0uhi2002254 for ; Wed, 20 Apr 2005 17:56:43 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3L0uh1E002253 for ietf-usefor-skb; Wed, 20 Apr 2005 17:56:43 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtp2.Stanford.EDU (smtp2.Stanford.EDU [171.67.16.125]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3L0ugPx002247 for ; Wed, 20 Apr 2005 17:56:42 -0700 (PDT) (envelope-from rra@stanford.edu) Received: from windlord.stanford.edu (windlord.Stanford.EDU [171.64.19.147]) by smtp2.Stanford.EDU (8.12.11/8.12.11) with SMTP id j3L0ugCD005051 for ; Wed, 20 Apr 2005 17:56:42 -0700 Received: (qmail 9461 invoked by uid 1000); 21 Apr 2005 00:56:41 -0000 To: ietf-usefor@imc.org Subject: Re: Differences between RFC 2822 and Usefor (fwd) In-Reply-To: (John Stanley's message of "Wed, 20 Apr 2005 17:24:47 -0700 (PDT)") References: From: Russ Allbery Organization: The Eyrie Date: Wed, 20 Apr 2005 17:56:41 -0700 Message-ID: <87ekd5t0ae.fsf@windlord.stanford.edu> User-Agent: Gnus/5.1007 (Gnus v5.10.7) XEmacs/21.4 (Jumbo Shrimp, linux) MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: John Stanley writes: > USEFOR-03 says it is defined as found in RFC2822, and is > optional. Optional means "not mandatory". There is no MUST. I think it's obvious that the References header has to be optional in that sense, since you can construct a well-formed Usenet message that doesn't have a References header. I'll repeat my earlier suggestion here, which is to simply define followups in such a way as to say that if something doesn't have a References header, it's not a followup. That neatly avoids the whole issue without needing to worry about MUSTs and SHOULDs. It becomes a definitional issue. If it doesn't have References, it's not a followup. > Like I already said, once you remove the ability to detect followups by > making References optional, it doesn't matter when it may or may not be > used. In fact, it may be used anytime (that's what "optional" means), > even apparently when there is nothing at all for it to convey. > By changing the requirement for References, we've solve a whole passle > of things that should be be CHANGED BACK to the way they were before you > decided to take it upon yourself to change them. Like "non-followups MAY > contain" References headers, which, despite your repeated claims to the > contrary, is a major change that nobody asked for. Well, to be fair, I think it's a reasonable way of dealing with the issue. I'm not sure if that means I was asking for it, but I can see the logic behind it. The way I look at this, there are two separate issues that we keep mixing together: * What a news reader should do when constructing a posted reply to a particular post. * How a news reader should display messages with References headers. The former is what we mean by a followup, and by definition it has to build a References header. If it doesn't, it's doing something other than posting a followup. The latter is a separate issue, a UI issue. It is expected that messages connected by References will be shown in a thread, subject to various user configuration and choices like whether changing the Subject should result in a separate thread in the user's display (something that should be a configuration option, not something the standard should be mandating either way, as there are people who prefer it both ways and there is no interoperability issue). There are various reasons why sometimes people like to connect messages that are not posted replies to each other, and that do not, or only barely do, fit into the notion of a followup. I don't think there's really anything else we can say that's more specific than that. -- Russ Allbery (rra@stanford.edu) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3L0W4x9098581 for ; Wed, 20 Apr 2005 17:32:04 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3L0W4YW098580 for ietf-usefor-skb; Wed, 20 Apr 2005 17:32:04 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtp2.Stanford.EDU (smtp2.Stanford.EDU [171.67.16.125]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3L0W38A098574 for ; Wed, 20 Apr 2005 17:32:03 -0700 (PDT) (envelope-from rra@stanford.edu) Received: from windlord.stanford.edu (windlord.Stanford.EDU [171.64.19.147]) by smtp2.Stanford.EDU (8.12.11/8.12.11) with SMTP id j3L0W2g5031344 for ; Wed, 20 Apr 2005 17:32:02 -0700 Received: (qmail 8989 invoked by uid 1000); 21 Apr 2005 00:32:02 -0000 To: ietf-usefor@imc.org Subject: Re: References In-Reply-To: <4266EF35.21A5@xyzzy.claranet.de> (Frank Ellermann's message of "Thu, 21 Apr 2005 02:09:25 +0200") References: <4266EF35.21A5@xyzzy.claranet.de> From: Russ Allbery Organization: The Eyrie Date: Wed, 20 Apr 2005 17:32:02 -0700 Message-ID: <87mzrtt1fh.fsf@windlord.stanford.edu> User-Agent: Gnus/5.1007 (Gnus v5.10.7) XEmacs/21.4 (Jumbo Shrimp, linux) MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Frank Ellermann writes: > Based on this premise we don't have (and IMHO don't want): > References => follow-up > That's elementary logic as far as I'm concerned. We don't > need your co-editor or Alexey for this issue, I'm 100% sure. Sounds right to me. If it doesn't have References, it's not a follow-up, but in practice people use References for things other than follow-ups, and I don't see any reason to ban those uses. So the question is whether we want to artificially define all those things as follow-ups as well, or not worry about it and just say that References is required for follow-ups but may also be used for other things. The second seems clearer and simpler to me. > Now to the claim that it's exactly the same as in RfC 2822: > That's not exactly true. We truncate References, RfC 2822 > does not, that's a Usepro issue. We separate References by > at least one WSP, RfC 2822 does not, that's a Usefor issue. Yup. > RfC 2822 fixes In-Reply-To without References, we'd expect > the same from a mail2news gateway, but don't need it within > news. It doesn't hurt, but it's not required. Yup. > So what's the problem ? It's an optional header field as > in RfC 2822, and its presence does not automatically mean > "follow-up", it could be also something else. Exactly. -- Russ Allbery (rra@stanford.edu) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3L0Os0L097583 for ; Wed, 20 Apr 2005 17:24:54 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3L0OsHd097582 for ietf-usefor-skb; Wed, 20 Apr 2005 17:24:54 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from b.mail.peak.org (b.mail.peak.org [69.59.192.42]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3L0Orf5097573 for ; Wed, 20 Apr 2005 17:24:53 -0700 (PDT) (envelope-from stanley@peak.org) Received: from a.shell.peak.org ([69.59.192.81]) by b.mail.peak.org (8.12.10/8.12.8) with ESMTP id j3L0OlI5021878 (version=TLSv1/SSLv3 cipher=EDH-RSA-DES-CBC3-SHA bits=168 verify=NO) for ; Wed, 20 Apr 2005 17:24:48 -0700 (PDT) Date: Wed, 20 Apr 2005 17:24:47 -0700 (PDT) From: John Stanley X-X-Sender: stanley@a.shell.peak.org To: ietf-usefor@imc.org Subject: Re: Differences between RFC 2822 and Usefor (fwd) Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Spam-Score: 0 () X-Scanned-By: MIMEDefang 2.39 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: "Charles Lindsey" : >There has been no WG decision to >change the requirement for a References header in followups. Not according to USEFOR-03. >Our Chair has >taken the view that Usefor is not the place to say it, but that Usepro IS >the proper place (and indeed there is already a corresponding MUST in >Usepro, though it may need a little more amplification). USEFOR-03 says it is defined as found in RFC2822, and is optional. Optional means "not mandatory". There is no MUST. >But in that case, I will have to find somewhere in Usepro to say that the >References header MAY also be used in other cases such as multipart FAQs, If it is optional, then it is optional. You don't need to say it MAY be used when USEFOR-03 already says it may be used. Like I already said, once you remove the ability to detect followups by making References optional, it doesn't matter when it may or may not be used. In fact, it may be used anytime (that's what "optional" means), even apparently when there is nothing at all for it to convey. By changing the requirement for References, we've solve a whole passle of things that should be be CHANGED BACK to the way they were before you decided to take it upon yourself to change them. Like "non-followups MAY contain" References headers, which, despite your repeated claims to the contrary, is a major change that nobody asked for. >News readers that try to do threading care very much if it is missing. That is not an interoperability issue, and you know it. It doesn't cause the news reader to break, unless it is a VERY poorly designed news reader. How does it deal with non-followups today, which do not contain References headers, or broken followups, which also do not contain them, if they cannot deal with any article tomorrow that does not contain one? If they break today, they are patently broken and need to be fixed. I know of none that are that broken. Do you? And, by the way, it's software. Until it gains sentience, it doesn't care about anything. >No, the requirement has not gone. Read the draft-03. It is gone. Explicit language that makes References optional. "As defined in RFC2822". No MUST. A MUST that doesn't exist is not a requirement. >The mandate in question is already in USEPRO. And always has been. But is missing completely in USEFOR, where the header is defined. It's gone. It's time to stop pretending that it's there, or that it wasn't a deliberate change to remove it. I'm ok with it gone; why are you suddenly trying to say that it needs to be put back? You didn't say that prior to my commenting on it, it must have been ok -- you had ample time to comment on the draft prior to my speaking up, and did not. Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3L0ESqY095750 for ; Wed, 20 Apr 2005 17:14:28 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3L0ESxh095749 for ietf-usefor-skb; Wed, 20 Apr 2005 17:14:28 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from ciao.gmane.org (main.gmane.org [80.91.229.2]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3L0EQX0095742 for ; Wed, 20 Apr 2005 17:14:27 -0700 (PDT) (envelope-from usenet-format@gmane.org) Received: from list by ciao.gmane.org with local (Exim 4.43) id 1DOPGU-0002Cv-QZ for ietf-usefor@imc.org; Thu, 21 Apr 2005 02:09:54 +0200 Received: from 212.82.251.239 ([212.82.251.239]) by main.gmane.org with esmtp (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Thu, 21 Apr 2005 02:09:54 +0200 Received: from nobody by 212.82.251.239 with local (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Thu, 21 Apr 2005 02:09:54 +0200 X-Injected-Via-Gmane: http://gmane.org/ To: ietf-usefor@imc.org From: Frank Ellermann Subject: References (was: Differences between RFC 2822 and Usefor (fwd)) Date: Thu, 21 Apr 2005 02:09:25 +0200 Organization: Lines: 46 Message-ID: <4266EF35.21A5@xyzzy.claranet.de> References: Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Complaints-To: usenet@sea.gmane.org X-Gmane-NNTP-Posting-Host: 212.82.251.239 X-Mailer: Mozilla 3.0 (OS/2; U) Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Charles Lindsey wrote: > I do not want to be seen arguing with my co-editor in public > if it can be avoided. It would be rather nice if your co-editor says something in public, like "hi there" for starters. > There has been no WG decision to change the requirement for > a References header in followups. I'm completely lost with this stuff. What's the precise problem ? Something without References is never a follow-up. No References => no follow-up. If that's the case we obviously also have: Follow-up => references. Based on this premise we don't have (and IMHO don't want): References => follow-up That's elementary logic as far as I'm concerned. We don't need your co-editor or Alexey for this issue, I'm 100% sure. Now to the claim that it's exactly the same as in RfC 2822: That's not exactly true. We truncate References, RfC 2822 does not, that's a Usepro issue. We separate References by at least one WSP, RfC 2822 does not, that's a Usefor issue. RfC 2822 fixes In-Reply-To without References, we'd expect the same from a mail2news gateway, but don't need it within news. It doesn't hurt, but it's not required. So what's the problem ? It's an optional header field as in RfC 2822, and its presence does not automatically mean "follow-up", it could be also something else. > I do not agree with the Chair on this one I'm not even sure where you disagree. Is it related to anything I said above ? Bye, Frank Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3KNij22088559 for ; Wed, 20 Apr 2005 16:44:45 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3KNijiM088558 for ietf-usefor-skb; Wed, 20 Apr 2005 16:44:45 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtp2.Stanford.EDU (smtp2.Stanford.EDU [171.67.16.125]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3KNiiND088549 for ; Wed, 20 Apr 2005 16:44:45 -0700 (PDT) (envelope-from rra@stanford.edu) Received: from windlord.stanford.edu (windlord.Stanford.EDU [171.64.19.147]) by smtp2.Stanford.EDU (8.12.11/8.12.11) with SMTP id j3KNiirY017262 for ; Wed, 20 Apr 2005 16:44:44 -0700 Received: (qmail 7383 invoked by uid 1000); 20 Apr 2005 23:44:43 -0000 To: ietf-usefor@imc.org Subject: Re: Fwd: http://www.ietf.org/internet-drafts/draft-ietf-usefor-usefor-03.txt In-Reply-To: (Charles Lindsey's message of "Wed, 20 Apr 2005 19:45:12 GMT") References: <200504111212.j3BCChM14516@clerew.man.ac.uk> <87psx1gnjc.fsf@windlord.stanford.edu> <87fyxvx1m6.fsf@windlord.stanford.edu> <87mzs2a1hu.fsf@windlord.stanford.edu> From: Russ Allbery Organization: The Eyrie Date: Wed, 20 Apr 2005 16:44:43 -0700 Message-ID: <87wtqxui6s.fsf@windlord.stanford.edu> User-Agent: Gnus/5.1007 (Gnus v5.10.7) XEmacs/21.4 (Jumbo Shrimp, linux) MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Charles Lindsey writes: > Ah! But in that case I don't think servers should be rejecting such > headers. If the article has been wandering in any out of both email and > news systems, then it is sometimes useful to have the full history > recorded (though I might agree that if there was a mailing list involved > I might not want to know what happened before it reached the mailing > list). Then I don't understand the distinction you're drawing between obsolete and obsolescent. As near as I could tell, the only difference in practice was that servers were rejecting the former headers, but now you're saying that they shouldn't even do that. -- Russ Allbery (rra@stanford.edu) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3KMvLQN078219 for ; Wed, 20 Apr 2005 15:57:21 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3KMvL9L078218 for ietf-usefor-skb; Wed, 20 Apr 2005 15:57:21 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from probity.mcc.ac.uk (probity.mcc.ac.uk [130.88.200.94]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3KMvKiY078211 for ; Wed, 20 Apr 2005 15:57:20 -0700 (PDT) (envelope-from chl@clerew.man.ac.uk) Received: from dialup027.mcc.ac.uk ([130.88.69.27]) by probity.mcc.ac.uk with esmtp (Exim 4.20) id 1DOO8F-000CWp-Dk for ietf-usefor@imc.org; Wed, 20 Apr 2005 23:57:19 +0100 Received: (from chl@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3KMUe715315; Wed, 20 Apr 2005 23:30:40 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20688 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: Differences between RFC 2822 and Usefor (fwd) Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: Date: Wed, 20 Apr 2005 20:03:27 GMT Lines: 60 MIME-Version: 1.0 X-Scanner: exiscan for exim4 (http://duncanthrax.net/exiscan/) *1DOO8F-000CWp-Dk*h0tR2iM6i9I* Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In John Stanley writes: >"Charles Lindsey" : >>The "MUST" in question arises from the long held consensus in this Working >>Group that the References header in News is NOT an "optional extra". >Not according to draft-usefor-03. According to draft-usefor-03, >References is defined to be the same as in RFC2822, with the addition of a >few bits that are not mandates for using it. This consensus you claim >exists has a very odd way of being expressed in our products. Almost as if >it doesn't really exist. >The fact that I am the only one who even commented on it being gone is a >pretty good sign that nobody cares that it is gone. No, I knew that it wasn't there, and there are a lot of other things that should be in the Usefor draft and that have not made it there yet. But I do not want to be seen arguing with my co-editor in public if it can be avoided. Especially on issues that do not appear to be the subject of controversy. >>That clear requirement has been in all our drafts up to article-13. >It is no longer a requirement at all in our draft. Indeed, but a draft is only a draft. There has been no WG decision to change the requirement for a References header in followups. Our Chair has taken the view that Usefor is not the place to say it, but that Usepro IS the proper place (and indeed there is already a corresponding MUST in Usepro, though it may need a little more amplification). I do not agree with the Chair on this one, but if nobody else speaks up and if his ruling stands, then Usepro is where it will be said (which also, incidentally, removes most of the cause of the disagreement there has been about how to word it all). But in that case, I will have to find somewhere in Usepro to say that the References header MAY also be used in other cases such as multipart FAQs, as well as pointing out the divergence from the SHOULD in RFC 2822. >At this point, the mandate is not part of the draft, and putting one in >would (or at least, should, in an honest system) require a strong >demonstration of interoperability issues. Since news systems simply do not >care if the header is in the article or not, there is no way to >demonstrate an interoperability issue, and thus RFC2119 language is not >appropriate. News readers that try to do threading care very much if it is missing. -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3KMvJ97078201 for ; Wed, 20 Apr 2005 15:57:19 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3KMvJIQ078200 for ietf-usefor-skb; Wed, 20 Apr 2005 15:57:19 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from probity.mcc.ac.uk (probity.mcc.ac.uk [130.88.200.94]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3KMvINk078191 for ; Wed, 20 Apr 2005 15:57:18 -0700 (PDT) (envelope-from chl@clerew.man.ac.uk) Received: from dialup027.mcc.ac.uk ([130.88.69.27]) by probity.mcc.ac.uk with esmtp (Exim 4.20) id 1DOO8D-000CWp-FM for ietf-usefor@imc.org; Wed, 20 Apr 2005 23:57:17 +0100 Received: (from chl@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3KMWsv15334; Wed, 20 Apr 2005 23:32:54 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20686 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: Fixed (was: Broken Message-ID syntax) Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: <4255153E.3080500@isode.com> <42594879.3B92@xyzzy.claranet.de> <425A8A63.268A@xyzzy.claranet.de> <425C607E.1B5A@xyzzy.claranet.de> <425D8031.D57@xyzzy.claranet.de> Date: Wed, 20 Apr 2005 19:42:16 GMT Lines: 38 MIME-Version: 1.0 X-Scanner: exiscan for exim4 (http://duncanthrax.net/exiscan/) *1DOO8D-000CWp-FM*QJF8NzufRBI* Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In <425D8031.D57@xyzzy.claranet.de> Frank Ellermann writes: >Charles Lindsey wrote: >> Yes, you have convinced me now. >Great, we have a msg-id syntax. >> there remains the issue of the naming of the syntax rules, >We're talking about seven names. So far we agree on msg-id, >and we disagree on names for LHS, RHS, and address-literal. >Maybe we can solve it for the remaining three names without >bothering Alexey, Henry, or Ken: Essentially, we either use the same names as RFC 2822 uses, saying that "the following syntax rules replace the corresponding rules in RFC 2822" (giving us id-left, id-right, no-fold-quote, no-fold literal, etc), or else we depart from the RFC 2822 names entirely, except for the msg-id at the top, in which case your names are as good as any. I don't see that there is anything in-between. My preference is to retain the RFC 2822 naming, which accords with the general principle that we do not change anything in RFC 2822 that we do not need to change. If our Chair says that we can change these names, then let it be so. Alexey? -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3KMvI8Z078193 for ; Wed, 20 Apr 2005 15:57:18 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3KMvITF078192 for ietf-usefor-skb; Wed, 20 Apr 2005 15:57:18 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from probity.mcc.ac.uk (probity.mcc.ac.uk [130.88.200.94]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3KMvHNl078183 for ; Wed, 20 Apr 2005 15:57:17 -0700 (PDT) (envelope-from chl@clerew.man.ac.uk) Received: from dialup027.mcc.ac.uk ([130.88.69.27]) by probity.mcc.ac.uk with esmtp (Exim 4.20) id 1DOO8C-000CWp-FD for ietf-usefor@imc.org; Wed, 20 Apr 2005 23:57:16 +0100 Received: (from chl@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3KMXf015340; Wed, 20 Apr 2005 23:33:41 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20687 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: Fwd: http://www.ietf.org/internet-drafts/draft-ietf-usefor-usefor-03.txt Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: <200504111212.j3BCChM14516@clerew.man.ac.uk> <87psx1gnjc.fsf@windlord.stanford.edu> <87fyxvx1m6.fsf@windlord.stanford.edu> <87mzs2a1hu.fsf@windlord.stanford.edu> Date: Wed, 20 Apr 2005 19:45:12 GMT Lines: 32 MIME-Version: 1.0 X-Scanner: exiscan for exim4 (http://duncanthrax.net/exiscan/) *1DOO8C-000CWp-FD*4wwlrdD5M7M* Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In <87mzs2a1hu.fsf@windlord.stanford.edu> Russ Allbery writes: >Charles Lindsey writes: >> Russ Allbery writes: >>> I think I see what you're saying -- it's not obsolete in the sense that >>> Path is obsolete (where news servers actually reject messages containing >>> the header). I'm fine with the wording; I guess I don't really care >>> whether that means "obsolete" or "obsolescent" as long as the key point is >>> expressed: don't use the header and don't bother creating it. >> Eh? Since when has Path been obsolete. Are you confusing it with some >> other ancient header? >Er, sorry, not Path, Received. Bleh. Ah! But in that case I don't think servers should be rejecting such headers. If the article has been wandering in any out of both email and news systems, then it is sometimes useful to have the full history recorded (though I might agree that if there was a mailing list involved I might not want to know what happened before it reached the mailing list). -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3KMvHYw078182 for ; Wed, 20 Apr 2005 15:57:17 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3KMvHcP078181 for ietf-usefor-skb; Wed, 20 Apr 2005 15:57:17 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from probity.mcc.ac.uk (probity.mcc.ac.uk [130.88.200.94]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3KMvGsY078161 for ; Wed, 20 Apr 2005 15:57:17 -0700 (PDT) (envelope-from chl@clerew.man.ac.uk) Received: from dialup027.mcc.ac.uk ([130.88.69.27]) by probity.mcc.ac.uk with esmtp (Exim 4.20) id 1DOO8B-000CWp-KZ for ietf-usefor@imc.org; Wed, 20 Apr 2005 23:57:15 +0100 Received: (from chl@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3KMYVY15348; Wed, 20 Apr 2005 23:34:31 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20690 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: Differences between RFC 2822 and Usefor Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: <425BF0A3.3020708@isode.com> <4262CF9B.40207@isode.com> Date: Wed, 20 Apr 2005 20:12:34 GMT Lines: 41 MIME-Version: 1.0 X-Scanner: exiscan for exim4 (http://duncanthrax.net/exiscan/) *1DOO8B-000CWp-KZ*E4nNcJZLgk2* Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In <4262CF9B.40207@isode.com> Alexey Melnikov writes: >Charles Lindsey wrote: >>And why does the Usefor draft contain the words >> >> None of the headers appearing in this section is required to appear >> in every article but some of them are required in certain types of >> article, such as followups. >> >Which is also followed by the following sentence: > Further discussion of these requirements appears in [USEPRO] >and [USEAGE]. Which sentence should not be there IMHO. >This is a clear indication that USEPRO is a proper place to define any >additional requirements. >You can add it to USEPRO to "Duties of a Followup Agent" as per above. Yes, that is possible. Indeed it is already there, but might need a little more saying about it there, including pointing out the divergence from RFC 2822 and something about using it in non-followup situations such as mulipart FAQs and message/partial. Incidentally, Usefor also needs the semantics of this header explaining (such as the significance attaching to the order in which the precursors are listed), so maybe the mention of multipart FAQs and message/partial could be slipped in there. -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3KMvHnL078173 for ; Wed, 20 Apr 2005 15:57:17 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3KMvHhi078172 for ietf-usefor-skb; Wed, 20 Apr 2005 15:57:17 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from probity.mcc.ac.uk (probity.mcc.ac.uk [130.88.200.94]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3KMvG6W078160 for ; Wed, 20 Apr 2005 15:57:16 -0700 (PDT) (envelope-from chl@clerew.man.ac.uk) Received: from dialup027.mcc.ac.uk ([130.88.69.27]) by probity.mcc.ac.uk with esmtp (Exim 4.20) id 1DOO8A-000CWp-UH for ietf-usefor@imc.org; Wed, 20 Apr 2005 23:57:15 +0100 Received: (from chl@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3KMZAq15354; Wed, 20 Apr 2005 23:35:10 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20689 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: Differences between RFC 2822 and Usefor Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: Date: Wed, 20 Apr 2005 20:06:14 GMT Lines: 30 MIME-Version: 1.0 X-Scanner: exiscan for exim4 (http://duncanthrax.net/exiscan/) *1DOO8A-000CWp-UH*Hqf6Bbhi9qw* Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In John Stanley writes: >The side effect is now that THIS requirement is gone, there is no way of >identifying what is and is not a followup no matter WHAT definition of >followup may be used, so there is no reason to argue about what is and is >not a followup anymore. If you think that not being able to identify what >is a followup isn't important (i.e., it's ok for non-followups to contain >References headers) then you cannot possibly think it is important enough >to identify followups that the header designed for that purpose has to be >mandatory. Be consistent -- either it's important or it isn't. No, the requirement has not gone. >Don't think that the mandate can be slipped back into USEPRO, since any >mandate in USEPRO would contradict USEFOR, and that would look stupid. If >USEFOR says it's optional, then USEPRO cannot claim that it is an >interoperability issue and is thus not optional. The mandate in question is already in USEPRO. And always has been. -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3IA1MCd078552 for ; Mon, 18 Apr 2005 03:01:22 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3IA1Mf7078551 for ietf-usefor-skb; Mon, 18 Apr 2005 03:01:22 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from rufus.isode.com (rufus.isode.com [62.3.217.251]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3IA1JCi078517 for ; Mon, 18 Apr 2005 03:01:20 -0700 (PDT) (envelope-from alexey.melnikov-usefor@isode.com) Received: from [192.168.0.3] ([62.3.217.253]) by rufus.isode.com via TCP (internal) with ESMTPA; Mon, 18 Apr 2005 11:01:00 +0100 Message-ID: <4262CF9B.40207@isode.com> Date: Sun, 17 Apr 2005 22:05:31 +0100 From: Alexey Melnikov User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.7.2) Gecko/20040804 Netscape/7.2 (ax) X-Accept-Language: en-us, en To: Charles Lindsey CC: ietf-usefor@imc.org Subject: Re: Differences between RFC 2822 and Usefor References: <425BF0A3.3020708@isode.com> In-Reply-To: MIME-version: 1.0 Content-type: text/plain; charset="ISO-8859-1"; format="flowed" Content-transfer-encoding: 7bit Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Charles Lindsey wrote: >In <425BF0A3.3020708@isode.com> Alexey Melnikov writes: > >>Charles Lindsey wrote: >> >> >>>The text regarding the References header in the new usefor-03 draft is >>>just plain WRONG. >>> >>> >>>But the requirement for >>>the header to be present for followups has been omitted, and it needs to >>>be put back (modulo the relatively minor matter of the precise wording). >>> >>> >>No it is not, because this is not a USEFOR business to say anything >>about followups. Any given article doesn't become valid/invalid because >>of presence or lack of References. USEFOR doesn't deal with a thread of >>messages, this is a protocol issue. >> >> > >Then why is RFC 2822 allowed to say that the References header SHOULD be >present in the corresponding email situation? > Because RFC 2822 is very much like a combined USEFOR+USEPRO. All I am trying to say is that USEFOR is not the proper place to state any requirement that doesn't apply to all articles. >And why does the word "followup" occur 10 times in the Usefor draft? > 6 of them are used when defining different terms, 3 are used when describing Followup-to (naturally), 1 is used in the sentence you've quoted below. IMHO, this is quite reasonable. I would like to avoid any use of the term in USEFOR, but this is just not practical. >And why does the Usefor draft contain the words > > None of the headers appearing in this section is required to appear > in every article but some of them are required in certain types of > article, such as followups. > Which is also followed by the following sentence: Further discussion of these requirements appears in [USEPRO] and [USEAGE]. This is a clear indication that USEPRO is a proper place to define any additional requirements. >And it says > > The References header is the same as that specified in Section 3.6.4 > of [RFC2822] with the added restrictions detailed in Section 2.2 and > those listed below: > >Which clearly implies that we have signed up to everything in RFC 2822 >Section 3.6.4 except as detailed otherwise, which means that we have >signed up to that "SHOULD", which we haven't. > > You can add it to USEPRO to "Duties of a Followup Agent" as per above. >Netnews departs from RFC 2822 in various ways, which we document. Changing >that "SHOULD" to a "MUST" is one of them (the identical SHOULD->MUST for >Message-ID is another, but I do not see any objection to saying that). >I understand that we intend to document these differences in an Appendix >to Usefor, but how can we document differences that we are not allowed to >describe? > > Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3I44L7o039856 for ; Sun, 17 Apr 2005 21:04:21 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3I44KKC039855 for ietf-usefor-skb; Sun, 17 Apr 2005 21:04:20 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from spsystems.net (spsystems.net [216.126.83.115]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3I44K0A039848 for ; Sun, 17 Apr 2005 21:04:20 -0700 (PDT) (envelope-from henry@spsystems.net) Received: from spsystems.net (henry@localhost [127.0.0.1]) by spsystems.net (8.12.10/8.12.10) with ESMTP id j3I43vVO001814; Mon, 18 Apr 2005 00:03:57 -0400 (EDT) Received: (from henry@localhost) by spsystems.net (8.12.10/8.12.10/Submit) id j3I43pU4001813; Mon, 18 Apr 2005 00:03:51 -0400 (EDT) Date: Mon, 18 Apr 2005 00:03:51 -0400 (EDT) From: Henry Spencer To: Usefor Mailing List Subject: Re: Path header delimiters In-Reply-To: <425AD6EF.448A@xyzzy.claranet.de> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: On Mon, 11 Apr 2005, Frank Ellermann wrote: > More fun: Any news about publishing s-o-1036 as RfC 4036 ? Despite good intentions, I've been distracted and haven't gotten to the point of making a submission. Early next month, I hope. (No chance between now and then, too much to do.) > Less fun: Charles and I need a volunteer for some kind of > "rough consensus" about the names (sic!) of some msg-id ABNF > productions, could you please toss a coin or add a comment ? I've been hoping that stuff would sort itself out, because I haven't been paying enough attention to have an opinion on it. Henry Spencer henry@spsystems.net Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3DLdldE097279 for ; Wed, 13 Apr 2005 14:39:47 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3DLdliX097278 for ietf-usefor-skb; Wed, 13 Apr 2005 14:39:47 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from b.mail.peak.org (b.mail.peak.org [69.59.192.42]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3DLdkD5097271 for ; Wed, 13 Apr 2005 14:39:46 -0700 (PDT) (envelope-from stanley@peak.org) Received: from a.shell.peak.org ([69.59.192.81]) by b.mail.peak.org (8.12.10/8.12.8) with ESMTP id j3DLde0j001422 (version=TLSv1/SSLv3 cipher=EDH-RSA-DES-CBC3-SHA bits=168 verify=NO) for ; Wed, 13 Apr 2005 14:39:41 -0700 (PDT) Date: Wed, 13 Apr 2005 14:39:40 -0700 (PDT) From: John Stanley X-X-Sender: stanley@a.shell.peak.org To: ietf-usefor@imc.org Subject: Re: Differences between RFC 2822 and Usefor Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Spam-Score: 0 () X-Scanned-By: MIMEDefang 2.39 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: "Charles Lindsey" : >And why does the Usefor draft contain the words > None of the headers appearing in this section is required to appear > in every article but some of them are required in certain types of > article, such as followups. Because someone (an editor, perhaps?) overlooked this statement when removing the mandate for References? Remove the "such as" clause and the problem is fixed. Do it and lets move on. Problem solved. >Which clearly implies that we have signed up to everything in RFC 2822 >Section 3.6.4 except as detailed otherwise, which means that we have >signed up to that "SHOULD", which we haven't. Clearly the text in usefor-03 says we HAVE signed up for RFC2822, since it says that. "as defined in ... with these limited exceptions" means "as defined in" with a few exceptions, none of which is a mandate for use in any kind of article, only mandates about what it contains when it is used. >Netnews departs from RFC 2822 in various ways, which we document. Yep. And one of those ways is no longer a mandatory References header in this nebulous "followup" thing. Undocumented "MUSTS" are not MUSTS. "Charles says..." is not RFC2119 justification. The side effect is now that THIS requirement is gone, there is no way of identifying what is and is not a followup no matter WHAT definition of followup may be used, so there is no reason to argue about what is and is not a followup anymore. If you think that not being able to identify what is a followup isn't important (i.e., it's ok for non-followups to contain References headers) then you cannot possibly think it is important enough to identify followups that the header designed for that purpose has to be mandatory. Be consistent -- either it's important or it isn't. Don't think that the mandate can be slipped back into USEPRO, since any mandate in USEPRO would contradict USEFOR, and that would look stupid. If USEFOR says it's optional, then USEPRO cannot claim that it is an interoperability issue and is thus not optional. Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3DLM9la096060 for ; Wed, 13 Apr 2005 14:22:09 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3DLM9vi096059 for ietf-usefor-skb; Wed, 13 Apr 2005 14:22:09 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from a.mail.peak.org (a.mail.peak.org [69.59.192.41]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3DLM8NS096047 for ; Wed, 13 Apr 2005 14:22:09 -0700 (PDT) (envelope-from stanley@peak.org) Received: from a.shell.peak.org ([69.59.192.81]) by a.mail.peak.org (8.12.10/8.12.8) with ESMTP id j3DLM2mO051869 (version=TLSv1/SSLv3 cipher=EDH-RSA-DES-CBC3-SHA bits=168 verify=NO) for ; Wed, 13 Apr 2005 14:22:03 -0700 (PDT) Date: Wed, 13 Apr 2005 14:22:02 -0700 (PDT) From: John Stanley X-X-Sender: stanley@a.shell.peak.org To: ietf-usefor@imc.org Subject: Re: Differences between RFC 2822 and Usefor (fwd) Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Spam-Score: 0 () X-Scanned-By: MIMEDefang 2.39 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Third attempt at one message. This "open list" does not seem to be accepting messages ... "Charles Lindsey" : >The "SHOULD" in question is in the first paragraph of section 3.6.4 of RFC >2822. I know where the SHOULD comes from. RFC2822 isn't "in question", it is an existing standard. The question is where you see a MUST in USEFOR. I don't see one. It simply isn't there. >The "MUST" in question arises from the long held consensus in this Working >Group that the References header in News is NOT an "optional extra". Not according to draft-usefor-03. According to draft-usefor-03, References is defined to be the same as in RFC2822, with the addition of a few bits that are not mandates for using it. This consensus you claim exists has a very odd way of being expressed in our products. Almost as if it doesn't really exist. The fact that I am the only one who even commented on it being gone is a pretty good sign that nobody cares that it is gone. >And you yourself have been at the forefront in upholding that consensus. Hardly. I've repeatedly said "IF we are going to say that References qualifies for RFC2119 mandates, THEN we should not water it down by saying that non-followups MAY contain them." That first clause is called a "conditional". If we DON'T say that References qualifies for such a mandate (and I've argued that it is hard to justify one, if you recall), then the situation is different. Like it is now. We currently do NOT say it is mandatory. Once we say that, the rest of the argument about WHEN it is mandatory goes away. >That clear requirement has been in all our drafts up to article-13. It is no longer a requirement at all in our draft. As you have been quick to tell me when I object to changes YOU make, things change. Our current draft does NOT make it a requirement; it is lunacy to claim that this lack of mandate reflects a "consensus" that there is a mandate, or that the complete lack of comment about the change (other than your expressed belief that a missing mandate is still a mandate) is a better sign of consensus than your word for it. At this point, the mandate is not part of the draft, and putting one in would (or at least, should, in an honest system) require a strong demonstration of interoperability issues. Since news systems simply do not care if the header is in the article or not, there is no way to demonstrate an interoperability issue, and thus RFC2119 language is not appropriate. I've accepted the change, why is the editor whose name is on it as author trying to claim it doesn't exist? >The text regarding the References header in the new usefor-03 draft is >just plain WRONG. Says you. If it was wrong, it should not have been sent off as our work product without a single person here seeing it. Well, ok, maybe someone here did see it, but it sure wasn't announced for review prior to it becoming official. >But the requirement for >the header to be present for followups has been omitted, and it needs to >be put back (modulo the relatively minor matter of the precise wording). You need to PROVE that NEED before you make a change to the draft. You saying "we need this" doesn't meet the requirements of RFC2119 and you know it. Were that all it takes, then we would be documenting PAM and MCT and making THEM mandatory. But, no. It doesn't work that way. You're taking all mention of those headers OUT of the drafts that still refer to them. >And there is also an error in the syntax of the References header, which >Frank Ellermann pointed out some while back, and which needs to be >corrected. Please stop trying to equate a syntax error with a change you want that has nothing to do with syntax. Making References optional (in accordance with RFC2822) is a different kind of change, and it's been made. Making it MANDATORY needs strong justification, and I'll await your attempts at same. "Because it was that way before" doesn't work when I object to a change you decide to make, don't pretend that you can use it as an excuse. Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3DKYM4S089463 for ; Wed, 13 Apr 2005 13:34:22 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3DKYMFk089462 for ietf-usefor-skb; Wed, 13 Apr 2005 13:34:22 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from ciao.gmane.org (main.gmane.org [80.91.229.2]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3DKYKOs089453 for ; Wed, 13 Apr 2005 13:34:20 -0700 (PDT) (envelope-from usenet-format@gmane.org) Received: from list by ciao.gmane.org with local (Exim 4.43) id 1DLoUS-00012v-2F for ietf-usefor@imc.org; Wed, 13 Apr 2005 22:29:36 +0200 Received: from 212.82.251.217 ([212.82.251.217]) by main.gmane.org with esmtp (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Wed, 13 Apr 2005 22:29:36 +0200 Received: from nobody by 212.82.251.217 with local (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Wed, 13 Apr 2005 22:29:36 +0200 X-Injected-Via-Gmane: http://gmane.org/ To: ietf-usefor@imc.org From: Frank Ellermann Subject: Fixed (was: Broken Message-ID syntax) Date: Wed, 13 Apr 2005 22:25:21 +0200 Organization: Lines: 96 Message-ID: <425D8031.D57@xyzzy.claranet.de> References: <4255153E.3080500@isode.com> <42594879.3B92@xyzzy.claranet.de> <425A8A63.268A@xyzzy.claranet.de> <425C607E.1B5A@xyzzy.claranet.de> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Complaints-To: usenet@sea.gmane.org X-Gmane-NNTP-Posting-Host: 212.82.251.217 X-Mailer: Mozilla 3.0 (OS/2; U) Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Charles Lindsey wrote: > Yes, you have convinced me now. Great, we have a msg-id syntax. > there remains the issue of the naming of the syntax rules, We're talking about seven names. So far we agree on msg-id, and we disagree on names for LHS, RHS, and address-literal. Maybe we can solve it for the remaining three names without bothering Alexey, Henry, or Ken: 1 - "unique-quote", the LHS if it must be quoted. If you hate "unique" a very traditional name would be "local". For a post-modern touch we could take "id". Maybe "quote" was too sloppy and "quoted" is better. Six possibilities: unique-quote local-quote id-quote unique-quoted local-quoted id-quoted 2 - "unique-part", the syntactically irrelevant stuff for the question "to quote or not to quote". Obviously we'd want the same prefix as in (1). Maybe "text" is better if we don't need it in (3): unique-part local-part id-part unique-text local-text id-text 3 - "unique-literal", something that requires quoting. The "literal" is somewhat dubious, 2822 and usefor-03 use it for the RHS, but here we are in the LHS Let's say that "literal" was a bad idea. Depending on (2) "text" could be fine. I also like "special" (as in your "mqspecial"): unique-text local-text id-text unique-special local-special id-special JFTR the three names where we are unable to find a compromise: 4 - "address-literal" vs. "no-fold-literal". You want the same name as in 2822 with a different syntax. For this reason I want a different name reflecting the semantics. We've established that the semantics is a "domain-literal" in RfC 2822, but that's already used for an 2822-construct with CFWS and FWS. A related source is RfC 2821, there the name is "address-literal". IMHO a typical case of "if it quacks like a duck...". 5 - "mdomain" vs. "id-right". As in (4), you want the name as in RfC 2822 with a different syntax. I want a name that reflects the semantics "domain". We've spent hours, days, and weeks with the prolem of the RHS in a Message-ID, because you said that it's no domain. I've shown that that's _not_ what 2822 means. AFAIK Bruce also disagrees with you. STD 11, RfC 1036, and s-o-1036 disagree with you. Abusing foreign namespaces is a serious problem. If you don't like "mdomain" you could pick "msg-domain" or "id-domain". But "id-right" is OUT as far as I'm concerned. 6 - "unique" vs. "id-left". A similar case as in (5). Maybe you feel that "unique" is too traditional, and "local" is even more traditional. OTOH it's true, so we could only beautify it a bit: msg-unique id-unique msg-local id-local We can use the prefix "id-" in (6), (5), (3), (2), and (1). Or the prefix "msg-" in (6), (5), we already have it in msg-id. > which I think needs a ruling from the Chair. Alexey as King Canute decreeing that the RHS is NOT a domain ? I don't think so. > the wording needs to explain this additional problem Not necessarily, leading / trailing / adjacent dots aren't new, it's exactly the same problem as in RfC 2822. > also the reason for exlcusion of NO-WS-CONTROL Maybe, but please state them all: It was never used, it never worked, it would be a PITA in news-URLs, it wasn't allowed in any news standard. More than one line of text, so that we have a line between these lines saying "2822 GOT IT WRONG - PERIOD." Bye, Frank Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3DK2sfk086901 for ; Wed, 13 Apr 2005 13:02:54 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3DK2s7b086900 for ietf-usefor-skb; Wed, 13 Apr 2005 13:02:54 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtp3.Stanford.EDU (smtp3.Stanford.EDU [171.67.16.138]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3DK2sig086894 for ; Wed, 13 Apr 2005 13:02:54 -0700 (PDT) (envelope-from rra@stanford.edu) Received: from windlord.stanford.edu (windlord.Stanford.EDU [171.64.19.147]) by smtp3.Stanford.EDU (8.12.11/8.12.11) with SMTP id j3DK2r0q031092 for ; Wed, 13 Apr 2005 13:02:53 -0700 Received: (qmail 13475 invoked by uid 1000); 13 Apr 2005 20:02:53 -0000 To: ietf-usefor@imc.org Subject: Re: Fwd: http://www.ietf.org/internet-drafts/draft-ietf-usefor-usefor-03.txt In-Reply-To: (Charles Lindsey's message of "Wed, 13 Apr 2005 15:27:14 GMT") References: <200504111212.j3BCChM14516@clerew.man.ac.uk> <87psx1gnjc.fsf@windlord.stanford.edu> <87fyxvx1m6.fsf@windlord.stanford.edu> From: Russ Allbery Organization: The Eyrie Date: Wed, 13 Apr 2005 13:02:53 -0700 Message-ID: <87mzs2a1hu.fsf@windlord.stanford.edu> User-Agent: Gnus/5.1007 (Gnus v5.10.7) XEmacs/21.4 (Jumbo Shrimp, linux) MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Charles Lindsey writes: > Russ Allbery writes: >> I think I see what you're saying -- it's not obsolete in the sense that >> Path is obsolete (where news servers actually reject messages containing >> the header). I'm fine with the wording; I guess I don't really care >> whether that means "obsolete" or "obsolescent" as long as the key point is >> expressed: don't use the header and don't bother creating it. > Eh? Since when has Path been obsolete. Are you confusing it with some > other ancient header? Er, sorry, not Path, Received. Bleh. -- Russ Allbery (rra@stanford.edu) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3DGCbfZ072659 for ; Wed, 13 Apr 2005 09:12:37 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3DGCbPu072658 for ietf-usefor-skb; Wed, 13 Apr 2005 09:12:37 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtp803.mail.ukl.yahoo.com (smtp803.mail.ukl.yahoo.com [217.12.12.140]) by above.proper.com (8.12.11/8.12.9) with SMTP id j3DGCaOX072623 for ; Wed, 13 Apr 2005 09:12:37 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from unknown (HELO host81-144-75-78.midband.mdip.bt.net) (ietf-usefor@imc.org@81.144.75.78 with poptime) by smtp803.mail.ukl.yahoo.com with SMTP; 13 Apr 2005 16:12:30 -0000 Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3DGCDE00006 for ietf-usefor@imc.org; Wed, 13 Apr 2005 17:12:13 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20676 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: Differences between RFC 2822 and Usefor Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: <425BF0A3.3020708@isode.com> Date: Wed, 13 Apr 2005 15:59:00 GMT Lines: 55 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In <425BF0A3.3020708@isode.com> Alexey Melnikov writes: >Charles Lindsey wrote: >>The text regarding the References header in the new usefor-03 draft is >>just plain WRONG. >>But the requirement for >>the header to be present for followups has been omitted, and it needs to >>be put back (modulo the relatively minor matter of the precise wording). >> >> >No it is not, because this is not a USEFOR business to say anything >about followups. Any given article doesn't become valid/invalid because >of presence or lack of References. USEFOR doesn't deal with a thread of >messages, this is a protocol issue. Then why is RFC 2822 allowed to say that the References header SHOULD be present in the corresponding email situation? And why does the word "followup" occur 10 times in the Usefor draft? And why does the Usefor draft contain the words None of the headers appearing in this section is required to appear in every article but some of them are required in certain types of article, such as followups. And it says The References header is the same as that specified in Section 3.6.4 of [RFC2822] with the added restrictions detailed in Section 2.2 and those listed below: Which clearly implies that we have signed up to everything in RFC 2822 Section 3.6.4 except as detailed otherwise, which means that we have signed up to that "SHOULD", which we haven't. Netnews departs from RFC 2822 in various ways, which we document. Changing that "SHOULD" to a "MUST" is one of them (the identical SHOULD->MUST for Message-ID is another, but I do not see any objection to saying that). I understand that we intend to document these differences in an Appendix to Usefor, but how can we document differences that we are not allowed to describe? -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3DGCaWA072649 for ; Wed, 13 Apr 2005 09:12:36 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3DGCaMI072648 for ietf-usefor-skb; Wed, 13 Apr 2005 09:12:36 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtp803.mail.ukl.yahoo.com (smtp803.mail.ukl.yahoo.com [217.12.12.140]) by above.proper.com (8.12.11/8.12.9) with SMTP id j3DGCZ2l072621 for ; Wed, 13 Apr 2005 09:12:35 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from unknown (HELO host81-144-75-78.midband.mdip.bt.net) (ietf-usefor@imc.org@81.144.75.78 with poptime) by smtp803.mail.ukl.yahoo.com with SMTP; 13 Apr 2005 16:12:29 -0000 Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3DGCBV29988 for ietf-usefor@imc.org; Wed, 13 Apr 2005 17:12:11 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20674 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: Path header delimiters Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: <87br8jx1ej.fsf@windlord.stanford.edu> Date: Wed, 13 Apr 2005 15:24:17 GMT Lines: 31 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In <87br8jx1ej.fsf@windlord.stanford.edu> Russ Allbery writes: >Charles Lindsey writes: >> Anyway, it a minor detail as regards scheme E, which seems to be gaining >> added consenus. Does anyone else want to play? >Giving semantic meaning to a doubled delimiter makes me nervous, just >because that's something one generally doesn't do in protocols of this >type. But that's just a vague, untargetted concern and I can't think of >concrete problems that it would solve. >It's the best solution I've seen so far, so I'm happy to go with it unless >someone comes up with something better. I do think that this is an >important problem for us to solve. OK, that makes it quite a smooth consensus compared to some we have had :-) . I shall write some text to incorporate it next week (no time at the moment - visiting grandchildren and a daughter getting married at the weekend). -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3DGCZwv072641 for ; Wed, 13 Apr 2005 09:12:35 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3DGCZ7T072640 for ietf-usefor-skb; Wed, 13 Apr 2005 09:12:35 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtp803.mail.ukl.yahoo.com (smtp803.mail.ukl.yahoo.com [217.12.12.140]) by above.proper.com (8.12.11/8.12.9) with SMTP id j3DGCYCa072616 for ; Wed, 13 Apr 2005 09:12:34 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from unknown (HELO host81-144-75-78.midband.mdip.bt.net) (ietf-usefor@imc.org@81.144.75.78 with poptime) by smtp803.mail.ukl.yahoo.com with SMTP; 13 Apr 2005 16:12:28 -0000 Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3DGCDV00010 for ietf-usefor@imc.org; Wed, 13 Apr 2005 17:12:13 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20677 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: Broken Message-ID syntax Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: <4255153E.3080500@isode.com> <42594879.3B92@xyzzy.claranet.de> <425A8A63.268A@xyzzy.claranet.de> <425C607E.1B5A@xyzzy.claranet.de> Date: Wed, 13 Apr 2005 16:04:57 GMT Lines: 34 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In <425C607E.1B5A@xyzzy.claranet.de> Frank Ellermann writes: >Module @ ~ unique-literal it can match in two different ways: >@ ~ unique-literal @. ~ unique-part >.@ ~ unique-part @ ~ unique-literal >| unique-part = 1*( atext / "." / unique-literal ) >Note the dot "." in unique-part, this should work as desiged. Yes, you have convinced me now. So there remains the issue of the naming of the syntax rules, which I think needs a ruling from the Chair. Also, the wording needs to explain this additional problem (and also the reason for exlcusion of NO-WS-CONTROL). At the very least, the three examples given should be changed to: <"ab.cd"@example.com> <"ab.\cd"@example.com> -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3DGCXNP072632 for ; Wed, 13 Apr 2005 09:12:33 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3DGCX7A072631 for ietf-usefor-skb; Wed, 13 Apr 2005 09:12:33 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtp803.mail.ukl.yahoo.com (smtp803.mail.ukl.yahoo.com [217.12.12.140]) by above.proper.com (8.12.11/8.12.9) with SMTP id j3DGCVux072613 for ; Wed, 13 Apr 2005 09:12:32 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from unknown (HELO host81-144-75-78.midband.mdip.bt.net) (ietf-usefor@imc.org@81.144.75.78 with poptime) by smtp803.mail.ukl.yahoo.com with SMTP; 13 Apr 2005 16:12:25 -0000 Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3DGCCp29996 for ietf-usefor@imc.org; Wed, 13 Apr 2005 17:12:12 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20675 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: Fwd: http://www.ietf.org/internet-drafts/draft-ietf-usefor-usefor-03.txt Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: <200504111212.j3BCChM14516@clerew.man.ac.uk> <87psx1gnjc.fsf@windlord.stanford.edu> <87fyxvx1m6.fsf@windlord.stanford.edu> Date: Wed, 13 Apr 2005 15:27:14 GMT Lines: 21 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In <87fyxvx1m6.fsf@windlord.stanford.edu> Russ Allbery writes: >I think I see what you're saying -- it's not obsolete in the sense that >Path is obsolete (where news servers actually reject messages containing >the header). I'm fine with the wording; I guess I don't really care >whether that means "obsolete" or "obsolescent" as long as the key point is >expressed: don't use the header and don't bother creating it. Eh? Since when has Path been obsolete. Are you confusing it with some other ancient header? -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3CNxlQ5066798 for ; Tue, 12 Apr 2005 16:59:47 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3CNxlYm066797 for ietf-usefor-skb; Tue, 12 Apr 2005 16:59:47 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from ciao.gmane.org (main.gmane.org [80.91.229.2]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3CNxiDr066783 for ; Tue, 12 Apr 2005 16:59:45 -0700 (PDT) (envelope-from usenet-format@gmane.org) Received: from list by ciao.gmane.org with local (Exim 4.43) id 1DLVEu-00086t-KU for ietf-usefor@imc.org; Wed, 13 Apr 2005 01:56:16 +0200 Received: from 212.82.251.36 ([212.82.251.36]) by main.gmane.org with esmtp (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Wed, 13 Apr 2005 01:56:16 +0200 Received: from nobody by 212.82.251.36 with local (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Wed, 13 Apr 2005 01:56:16 +0200 X-Injected-Via-Gmane: http://gmane.org/ To: ietf-usefor@imc.org From: Frank Ellermann Subject: Re: Broken Message-ID syntax Date: Wed, 13 Apr 2005 01:57:50 +0200 Organization: Lines: 47 Message-ID: <425C607E.1B5A@xyzzy.claranet.de> References: <4255153E.3080500@isode.com> <42594879.3B92@xyzzy.claranet.de> <425A8A63.268A@xyzzy.claranet.de> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Complaints-To: usenet@sea.gmane.org X-Gmane-NNTP-Posting-Host: 212.82.251.36 X-Mailer: Mozilla 3.0 (OS/2; U) Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Charles Lindsey wrote: > anything with an "@" in it (or various other weird > characters) clearly MUST be quoted, ACK > your syntax does not allow me to say <"@.@"@example.com> NAK | msg-id = "<" unique "@" mdomain ">" | unique = dot-atom-text / ( DQUOTE unique-quote DQUOTE ) Skipping some details for your msg-id <"@.@"@example.com> I get example.com ~ mdomain "@.@" ~ ( DQUOTE unique-quote DQUOTE ) Modulo DQUOTE that's @.@ ~ unique-quote | unique-quote = ( "." [unique-part] ) / | ( [unique-part] "." ) / | ( [unique-part] unique-literal [unique-part] ) @.@ doesn't start with a dot, it doesn't end with a dot, so let's try: @.@ ~ [unique-part] unique-literal [unique-part] Module @ ~ unique-literal it can match in two different ways: @ ~ unique-literal @. ~ unique-part .@ ~ unique-part @ ~ unique-literal | unique-part = 1*( atext / "." / unique-literal ) Note the dot "." in unique-part, this should work as desiged. > though it would allow <"@..@"@example.com> Any number of dots, zero, one, two, or more. > it will be even uglier after the "@.@" bug is fixed :-( . First I want to see a bug before we make it uglier than it is. It's already ambiguous, it can't get worse, unless it's plain wrong. Bye, Frank Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3CJLb6o048838 for ; Tue, 12 Apr 2005 12:21:37 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3CJLbw9048837 for ietf-usefor-skb; Tue, 12 Apr 2005 12:21:37 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from anchor-post-34.mail.demon.net (anchor-post-34.mail.demon.net [194.217.242.92]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3CJLakt048829 for ; Tue, 12 Apr 2005 12:21:36 -0700 (PDT) (envelope-from richard@highwayman.com) Received: from gti.noc.demon.net ([195.11.55.101] helo=happyday.al.cl.cam.ac.uk) by anchor-post-34.mail.demon.net with esmtp (Exim 4.42) id 1DLQx5-000MnJ-Cv; Tue, 12 Apr 2005 19:21:35 +0000 Message-ID: Date: Tue, 12 Apr 2005 20:20:10 +0100 To: Russ Allbery Cc: Dr John Stockton , ietf-usefor@imc.org From: Richard Clayton Subject: Re: Fwd: http://www.ietf.org/internet-drafts/draft-ietf-usefor-usefor-03.txt References: <200504111212.j3BCChM14516@clerew.man.ac.uk> <87psx1gnjc.fsf@windlord.stanford.edu> <4XACZ2FxV$WCFwUM@merlyn.demon.co.uk> <87oecjx1xm.fsf@windlord.stanford.edu> In-Reply-To: <87oecjx1xm.fsf@windlord.stanford.edu> MIME-Version: 1.0 X-Mailer: Turnpike Integrated Version 5.02 M Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 In message <87oecjx1xm.fsf@windlord.stanford.edu>, Russ Allbery writes > >John Stockton writes: > >> I use Turnpike. It has kill-rules acting on article headers. With a >> general rule using a RegExp one can kill or accept by a wide variety of >> criteria. But in the controlling dialogue of the version that I use >> there is specific provision for "Number of lines ... (as recorded in the >> header)" - enter a number, rather than compose a RegExp - evidently, an >> Internet Expert must have considered Lines to be useful information.. > >> I find this useful in newsgroups where a few users tend to post whole >> web pages, computer programs, or essays. > >Yes. I'm not disagreeing about the usefulness of that sort of killfile >rule. I'm saying that Turnpike should apply that rule against the >overview information, which is guaranteed to be correct. which it can already do :) .... assuming of course that its heuristics (and configuration) mean that it is indeed fetching the overview info and it isn't just using HEAD commands (that will always "work") ... in the real world, clients currently have to guess (or be told) what facilities to use; and tend to err in favour of features that are bound to be available... ... hence the value of NNTP v2 :) >If it's using the Lines header in the article, it's getting occasionally >bogus results right now. The standard is intended to tell Turnpike and >other news readers that do this to stop and use the overview information >instead. Declaring this to be a bug in Turnpike is fully intentional. John should add USEXOVER=YES to the appropriate [NEWS xxx] part of his configuration file :) and the "bug" will not bite him - -- richard @ highwayman . com "Nothing seems the same Still you never see the change from day to day And no-one notices the customs slip away" -----BEGIN PGP SIGNATURE----- Version: PGPsdk version 1.7.1 iQA/AwUBQlwfapoAxkTY1oPiEQJrRwCfW7Md/LVN74iV2bP1SbCfHaZmqs0An376 8uo4K/bZlJr1+0lljqx46KYU =smJq -----END PGP SIGNATURE----- Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3CJ464C047876 for ; Tue, 12 Apr 2005 12:04:06 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3CJ466b047875 for ietf-usefor-skb; Tue, 12 Apr 2005 12:04:06 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtp1.Stanford.EDU (smtp1.Stanford.EDU [171.67.16.123]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3CJ45vX047868 for ; Tue, 12 Apr 2005 12:04:05 -0700 (PDT) (envelope-from rra@stanford.edu) Received: from windlord.stanford.edu (windlord.Stanford.EDU [171.64.19.147]) by smtp1.Stanford.EDU (8.12.11/8.12.11) with SMTP id j3CJ45Lt011240 for ; Tue, 12 Apr 2005 12:04:05 -0700 Received: (qmail 23592 invoked by uid 1000); 12 Apr 2005 19:04:04 -0000 To: ietf-usefor@imc.org Subject: Re: Path header delimiters In-Reply-To: (Charles Lindsey's message of "Tue, 12 Apr 2005 11:36:53 GMT") References: From: Russ Allbery Organization: The Eyrie Date: Tue, 12 Apr 2005 12:04:04 -0700 Message-ID: <87br8jx1ej.fsf@windlord.stanford.edu> User-Agent: Gnus/5.1007 (Gnus v5.10.7) XEmacs/21.4 (Jumbo Shrimp, linux) MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Charles Lindsey writes: > Anyway, it a minor detail as regards scheme E, which seems to be gaining > added consenus. Does anyone else want to play? Giving semantic meaning to a doubled delimiter makes me nervous, just because that's something one generally doesn't do in protocols of this type. But that's just a vague, untargetted concern and I can't think of concrete problems that it would solve. It's the best solution I've seen so far, so I'm happy to go with it unless someone comes up with something better. I do think that this is an important problem for us to solve. -- Russ Allbery (rra@stanford.edu) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3CIxVr1047677 for ; Tue, 12 Apr 2005 11:59:31 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3CIxV1x047676 for ietf-usefor-skb; Tue, 12 Apr 2005 11:59:31 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtp1.Stanford.EDU (smtp1.Stanford.EDU [171.67.16.123]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3CIxUIp047669 for ; Tue, 12 Apr 2005 11:59:30 -0700 (PDT) (envelope-from rra@stanford.edu) Received: from windlord.stanford.edu (windlord.Stanford.EDU [171.64.19.147]) by smtp1.Stanford.EDU (8.12.11/8.12.11) with SMTP id j3CIxTo8009174 for ; Tue, 12 Apr 2005 11:59:30 -0700 Received: (qmail 22722 invoked by uid 1000); 12 Apr 2005 18:59:29 -0000 To: ietf-usefor@imc.org Subject: Re: Fwd: http://www.ietf.org/internet-drafts/draft-ietf-usefor-usefor-03.txt In-Reply-To: (Charles Lindsey's message of "Tue, 12 Apr 2005 11:23:40 GMT") References: <200504111212.j3BCChM14516@clerew.man.ac.uk> <87psx1gnjc.fsf@windlord.stanford.edu> From: Russ Allbery Organization: The Eyrie Date: Tue, 12 Apr 2005 11:59:29 -0700 Message-ID: <87fyxvx1m6.fsf@windlord.stanford.edu> User-Agent: Gnus/5.1007 (Gnus v5.10.7) XEmacs/21.4 (Jumbo Shrimp, linux) MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Charles Lindsey writes: > Russ Allbery writes: >> Kill Rules are based on overview information, not on article headers >> except in unusual situations (since killing on full headers is very >> slow). Overview contains its own line count that has nothing to do >> with an article header; this is specified by the NNTP protocol. > I think it is up to individual implementors of reading agents whether to > provide kill file facilities on all headers, or only on those in the > overview; some do, some don't. Except for Lines. Lines should always be applied against overview. > Indeed, but any killfile rule is likely to be expressed as "kill all > articles where Lines > 1000", in which case absolute accuracy of the > line count hardly matters. But I agree that there are better ways of > killing overly long articles. Like EXACTLY THAT KILLFILE RULE, but applied against the guaranteed-accurate overview information. > The wording actually used, which has not changed since our very early > drafts, was > This header is to be regarded as obsolete, and it will likely be > removed entirely in a future version of this standard. In the > meantime, its use is deprecated. > Which to me indicates "It is not obsolete yet, but it soon will be, so > it is not really a good idea to use it". I think we did it that way > because we wanted to document the correct way to do the count (because > implementations regularly did it wrong), but there was still a > widespread usage of it, which was likely to continue for some while. > To my mind, the proper word to describe that situation is "obsolescent", > rather than "obsolete". I think I see what you're saying -- it's not obsolete in the sense that Path is obsolete (where news servers actually reject messages containing the header). I'm fine with the wording; I guess I don't really care whether that means "obsolete" or "obsolescent" as long as the key point is expressed: don't use the header and don't bother creating it. -- Russ Allbery (rra@stanford.edu) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3CIqeQQ047274 for ; Tue, 12 Apr 2005 11:52:40 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3CIqefV047273 for ietf-usefor-skb; Tue, 12 Apr 2005 11:52:40 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtp3.Stanford.EDU (smtp3.Stanford.EDU [171.67.16.138]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3CIqd33047261 for ; Tue, 12 Apr 2005 11:52:39 -0700 (PDT) (envelope-from rra@stanford.edu) Received: from windlord.stanford.edu (windlord.Stanford.EDU [171.64.19.147]) by smtp3.Stanford.EDU (8.12.11/8.12.11) with SMTP id j3CIqc7b001663 for ; Tue, 12 Apr 2005 11:52:38 -0700 Received: (qmail 22549 invoked by uid 1000); 12 Apr 2005 18:52:38 -0000 To: Dr John Stockton Cc: ietf-usefor@imc.org Subject: Re: Fwd: http://www.ietf.org/internet-drafts/draft-ietf-usefor-usefor-03.txt In-Reply-To: <4XACZ2FxV$WCFwUM@merlyn.demon.co.uk> (John Stockton's message of "Tue, 12 Apr 2005 17:21:05 +0100") References: <200504111212.j3BCChM14516@clerew.man.ac.uk> <87psx1gnjc.fsf@windlord.stanford.edu> <4XACZ2FxV$WCFwUM@merlyn.demon.co.uk> From: Russ Allbery Organization: The Eyrie Date: Tue, 12 Apr 2005 11:52:37 -0700 Message-ID: <87oecjx1xm.fsf@windlord.stanford.edu> User-Agent: Gnus/5.1007 (Gnus v5.10.7) XEmacs/21.4 (Jumbo Shrimp, linux) MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: John Stockton writes: > I use Turnpike. It has kill-rules acting on article headers. With a > general rule using a RegExp one can kill or accept by a wide variety of > criteria. But in the controlling dialogue of the version that I use > there is specific provision for "Number of lines ... (as recorded in the > header)" - enter a number, rather than compose a RegExp - evidently, an > Internet Expert must have considered Lines to be useful information.. > I find this useful in newsgroups where a few users tend to post whole > web pages, computer programs, or essays. Yes. I'm not disagreeing about the usefulness of that sort of killfile rule. I'm saying that Turnpike should apply that rule against the overview information, which is guaranteed to be correct. You'll get exactly the same results, faster, and more reliably. If it's using the Lines header in the article, it's getting occasionally bogus results right now. The standard is intended to tell Turnpike and other news readers that do this to stop and use the overview information instead. Declaring this to be a bug in Turnpike is fully intentional. >> I believe that we did. We intended it to be declared obsolete in the >> sense that nothing should generate it and nothing should use it. > That will not meet with universal agreement. It generally does as soon as people understand what I mean when I say that a better version of that information is already in the overview data. -- Russ Allbery (rra@stanford.edu) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3CGxsKS037542 for ; Tue, 12 Apr 2005 09:59:54 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3CGxsQ2037541 for ietf-usefor-skb; Tue, 12 Apr 2005 09:59:54 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from mail1.panix.com (mail1.panix.com [166.84.1.72]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3CGxrQA037532 for ; Tue, 12 Apr 2005 09:59:54 -0700 (PDT) (envelope-from sethb@panix.com) Received: from panix5.panix.com (panix5.panix.com [166.84.1.5]) by mail1.panix.com (Postfix) with ESMTP id 936CA58B3B for ; Tue, 12 Apr 2005 12:59:52 -0400 (EDT) Received: (from sethb@localhost) by panix5.panix.com (8.11.6p3/8.8.8/PanixN1.1) id j3CGxq628085; Tue, 12 Apr 2005 12:59:52 -0400 (EDT) Date: Tue, 12 Apr 2005 12:59:52 -0400 (EDT) Message-Id: <200504121659.j3CGxq628085@panix5.panix.com> From: Seth Breidbart To: ietf-usefor@imc.org In-reply-to: (chl@clerew.man.ac.uk) Subject: Re: Path header delimiters References: Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: > Anyway, it a minor detail as regards scheme E, which seems to be > gaining added consenus. Does anyone else want to play? Sure. consensus++ Seth Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3CG0elD031831 for ; Tue, 12 Apr 2005 09:00:40 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3CG0edL031830 for ietf-usefor-skb; Tue, 12 Apr 2005 09:00:40 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from rufus.isode.com (rufus.isode.com [62.3.217.251]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3CG0cUH031822 for ; Tue, 12 Apr 2005 09:00:39 -0700 (PDT) (envelope-from alexey.melnikov-usefor@isode.com) Received: from [172.16.2.185] (shiny.isode.com [62.3.217.250]) by rufus.isode.com via TCP (internal) with ESMTPA; Tue, 12 Apr 2005 17:00:35 +0100 Message-ID: <425BF0A3.3020708@isode.com> Date: Tue, 12 Apr 2005 17:00:35 +0100 From: Alexey Melnikov User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.7.2) Gecko/20040804 Netscape/7.2 (ax) X-Accept-Language: en-us, en MIME-Version: 1.0 To: Charles Lindsey CC: ietf-usefor@imc.org Subject: Re: Differences between RFC 2822 and Usefor References: In-Reply-To: Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Charles Lindsey wrote: >>>3.2.1 References >>> >>> >>> The References header is the same as that specified in Section 3.6.4 >>> of [RFC2822] with the added restrictions detailed in Section 2.2 and >>> those listed below: >>> >>> >>Section 2.2 does not say anything about specific headers (and nothing at >>all about References), and the remaining part of 3.2.1 says nothing about >>followups and MUST. RFC2822 says "SHOULD". >> >> > >The text regarding the References header in the new usefor-03 draft is >just plain WRONG. > >Yes. the bit about section 2.2 covers the obligatory SP after the ':', as >required in all headers, and the "listed below" refers to the avoidance of >comments (MUST accept but do not generate yet). > >But the requirement for >the header to be present for followups has been omitted, and it needs to >be put back (modulo the relatively minor matter of the precise wording). > > No it is not, because this is not a USEFOR business to say anything about followups. Any given article doesn't become valid/invalid because of presence or lack of References. USEFOR doesn't deal with a thread of messages, this is a protocol issue. >And there is also an error in the syntax of the References header, which >Frank Ellermann pointed out some while back, and which needs to be >corrected. > This should be fixed or at least discussed. Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3CC1MUb095262 for ; Tue, 12 Apr 2005 05:01:22 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3CC1MUZ095261 for ietf-usefor-skb; Tue, 12 Apr 2005 05:01:22 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtp801.mail.ukl.yahoo.com (smtp801.mail.ukl.yahoo.com [217.12.12.138]) by above.proper.com (8.12.11/8.12.9) with SMTP id j3CC1FQI095153 for ; Tue, 12 Apr 2005 05:01:17 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from unknown (HELO host81-144-77-103.midband.mdip.bt.net) (ietf-usefor@imc.org@81.144.77.103 with poptime) by smtp801.mail.ukl.yahoo.com with SMTP; 12 Apr 2005 12:01:09 -0000 Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3CC0sr22211 for ietf-usefor@imc.org; Tue, 12 Apr 2005 13:00:54 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20664 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: Broken Message-ID syntax Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: <4255153E.3080500@isode.com> <42594879.3B92@xyzzy.claranet.de> <425A8A63.268A@xyzzy.claranet.de> Date: Tue, 12 Apr 2005 11:31:32 GMT Lines: 41 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In <425A8A63.268A@xyzzy.claranet.de> Frank Ellermann writes: >Charles Lindsey wrote: >> It prevents the msg-id >> <"foo.bar.baz"@example.com> >> as required >Yes, that's the idea: >Leading dot or trailing dot or unique-literal <=> quote >No leading/trailing dot and no unique-literal <=> don't quote >> it also prevents >> <"@.@"@example.com> >"@" is a "unique literal" => quote. Working as designed. >"@" is no atext => no dot-atom-text => unquoted is illegal. >Where's the problem ? The problem is that anything with an "@" in it (or various other weird characters) clearly MUST be quoted, but your syntax does not allow me to say <"@.@"@example.com> (though it would allow <"@..@"@example.com>). >In the proposed "unique" ABNF adjacent dots are handled by >".." in "unique-literal". Ugly like hell, but who cares if >it's correct. And I suspect it will be even uglier after the "@.@" bug is fixed :-( . But let us get a working correct syntax first, and then we can see whether there is a neater way to express it. -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3CC1Jtq095236 for ; Tue, 12 Apr 2005 05:01:19 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3CC1JS4095235 for ietf-usefor-skb; Tue, 12 Apr 2005 05:01:19 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtp801.mail.ukl.yahoo.com (smtp801.mail.ukl.yahoo.com [217.12.12.138]) by above.proper.com (8.12.11/8.12.9) with SMTP id j3CC1Isa095179 for ; Tue, 12 Apr 2005 05:01:18 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from unknown (HELO host81-144-77-103.midband.mdip.bt.net) (ietf-usefor@imc.org@81.144.77.103 with poptime) by smtp801.mail.ukl.yahoo.com with SMTP; 12 Apr 2005 12:01:11 -0000 Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3CC0rq22205 for ietf-usefor@imc.org; Tue, 12 Apr 2005 13:00:53 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20663 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: Fwd: http://www.ietf.org/internet-drafts/draft-ietf-usefor-usefor-03.txt Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: <200504111212.j3BCChM14516@clerew.man.ac.uk> <87psx1gnjc.fsf@windlord.stanford.edu> Date: Tue, 12 Apr 2005 11:23:40 GMT Lines: 61 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In <87psx1gnjc.fsf@windlord.stanford.edu> Russ Allbery writes: >Charles Lindsey writes: >> "Dr John Stockton" writes: >>> "3.3.1 Lines" >>> This may not be needed for the transmission of News; but it is very >>> useful for Kill Rules. Therefore, IMHO it should be compulsory, not >>> obsolete (unless replaced by, say, 'Bytes'). >Kill Rules are based on overview information, not on article headers >except in unusual situations (since killing on full headers is very slow). >Overview contains its own line count that has nothing to do with an >article header; this is specified by the NNTP protocol. I think it is up to individual implementors of reading agents whether to provide kill file facilities on all headers, or only on those in the overview; some do, some don't. >You will find, if you analyze articles in the wild, that a significant >percentage of Lines headers are wrong. Indeed, but any killfile rule is likely to be expressed as "kill all articles where Lines > 1000", in which case absolute accuracy of the line count hardly matters. But I agree that there are better ways of killing overly long articles. >> I think the WG agreed early on that this header had passed its >> sell-by-date, though I do not think we intended it to be declared >> "obsolete" as the current Usefor draft implies. >I believe that we did. We intended it to be declared obsolete in the >sense that nothing should generate it and nothing should use it. The wording actually used, which has not changed since our very early drafts, was This header is to be regarded as obsolete, and it will likely be removed entirely in a future version of this standard. In the meantime, its use is deprecated. Which to me indicates "It is not obsolete yet, but it soon will be, so it is not really a good idea to use it". I think we did it that way because we wanted to document the correct way to do the count (because implementations regularly did it wrong), but there was still a widespread usage of it, which was likely to continue for some while. To my mind, the proper word to describe that situation is "obsolescent", rather than "obsolete". -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3CC1IF8095225 for ; Tue, 12 Apr 2005 05:01:18 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3CC1IoO095223 for ietf-usefor-skb; Tue, 12 Apr 2005 05:01:18 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtp801.mail.ukl.yahoo.com (smtp801.mail.ukl.yahoo.com [217.12.12.138]) by above.proper.com (8.12.11/8.12.9) with SMTP id j3CC1Ee3095146 for ; Tue, 12 Apr 2005 05:01:17 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from unknown (HELO host81-144-77-103.midband.mdip.bt.net) (ietf-usefor@imc.org@81.144.77.103 with poptime) by smtp801.mail.ukl.yahoo.com with SMTP; 12 Apr 2005 12:01:08 -0000 Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3CC0tQ22216 for ietf-usefor@imc.org; Tue, 12 Apr 2005 13:00:55 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20665 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: Path header delimiters Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: Date: Tue, 12 Apr 2005 11:36:53 GMT Lines: 32 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In Henry Spencer writes: >On Mon, 11 Apr 2005, Frank Ellermann wrote: >> > I would suggest would be to use "AKA" -- Also Known As -- >> > instead of MISMATCH >> >> Two problems: MISMATCH is already "common practice"... >As a standalone item in the path content? I think not. There is some >vaguely-related existing practice of appending it to server names, which >might be considered a minor point in its favor. You are damning it with faint praise :-) . Actually, I think I prefer MISMATCH, not only because it is close to some existing practice, but also because it is a conspicuous word that will shout out at you "there is something fishy about this Path" whenever you see it. Anyway, it a minor detail as regards scheme E, which seems to be gaining added consenus. Does anyone else want to play? -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3CBCfS1078142 for ; Tue, 12 Apr 2005 04:12:41 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3CBCfp2078141 for ietf-usefor-skb; Tue, 12 Apr 2005 04:12:41 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtp811.mail.ukl.yahoo.com (smtp811.mail.ukl.yahoo.com [217.12.12.201]) by above.proper.com (8.12.11/8.12.9) with SMTP id j3CBCceg078094 for ; Tue, 12 Apr 2005 04:12:40 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from unknown (HELO host81-144-77-234.midband.mdip.bt.net) (ietf-usefor@imc.org@81.144.77.234 with poptime) by smtp811.mail.ukl.yahoo.com with SMTP; 12 Apr 2005 11:12:31 -0000 Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3CBCFj21322 for ietf-usefor@imc.org; Tue, 12 Apr 2005 12:12:15 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20662 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: Differences between RFC 2822 and Usefor Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: Date: Tue, 12 Apr 2005 11:08:04 GMT Lines: 62 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In John Stanley writes: >"Charles Lindsey" : >> Differences between RFC 2822 and Usefor. >> ---------------------------------------- >>2.11 A References header MUST be provided for followups (as opposed to >>SHOULD be provided for replies in RFC 2822). >Would you care to explain where you are getting this difference? In >draft-ietf-usefor-usefor-03.txt, the References header is defined thusly: The "SHOULD" in question is in the first paragraph of section 3.6.4 of RFC 2822. In consequence of that SHOULD, many MUAs still regard providing a References header as an "optional extra", with the consequence that attempts to display threading in mailing lists (including this one) usually succeed only in showing broken threads. The "MUST" in question arises from the long held consensus in this Working Group that the References header in News is NOT an "optional extra". And you yourself have been at the forefront in upholding that consensus. There is a disagreement as to how exactly that requirement should be expressed, and as to how the wording should reflect the use of the References header in the case of multipart FAQs and the like, but there is no disagreement at all that it MUST be provided in cases where a poster follows up (responds/replies/whatever) to an article from an earlier poster. That clear requirement has been in all our drafts up to article-13. >>3.2.1 References >> The References header is the same as that specified in Section 3.6.4 >> of [RFC2822] with the added restrictions detailed in Section 2.2 and >> those listed below: >Section 2.2 does not say anything about specific headers (and nothing at >all about References), and the remaining part of 3.2.1 says nothing about >followups and MUST. RFC2822 says "SHOULD". The text regarding the References header in the new usefor-03 draft is just plain WRONG. Yes. the bit about section 2.2 covers the obligatory SP after the ':', as required in all headers, and the "listed below" refers to the avoidance of comments (MUST accept but do not generate yet). But the requirement for the header to be present for followups has been omitted, and it needs to be put back (modulo the relatively minor matter of the precise wording). And there is also an error in the syntax of the References header, which Frank Ellermann pointed out some while back, and which needs to be corrected. -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3C9W4ur043058 for ; Tue, 12 Apr 2005 02:32:04 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3C9W4H3043057 for ietf-usefor-skb; Tue, 12 Apr 2005 02:32:04 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from rufus.isode.com (rufus.isode.com [62.3.217.251]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3C9W1WJ043010 for ; Tue, 12 Apr 2005 02:32:03 -0700 (PDT) (envelope-from alexey.melnikov-usefor@isode.com) Received: from [192.168.0.3] ([62.3.217.253]) by rufus.isode.com via TCP (internal) with ESMTPA; Tue, 12 Apr 2005 10:31:55 +0100 Message-ID: <425B958A.9010107@isode.com> Date: Tue, 12 Apr 2005 10:31:54 +0100 From: Alexey Melnikov User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.7.2) Gecko/20040804 Netscape/7.2 (ax) X-Accept-Language: en-us, en MIME-Version: 1.0 To: Russ Allbery CC: ietf-usefor@imc.org, Dr John Stockton Subject: Re: Fwd: http://www.ietf.org/internet-drafts/draft-ietf-usefor-usefor-03.txt References: <200504111212.j3BCChM14516@clerew.man.ac.uk> <87psx1gnjc.fsf@windlord.stanford.edu> In-Reply-To: <87psx1gnjc.fsf@windlord.stanford.edu> Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Russ Allbery wrote: >Charles Lindsey writes: > > >>"Dr John Stockton" writes: >> >> >>>"3.3.1 Lines" >>> >>> >>>This may not be needed for the transmission of News; but it is very >>>useful for Kill Rules. Therefore, IMHO it should be compulsory, not >>>obsolete (unless replaced by, say, 'Bytes'). >>> >>> > >Kill Rules are based on overview information, not on article headers >except in unusual situations (since killing on full headers is very slow). >Overview contains its own line count that has nothing to do with an >article header; this is specified by the NNTP protocol. > >You will find, if you analyze articles in the wild, that a significant >percentage of Lines headers are wrong. > > >>I think the WG agreed early on that this header had passed its >>sell-by-date, though I do not think we intended it to be declared >>"obsolete" as the current Usefor draft implies. >> >> >I believe that we did. We intended it to be declared obsolete in the >sense that nothing should generate it and nothing should use it. > > Exactly. I don't think any clarification in the document is needed. Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3BK02ZE069521 for ; Mon, 11 Apr 2005 13:00:02 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3BK02Zp069520 for ietf-usefor-skb; Mon, 11 Apr 2005 13:00:02 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from ciao.gmane.org (main.gmane.org [80.91.229.2]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3BK00VW069503 for ; Mon, 11 Apr 2005 13:00:01 -0700 (PDT) (envelope-from usenet-format@gmane.org) Received: from list by ciao.gmane.org with local (Exim 4.43) id 1DL51h-0005A7-Fi for ietf-usefor@imc.org; Mon, 11 Apr 2005 21:56:53 +0200 Received: from 212.82.251.33 ([212.82.251.33]) by main.gmane.org with esmtp (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Mon, 11 Apr 2005 21:56:53 +0200 Received: from nobody by 212.82.251.33 with local (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Mon, 11 Apr 2005 21:56:53 +0200 X-Injected-Via-Gmane: http://gmane.org/ To: ietf-usefor@imc.org From: Frank Ellermann Subject: Re: Path header delimiters Date: Mon, 11 Apr 2005 21:58:39 +0200 Organization: Lines: 25 Message-ID: <425AD6EF.448A@xyzzy.claranet.de> References: <425AC86F.2004@xyzzy.claranet.de> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Complaints-To: usenet@sea.gmane.org X-Gmane-NNTP-Posting-Host: 212.82.251.33 X-Mailer: Mozilla 3.0 (OS/2; U) Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Henry Spencer wrote: [MISMATCH] > As a standalone item in the path content? I think not. Probably I was wrong, a quick check confirmed your version ".MISMATCH" added to the host name, not "!MISMATCH". > This would be a nuisance for its proprietors, yes, but not > a problem for the net as a whole. Also true, the former host TV apparently found a better name. A potential host AKA after the introduction of TLD AKA could also do this as news server. Besides ...!news.aka!AKA!... would be fun. More fun: Any news about publishing s-o-1036 as RfC 4036 ? It's apparently not yet in the RfC editor queue. Less fun: Charles and I need a volunteer for some kind of "rough consensus" about the names (sic!) of some msg-id ABNF productions, could you please toss a coin or add a comment ? Bye, Frank Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3BJ9auO059661 for ; Mon, 11 Apr 2005 12:09:36 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3BJ9ab7059660 for ietf-usefor-skb; Mon, 11 Apr 2005 12:09:36 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from spsystems.net (spsystems.net [216.126.83.115]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3BJ9abd059650 for ; Mon, 11 Apr 2005 12:09:36 -0700 (PDT) (envelope-from henry@spsystems.net) Received: from spsystems.net (henry@localhost [127.0.0.1]) by spsystems.net (8.12.10/8.12.10) with ESMTP id j3BJ9WVO024343; Mon, 11 Apr 2005 15:09:32 -0400 (EDT) Received: (from henry@localhost) by spsystems.net (8.12.10/8.12.10/Submit) id j3BJ9WTO024342; Mon, 11 Apr 2005 15:09:32 -0400 (EDT) Date: Mon, 11 Apr 2005 15:09:32 -0400 (EDT) From: Henry Spencer To: Usefor Mailing List Subject: Re: Path header delimiters In-Reply-To: <425AC86F.2004@xyzzy.claranet.de> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: On Mon, 11 Apr 2005, Frank Ellermann wrote: > > I would suggest would be to use "AKA" -- Also Known As -- > > instead of MISMATCH > > Two problems: MISMATCH is already "common practice"... As a standalone item in the path content? I think not. There is some vaguely-related existing practice of appending it to server names, which might be considered a minor point in its favor. > and we > hope that no UUCP host MISMATCH existed, and that no TLD with > this name will be introduced. For "AKA" I'm less sure... We can quite reasonably have the same hopes for AKA. Remember also that it doesn't greatly matter whether AKA is potentially a valid server name, because such a server can be known by a nickname (e.g. "news.aka") for news purposes. This would be a nuisance for its proprietors, yes, but not a problem for the net as a whole. Henry Spencer henry@spsystems.net Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3BIwH36057762 for ; Mon, 11 Apr 2005 11:58:19 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3BIwHgK057761 for ietf-usefor-skb; Mon, 11 Apr 2005 11:58:17 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from ciao.gmane.org (main.gmane.org [80.91.229.2]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3BIwEQw057740 for ; Mon, 11 Apr 2005 11:58:15 -0700 (PDT) (envelope-from usenet-format@gmane.org) Received: from list by ciao.gmane.org with local (Exim 4.43) id 1DL43W-0004CA-5n for ietf-usefor@imc.org; Mon, 11 Apr 2005 20:54:42 +0200 Received: from 212.82.251.33 ([212.82.251.33]) by main.gmane.org with esmtp (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Mon, 11 Apr 2005 20:54:42 +0200 Received: from nobody by 212.82.251.33 with local (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Mon, 11 Apr 2005 20:54:42 +0200 X-Injected-Via-Gmane: http://gmane.org/ To: ietf-usefor@imc.org From: Frank Ellermann Subject: Re: Path header delimiters Date: Mon, 11 Apr 2005 20:56:47 +0200 Organization: Lines: 12 Message-ID: <425AC86F.2004@xyzzy.claranet.de> References: Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Complaints-To: usenet@sea.gmane.org X-Gmane-NNTP-Posting-Host: 212.82.251.33 X-Mailer: Mozilla 3.0 (OS/2; U) Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Henry Spencer wrote: > I would suggest would be to use "AKA" -- Also Known As -- > instead of MISMATCH Two problems: MISMATCH is already "common practice", and we hope that no UUCP host MISMATCH existed, and that no TLD with this name will be introduced. For "AKA" I'm less sure, funny host names are AI, IO, PH, PN, TM, and TW - in theory these hosts could be news servers. Bye, Frank Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3BIjjOX055327 for ; Mon, 11 Apr 2005 11:45:45 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3BIjjSW055326 for ietf-usefor-skb; Mon, 11 Apr 2005 11:45:45 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtp1.Stanford.EDU (smtp1.Stanford.EDU [171.67.16.123]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3BIjiff055319 for ; Mon, 11 Apr 2005 11:45:44 -0700 (PDT) (envelope-from rra@stanford.edu) Received: from windlord.stanford.edu (windlord.Stanford.EDU [171.64.19.147]) by smtp1.Stanford.EDU (8.12.11/8.12.11) with SMTP id j3BIjhW3009007 for ; Mon, 11 Apr 2005 11:45:43 -0700 Received: (qmail 32689 invoked by uid 1000); 11 Apr 2005 18:45:43 -0000 To: ietf-usefor@imc.org, "Dr John Stockton" Subject: Re: Fwd: http://www.ietf.org/internet-drafts/draft-ietf-usefor-usefor-03.txt In-Reply-To: <200504111212.j3BCChM14516@clerew.man.ac.uk> (Charles Lindsey's message of "Mon, 11 Apr 2005 13:12:43 +0100 (BST)") References: <200504111212.j3BCChM14516@clerew.man.ac.uk> From: Russ Allbery Organization: The Eyrie Date: Mon, 11 Apr 2005 11:45:43 -0700 Message-ID: <87psx1gnjc.fsf@windlord.stanford.edu> User-Agent: Gnus/5.1007 (Gnus v5.10.7) XEmacs/21.4 (Jumbo Shrimp, linux) MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Charles Lindsey writes: > "Dr John Stockton" writes: >> "3.3.1 Lines" >> This may not be needed for the transmission of News; but it is very >> useful for Kill Rules. Therefore, IMHO it should be compulsory, not >> obsolete (unless replaced by, say, 'Bytes'). Kill Rules are based on overview information, not on article headers except in unusual situations (since killing on full headers is very slow). Overview contains its own line count that has nothing to do with an article header; this is specified by the NNTP protocol. You will find, if you analyze articles in the wild, that a significant percentage of Lines headers are wrong. > I think the WG agreed early on that this header had passed its > sell-by-date, though I do not think we intended it to be declared > "obsolete" as the current Usefor draft implies. I believe that we did. We intended it to be declared obsolete in the sense that nothing should generate it and nothing should use it. -- Russ Allbery (rra@stanford.edu) Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3BHhaEF043044 for ; Mon, 11 Apr 2005 10:43:36 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3BHhaje043043 for ietf-usefor-skb; Mon, 11 Apr 2005 10:43:36 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from b.mail.peak.org (b.mail.peak.org [69.59.192.42]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3BHhaJ0043019 for ; Mon, 11 Apr 2005 10:43:36 -0700 (PDT) (envelope-from stanley@peak.org) Received: from a.shell.peak.org ([69.59.192.81]) by b.mail.peak.org (8.12.10/8.12.8) with ESMTP id j3BHhUTF066946 (version=TLSv1/SSLv3 cipher=EDH-RSA-DES-CBC3-SHA bits=168 verify=NO) for ; Mon, 11 Apr 2005 10:43:30 -0700 (PDT) Date: Mon, 11 Apr 2005 10:43:30 -0700 (PDT) From: John Stanley X-X-Sender: stanley@a.shell.peak.org To: ietf-usefor@imc.org Subject: Re: Issues outstanding Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Spam-Score: 0 () X-Scanned-By: MIMEDefang 2.39 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: "Charles Lindsey" : >(Oddly, I noticed that John Stanley was in favour of Posted-And-Mailed - Why is this so odd, Charles? I was one of the people who worked on the draft that tried to define PAM and MCT. Do you think I would have spent my time on that were I opposed to them? >he is quite right, but if Mail- Copies-To has to wait, then >Posted-And-Mailed must wait too.) If our task is to document existing practice, and if you were correct in writing: >It is still not clear (to me) what the objection to keeping them is, and I >see no merit at all in #2b (since these headers are in moderately common >use, and the "experiment" has, in effect, been done). as Alexy quoted you as saying, then it would be incorrect NOT to include them. Are they not "current practice", are they NOT in moderately common use, or are we sticking our heads in the sand and ignoring the task we are supposed to accomplish here? I notice that Alexy did not debate the validitiy of your statement, he just decided the issue is closed and we will not document existing practice here. Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3BHXvoP040743 for ; Mon, 11 Apr 2005 10:33:57 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3BHXvlI040742 for ietf-usefor-skb; Mon, 11 Apr 2005 10:33:57 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from a.mail.peak.org (a.mail.peak.org [69.59.192.41]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3BHXu7E040718 for ; Mon, 11 Apr 2005 10:33:56 -0700 (PDT) (envelope-from stanley@peak.org) Received: from a.shell.peak.org ([69.59.192.81]) by a.mail.peak.org (8.12.10/8.12.8) with ESMTP id j3BHXokK038935 (version=TLSv1/SSLv3 cipher=EDH-RSA-DES-CBC3-SHA bits=168 verify=NO) for ; Mon, 11 Apr 2005 10:33:51 -0700 (PDT) Date: Mon, 11 Apr 2005 10:33:50 -0700 (PDT) From: John Stanley X-X-Sender: stanley@a.shell.peak.org To: ietf-usefor@imc.org Subject: Re: Differences between RFC 2822 and Usefor Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Spam-Score: 0 () X-Scanned-By: MIMEDefang 2.39 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: "Charles Lindsey" : > Differences between RFC 2822 and Usefor. > ---------------------------------------- >2.11 A References header MUST be provided for followups (as opposed to >SHOULD be provided for replies in RFC 2822). Would you care to explain where you are getting this difference? In draft-ietf-usefor-usefor-03.txt, the References header is defined thusly: >3.2.1 References > The References header is the same as that specified in Section 3.6.4 > of [RFC2822] with the added restrictions detailed in Section 2.2 and > those listed below: Section 2.2 does not say anything about specific headers (and nothing at all about References), and the remaining part of 3.2.1 says nothing about followups and MUST. RFC2822 says "SHOULD". There is no section 2.11 at all, much less one that creates a difference between RFC2822 and USEFOR. So, it appears to me that the artificial requirement for References has been removed from the draft, as would be proper based upon a strict reading of RFC2119 and when a MUST is appropriate. So be it. That's one way of solving the debate between "followups MUST/non-followups MUST NOT" and "followups MUST/non-followups MAY". Not the way I would have picked, but one that is certainly RFC2119 conforming. Of course, now that this change has been made, it will require a full justification for putting back an extraneous (w.r.t. RFC2119) MUST condition. There is no interoperability issue, nothing breaks when it isn't there, so it will take some special arguments to justify replacing it. Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3BGChVj023522 for ; Mon, 11 Apr 2005 09:12:43 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3BGChfa023520 for ietf-usefor-skb; Mon, 11 Apr 2005 09:12:43 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtp802.mail.ukl.yahoo.com (smtp802.mail.ukl.yahoo.com [217.12.12.139]) by above.proper.com (8.12.11/8.12.9) with SMTP id j3BGCgPA023491 for ; Mon, 11 Apr 2005 09:12:42 -0700 (PDT) (envelope-from chl@clerew.man.ac.uk) Received: from unknown (HELO host81-144-77-126.midband.mdip.bt.net) (ietf-usefor@imc.org@81.144.77.126 with poptime) by smtp802.mail.ukl.yahoo.com with SMTP; 11 Apr 2005 16:12:35 -0000 Received: (from chl@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3BCChM14516; Mon, 11 Apr 2005 13:12:43 +0100 (BST) Date: Mon, 11 Apr 2005 13:12:43 +0100 (BST) From: Charles Lindsey Message-Id: <200504111212.j3BCChM14516@clerew.man.ac.uk> To: ietf-usefor@imc.org Cc: "Dr John Stockton" Subject: Re: Fwd: http://www.ietf.org/internet-drafts/draft-ietf-usefor-usefor-03.txt References: X-Newsreader: NN version 6.5.2 (NOV) Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In local.usefor you write: >------- Forwarded message ------- >From: "Dr John Stockton" >To: "Charles. H. Lindsey" >Subject: >http://www.ietf.org/internet-drafts/draft-ietf-usefor-usefor-03.txt >Date: Sun, 10 Apr 2005 19:44:55 +0100 >I've also been reading >http://www.ietf.org/internet-drafts/draft-ietf-usefor-usefor-03.txt, >April 2005. I think most of these points are for Ken Murchison to answer. So I shall just make a few comments. >"Changes since draft-ietf-usefor-usefor-01 >... > o Only allow "UT and "GMT" in Date header; disallow all other zone>." >AIUI, it is UTC and not UT that is distributed by time servers, in which >case a label UT would in practice be incorrect in principle. UT != UTC. >Anyway, IIRC, UT is GMT, except when it is the University of Texas. The only reason these two s remain in the draft is that they are in current widespread use (well, GMT is anyway). So they must continue to be accepted, even though they MUST NOT be generated in future. Generally speaking, neither news nor email software cares in the least about the subtleties of UTC vs GMT. Times, as they appear in Date headers, are just approximations to the time the article was written, and cannot be assumed to be accurate to better than a few minutes. >Note that while input of UTC should allow 23:59:60,... That format for time is already permitted by RFC 2822, though I doubt any operating system ever produces it in real practice. >"1.1 Basic Concepts > "Netnews" is a set of protocols for generating, storing and > retrieving news "articles" (which are a subset of Email messages)" >They are surely not a subset of E-mail messages. They might be >described as similar. But E-mail is now commonly HTML & News should not >be; I'd omit the parenthetic analogy. Yes, it might be better to say "(whose format is a subset of that for Email messages)". Ken? >"1.5 Definitions >... > When an article is posted to more than one newsgroup, it > is said to be "crossposted"; ..." >'When a single [copy of an ] article ...' ??? I am not convinced that needs changing. >"2.2 Headers" and elsewhere. >The common understanding of "Header[s]" is "That cabalistic stuff before >the message starts", whereas Header is evidently used here to refer to a >logical line starting with such as "Date: ". We took a conscious decision to use the terminology in widespread use, rather than that defined in RFC 2822. The new NNTP draft is following that same approach. So the "cabalistic stuff" is the "Headers" (plural) and the Date header, for example, is a "Header" (singular). >ISTM that, clear definitions notwithstanding, there is an undue risk of >reader error unless there is a very clear written distinction in all >references to the entire heading material, the logical (de-folded) >lines, and the physical (to CRLF) lines. That part got me confused. I think (hope) that the draft always makes it clear from the context when a single "header line" is being discussed. >"3.1.2 Date" >ISTM that "GMT" is useful to the human reader, as indicating a choice of >the world standard time rather than the local one. My Turnpike here >uses +0000 in Winter Time, which carries an expectation that it will use >+0100 in Summer Time, and that the relevant agent is in the London time >zone. GMT would carry no such implication. Then agents such as Google >should be encouraged to use GMT rather than -0800, if they cannot use >the author's zone. In other words, I suggest un-deprecation of GMT >(and/or UT/UTC). I don't think there has ever been any expectation that you can guess, from somebody's Date header, whether that somebody lives in a location where daylight saving applies. And I believe Madrid also uses +0000 in Winter (and if Madrid doesn't, then Lisbon surely will - and what about Casablanca)? Generally speaking, the poster's timezone, as written, should be preserved as the article propagates or gets stored (indeed, gratuitous changes of headers are forbidden by Usepro). I find it useful to know whether someone really was taking the trouble to stay up late to respond to my messages. >Presumably absence of a numeric offset indication, while not allowed, >would be treated as +0000. Possibly. >"3.3.1 Lines" >This may not be needed for the transmission of News; but it is very >useful for Kill Rules. Therefore, IMHO it should be compulsory, not >obsolete (unless replaced by, say, 'Bytes'). I think the WG agreed early on that this header had passed its sell-by-date, though I do not think we intended it to be declared "obsolete" as the current Usefor draft implies. I think "obsolescent" is the right term. At least the draft does define exactly which lines are supposed to get counted, which had always been a bit vague hitherto. >"5. Security Considerations >... > Agents that generate message-ids for news articles > SHOULD ensure that they are unpredictable." >It would be undesirable to lose the ability to kill-rule on partial >message-ID, or to kill spam addressed to message-IDs using partial ID, >or to search news databases on partial ID. Therefore, I suggest >something like "incompletely predictable". The right hand side in fact >should be predictable. RFC 2822 RECOMMENDS (which is a pretty strong term) that the RHS should be the domain of the sender (or of his ISP), and Usefor inherits that RECCOMENDATION. -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3BGCdGc023504 for ; Mon, 11 Apr 2005 09:12:39 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3BGCdrc023503 for ietf-usefor-skb; Mon, 11 Apr 2005 09:12:39 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtp802.mail.ukl.yahoo.com (smtp802.mail.ukl.yahoo.com [217.12.12.139]) by above.proper.com (8.12.11/8.12.9) with SMTP id j3BGCcVS023474 for ; Mon, 11 Apr 2005 09:12:38 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from unknown (HELO host81-144-77-126.midband.mdip.bt.net) (ietf-usefor@imc.org@81.144.77.126 with poptime) by smtp802.mail.ukl.yahoo.com with SMTP; 11 Apr 2005 16:12:32 -0000 Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3BGCEZ15275 for ietf-usefor@imc.org; Mon, 11 Apr 2005 17:12:14 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20651 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: http://www.ietf.org/internet-drafts/draft-ietf-usefor-useage-01.txt Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: <200504110250.j3B2oOf20818@panix5.panix.com> Date: Mon, 11 Apr 2005 12:14:32 GMT Lines: 29 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In <200504110250.j3B2oOf20818@panix5.panix.com> Seth Breidbart writes: >"Charles Lindsey" wrote: >> In "Dr John Stockton" >> writes: >>> Moreover, articles crossposted to many newsgroups SHOULD be >>> considered to have been read once they have been seen in any of >>> those groups." >> >>>But "many" is not the right word, surely. Cross-posting can be to 2 >>>groups, and 2 < a few < many. >> >> s/many/several/ >How about "multiple"? I think "several" is good enough. Anyone else preferring "multiple"? -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3BFpKbG019098 for ; Mon, 11 Apr 2005 08:51:20 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3BFpKk1019097 for ietf-usefor-skb; Mon, 11 Apr 2005 08:51:20 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from spsystems.net (spsystems.net [216.126.83.115]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3BFpJqN019080 for ; Mon, 11 Apr 2005 08:51:19 -0700 (PDT) (envelope-from henry@spsystems.net) Received: from spsystems.net (henry@localhost [127.0.0.1]) by spsystems.net (8.12.10/8.12.10) with ESMTP id j3BFpEVO022083; Mon, 11 Apr 2005 11:51:14 -0400 (EDT) Received: (from henry@localhost) by spsystems.net (8.12.10/8.12.10/Submit) id j3BFpEmU022082; Mon, 11 Apr 2005 11:51:14 -0400 (EDT) Date: Mon, 11 Apr 2005 11:51:13 -0400 (EDT) From: Henry Spencer To: Usefor Mailing List Subject: Re: Path header delimiters In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: On Sat, 9 Apr 2005, Charles Lindsey wrote: > E. A variant on scheme D > Path: good-site.com!!mallet.com!MISMATCH!dodgy.com!old-site.com! > new-site.com!!injector.com!POSTED!not-for-mail > This avoids any delimiter other than '!', but it assumes that two > delimiters in succession will not cause any trouble. But it is clear that > RFC 1036 permits such usage... This sounds good to me. The one slight amendment I would suggest would be to use "AKA" -- Also Known As -- instead of MISMATCH, making it a bit clearer just what's going on. (I struggled to find a concise way of writing "claimed to be", which is the underlying idea, and this was the best I could think of.) I support scheme E even if this suggestion is not accepted. Henry Spencer henry@spsystems.net Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3BEaqTJ003987 for ; Mon, 11 Apr 2005 07:36:52 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3BEaq4n003986 for ietf-usefor-skb; Mon, 11 Apr 2005 07:36:52 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from ciao.gmane.org (main.gmane.org [80.91.229.2]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3BEamHX003967 for ; Mon, 11 Apr 2005 07:36:50 -0700 (PDT) (envelope-from usenet-format@gmane.org) Received: from list by ciao.gmane.org with local (Exim 4.43) id 1DKzxX-0006do-Ce for ietf-usefor@imc.org; Mon, 11 Apr 2005 16:32:15 +0200 Received: from c-134-88-77.hh.dial.de.ignite.net ([62.134.88.77]) by main.gmane.org with esmtp (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Mon, 11 Apr 2005 16:32:15 +0200 Received: from nobody by c-134-88-77.hh.dial.de.ignite.net with local (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Mon, 11 Apr 2005 16:32:15 +0200 X-Injected-Via-Gmane: http://gmane.org/ To: ietf-usefor@imc.org From: Frank Ellermann Subject: Re: Broken Message-ID syntax Date: Mon, 11 Apr 2005 16:32:03 +0200 Organization: Lines: 44 Message-ID: <425A8A63.268A@xyzzy.claranet.de> References: <4255153E.3080500@isode.com> <42594879.3B92@xyzzy.claranet.de> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Complaints-To: usenet@sea.gmane.org X-Gmane-NNTP-Posting-Host: c-134-88-77.hh.dial.de.ignite.net X-Mailer: Mozilla 3.0 (OS/2; U) Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Charles Lindsey wrote: > you can count me out for any name change for and > . That's why we need a third opinion like Alexejs's about this point, or a better question. A recent joke on ietf.general: /\ /\ __/ \__ ___/ \___/\___ _/\__/\__/\_ consensus rough consensus bad question [ABNF] > Well not with those names :-( . As long as the RHS contains "domain" and the "address-literal" is an "address-literal" you could beutify it, all productions starting with "id-" could be a good idea (id-local, id-domain, etc.) > It prevents the msg-id > <"foo.bar.baz"@example.com> > as required Yes, that's the idea: Leading dot or trailing dot or unique-literal <=> quote No leading/trailing dot and no unique-literal <=> don't quote > it also prevents > <"@.@"@example.com> "@" is a "unique literal" => quote. Working as designed. "@" is no atext => no dot-atom-text => unquoted is illegal. Where's the problem ? > the "[[Adjacent dots should not be allowed]]" remark in > the draft is not a correct description of the problem. In the proposed "unique" ABNF adjacent dots are handled by ".." in "unique-literal". Ugly like hell, but who cares if it's correct. Bye, Frank Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3BBClG3032398 for ; Mon, 11 Apr 2005 04:12:47 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3BBClZU032397 for ietf-usefor-skb; Mon, 11 Apr 2005 04:12:47 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtp805.mail.ukl.yahoo.com (smtp805.mail.ukl.yahoo.com [217.12.12.195]) by above.proper.com (8.12.11/8.12.9) with SMTP id j3BBCjfk032346 for ; Mon, 11 Apr 2005 04:12:46 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from unknown (HELO host81-144-71-41.midband.mdip.bt.net) (ietf-usefor@imc.org@81.144.71.41 with poptime) by smtp805.mail.ukl.yahoo.com with SMTP; 11 Apr 2005 11:12:39 -0000 Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3B9wOe13681 for ietf-usefor@imc.org; Mon, 11 Apr 2005 10:58:24 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20650 Newsgroups: local.usefor Path: clerew!clerew!x Received: from clerew.man.ac.uk (localhost [127.0.0.1]) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) with ESMTP id j3B9uWG13620 for ; Mon, 11 Apr 2005 10:56:33 +0100 (BST) Subject: Fwd: http://www.ietf.org/internet-drafts/draft-ietf-usefor-usefor-03.txt References: Message-ID: To: local.usefor@clerew.man.ac.uk From: "Charles Lindsey" Content-Type: text/plain; format=flowed; delsp=yes; charset=iso-8859-1 MIME-Version: 1.0 Content-Transfer-Encoding: 8bit Date: Mon, 11 Apr 2005 10:56:25 +0100 In-Reply-To: User-Agent: Opera M2(BETA1)/8.00 (SunOS, build 913) Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: ------- Forwarded message ------- From: "Dr John Stockton" To: "Charles. H. Lindsey" Subject: http://www.ietf.org/internet-drafts/draft-ietf-usefor-usefor-03.txt Date: Sun, 10 Apr 2005 19:44:55 +0100 I've also been reading http://www.ietf.org/internet-drafts/draft-ietf-usefor-usefor-03.txt, April 2005. "Changes since draft-ietf-usefor-usefor-01 ... o Only allow "UT and "GMT" in Date header; disallow all other ." AIUI, it is UTC and not UT that is distributed by time servers, in which case a label UT would in practice be incorrect in principle. UT != UTC. Anyway, IIRC, UT is GMT, except when it is the University of Texas. Note that while input of UTC should allow 23:59:60, in this context output should not generate it, since an error of <= one second is better than a risk of a receiving agent getting upset about it. A posting agent can just wait for a leap second to have finished. However, I see that the tenor of 3.1.2 is incompatible with the introduction of UTC. "1.1 Basic Concepts "Netnews" is a set of protocols for generating, storing and retrieving news "articles" (which are a subset of Email messages)" They are surely not a subset of E-mail messages. They might be described as similar. But E-mail is now commonly HTML & News should not be; I'd omit the parenthetic analogy. "1.2 Scope ... An best ..." Ouch. "1.5 Definitions ... When an article is posted to more than one newsgroup, it is said to be "crossposted"; ..." 'When a single [copy of an ] article ...' ??? "2.2 Headers" and elsewhere. The common understanding of "Header[s]" is "That cabalistic stuff before the message starts", whereas Header is evidently used here to refer to a logical line starting with such as "Date: ". ISTM that, clear definitions notwithstanding, there is an undue risk of reader error unless there is a very clear written distinction in all references to the entire heading material, the logical (de-folded) lines, and the physical (to CRLF) lines. That part got me confused. "3.1.2 Date" ISTM that "GMT" is useful to the human reader, as indicating a choice of the world standard time rather than the local one. My Turnpike here uses +0000 in Winter Time, which carries an expectation that it will use +0100 in Summer Time, and that the relevant agent is in the London time zone. GMT would carry no such implication. Then agents such as Google should be encouraged to use GMT rather than -0800, if they cannot use the author's zone. In other words, I suggest un-deprecation of GMT (and/or UT/UTC). Presumably absence of a numeric offset indication, while not allowed, would be treated as +0000. "3.3.1 Lines" This may not be needed for the transmission of News; but it is very useful for Kill Rules. Therefore, IMHO it should be compulsory, not obsolete (unless replaced by, say, 'Bytes'). "5. Security Considerations ... Agents that generate message-ids for news articles SHOULD ensure that they are unpredictable." It would be undesirable to lose the ability to kill-rule on partial message-ID, or to kill spam addressed to message-IDs using partial ID, or to search news databases on partial ID. Therefore, I suggest something like "incompletely predictable". The right hand side in fact should be predictable. Example : the first two of the last four characters of a Turnpike message-ID are, for a given setup, constant over long periods of time; mine are currently Fw. Therefore, it could be useful for me to kill all mail to .*fw..@, provided that I have no user called halfwit@ or similar. AIUI, newer Turnpike allows the user to choose the last few characters of the left part, which is even better as a kill target. Regards, -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133   Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk      Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9      Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3BAN0oO010436 for ; Mon, 11 Apr 2005 03:23:00 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3BAN0th010435 for ietf-usefor-skb; Mon, 11 Apr 2005 03:23:00 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtps-vbr1.xs4all.nl (smtps-vbr1.xs4all.nl [194.109.24.19]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3BAMwsV010409 for ; Mon, 11 Apr 2005 03:22:59 -0700 (PDT) (envelope-from rvtol@isolution.nl) Received: from isop10 (velvet.isolution.nl [194.109.164.102]) (authenticated bits=0) by smtps-vbr1.xs4all.nl (8.12.11/8.12.11) with ESMTP id j3BAMp7s067131 (version=TLSv1/SSLv3 cipher=RC4-MD5 bits=128 verify=NO) for ; Mon, 11 Apr 2005 12:22:56 +0200 (CEST) (envelope-from rvtol@isolution.nl) Message-ID: <0f6c01c53e80$6d4372c0$0b01a8c0@isolution.nl> From: "Ruud H.G. van Tol" To: References: <200504110250.j3B2oOf20818@panix5.panix.com> Subject: Re: http://www.ietf.org/internet-drafts/draft-ietf-usefor-useage-01.txt Date: Mon, 11 Apr 2005 12:15:50 +0200 Organization: Chaos rules. MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-15" Content-Transfer-Encoding: 7bit X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 6.00.2800.1478 X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2800.1478 X-Virus-Scanned: by XS4ALL Virus Scanner Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Seth Breidbart: > Charles Lindsey: >> Dr John Stockton: >>> In particular, that the form used, after removal of any >>> ".invalid", MUST NOT be a possible address for someone else, >>> present or reasonable future. >> I am not sure. Sometimes people just add ".invalid" to their own >> genuine address. > That's why it says "someone _else_". Same observation here. The "for" didn't emphasize the issue of ownership (or permission) though, so it was easy to read it otherwise, like the "someone else" being the spammert. -- Grtz, Ruud Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3B2oQ7U072190 for ; Sun, 10 Apr 2005 19:50:26 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3B2oQno072189 for ietf-usefor-skb; Sun, 10 Apr 2005 19:50:26 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from mail1.panix.com (mail1.panix.com [166.84.1.72]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3B2oP1b072183 for ; Sun, 10 Apr 2005 19:50:25 -0700 (PDT) (envelope-from sethb@panix.com) Received: from panix5.panix.com (panix5.panix.com [166.84.1.5]) by mail1.panix.com (Postfix) with ESMTP id 7932E58AA7 for ; Sun, 10 Apr 2005 22:50:24 -0400 (EDT) Received: (from sethb@localhost) by panix5.panix.com (8.11.6p3/8.8.8/PanixN1.1) id j3B2oOf20818; Sun, 10 Apr 2005 22:50:24 -0400 (EDT) Date: Sun, 10 Apr 2005 22:50:24 -0400 (EDT) Message-Id: <200504110250.j3B2oOf20818@panix5.panix.com> From: Seth Breidbart To: ietf-usefor@imc.org In-reply-to: (chl@clerew.man.ac.uk) Subject: Re: http://www.ietf.org/internet-drafts/draft-ietf-usefor-useage-01.txt References: Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: "Charles Lindsey" wrote: > In "Dr John Stockton" > writes: >> In particular, that the form used, after removal of any >>".invalid", MUST NOT be a possible address for someone else, present >>or reasonable future. > > I am not sure. Sometimes people just add ".invalid" to their own genuine > address. That's why it says "someone _else_". >> Moreover, articles crossposted to many newsgroups SHOULD be >> considered to have been read once they have been seen in any of >> those groups." > >>But "many" is not the right word, surely. Cross-posting can be to 2 >>groups, and 2 < a few < many. > > s/many/several/ How about "multiple"? Seth Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3B2CbYZ068802 for ; Sun, 10 Apr 2005 19:12:37 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3B2CbXr068801 for ietf-usefor-skb; Sun, 10 Apr 2005 19:12:37 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtp809.mail.ukl.yahoo.com (smtp809.mail.ukl.yahoo.com [217.12.12.199]) by above.proper.com (8.12.11/8.12.9) with SMTP id j3B2Caes068769 for ; Sun, 10 Apr 2005 19:12:36 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from unknown (HELO host81-144-75-36.midband.mdip.bt.net) (ietf-usefor@imc.org@81.144.75.36 with poptime) by smtp809.mail.ukl.yahoo.com with SMTP; 11 Apr 2005 02:12:28 -0000 Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3B2CB209324 for ietf-usefor@imc.org; Mon, 11 Apr 2005 03:12:11 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20647 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: http://www.ietf.org/internet-drafts/draft-ietf-usefor-useage-01.txt Content-Type: text/plain; charset=iso-8859-1 Message-ID: Content-Transfer-Encoding: 8bit X-Newsreader: NN version 6.5.2 (NOV) References: Mime-Version: 1.0 Date: Sun, 10 Apr 2005 22:21:14 GMT Lines: 370 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In "Dr John Stockton" writes: >I've been reading >http://www.ietf.org/internet-drafts/draft-ietf-usefor-useage-01.txt, >March 2005. >I don't think I'm qualified to address the mailing list of the Usenet >Format Working Group; If this group were confined to "qualified" people, there would be no group :-). I have copied your message to the group, and am CCing this reply to you, but you will need to subscribe if you want to see any further responses. >"1.2. Objectives >... > NOTE: The extreme irritation caused to other readers by such > violations is not to be underestimated; ..." >And by false accusations of such violations! Yes, but don`t let's over-egg the pudding :-) . >"2.2. Textual Notations >... > NOTE: While such explanatory notes may seem superfluous in > principle, they often help the less-than-omniscient reader > understand the true intent of the specification in cases where > the wording is not entirely clear." >The wording should always be entirely clear, in the opinion of the >author! But the meaning may well not be entirely obvious to such a >reader. Yes. It's a nice quote by Henry Spencer from Son-of-1036 - too good not to use it. >"3.1. The Well-Behaved Posting Agent > The implementor of a posting agent SHOULD make it possible for a > suitably perseverent poster to generate any article, ..." >I don't believe that "perseverent" is a real word, except maybe in the >full OED; "perseverant" is in Chambers' Dictionary; OK, I will follow Chambers. >Last sentence of 3.1 contains "imppose". Fixed. >"3.1.1. Construction of Headers >... > Posting agents SHOULD permit the poster to include headers of > arbitrary length (and MUST permit at least 79 characters)." >You must mean "header lines" or "characters per line" ? No, I meant headers. The next sentence tells them to fold, preferably before 79, but certainly before 998. >... > "NOTE: ... standard 80-column screen" >Since use of text-only screens is now rare, how about something more >general, such as 80-column display, or viewport ? OK, I now say "display". >... > "... WSP ..." >Undefined in useage-01;... Defined in Usefor, by way of RFC 2822. >"3.1.1.1. Date >... > Date: Sat, 26 May 2001 11:13:00 -0500 (EST)" >To rub it in, you could observe (after checking) that EST is used in >Australia, probably (+1000). If someone can quote the Autralian rule authoritatively, and what "EST" stands for there, I might even be tempted to do just that. >I'd like to see full rigorous ISO 8601 dates being allowed, with day-of- >week as an optional but recommended comment - > Date : 2001-05-26 11:13:00 -0500 (EST) (Sat) It would be nice, but the email/news Date format is too entrenched to change it now. >In a time example, I'd choose an hour greater than 12, to expand the >experience of US readers. I think US readers are sufficiently familiar with the 24 hour clock as seen in Date headers. But I have changed it anyway. >"3.1.1.2. From" >Contains the word "ro". Oops! > "NOTE: ..." >ISTM that it might be better to bow to /force majeure/, and indicate, >maybe by reference to another document, what is and what is not >reasonable common practice. In particular, that the form used, after >removal of any ".invalid", MUST NOT be a possible address for someone >else, present or reasonable future. I am not sure. Sometimes people just add ".invalid" to their own genuine address. Not always spammer proof, but most spammers are stupid (Rule #3, or something) - I get massive amounts of spam directed at the Message-ID of my articles from 3 years back. The main purpose of that sentence was to encourage people NOT to munge their addresses WITHOUT adding .invalid as well. >"3.1.1.4. Subject >... > regognized" ! Oops! >"3.1.1.7. Organization >... > ... unless ... and unless ... " >ISTM that in English that's syntactically unreasonable! "unless ... or >... "? No, I meant 'and'. Applying De Morgan's rule, I said "DIScouage UNLESS acceptable AND UNLESS useful", which is the same as "ENcourage IF acceptable OR useful" (or would be if there wasn't a possible middle to exclude). >"3.1.2.1. Signatures > A "personal signature" is a short closing text automatically added" >ISTM that it is not necessarily automatic. OK. How about "added (usually automatically)"? >Since sigs are not normally quoted, you *might* agree with the >Implementors of Turnpike that sig lines can have as many as 79 >characters each. See my own signature below :-) . >"3.1.2.3. Content-Transfer-Encoding >... > moreover, Usenet articles are very likely to include trailing > whitespace in the form of a personal signature (3.1.2.1)." >3.1.2.1 is clear that the signature follows, and does not include, the >separator, which has a whitespace. OK. 'in the form of the "-- " which introduces a personal signature'. >"3.2.1.1. Subject >... > 1. Although the "Re" (which is an abbreviation for the Latin "In re", > meaning "in the matter of", and not an abbreviation of "Reference" > as is sometimes erroneously supposed) may be understood by English > speakers, and indeed by speakers of most European Languages, its > use in a newsgroup where articles were customarily written in > Arabic, or Hindi, or Chinese would be less than helpful." >That seems disrespectful to Eastern intelligence. If they can manage to >use News, they should be able to recognise the odd alien term (however >displayed) occurring always in a specific context. Apparently a lot or orientals are very unfamiliar with the Latin alphabet, but nevertheless are beginning to use mail and even news in their own languages. Granted they cannot avoid it 100% yet. There are domain-names (even if puny-coded), local-parts and newsgroup-names and much else that is still in ASCII, but those issues are being worked on (though slowly). >"3.2.1.1.1. Examples >... > Software can always recognize > that such changes have occurred from the References header." >I don't see that software can do that. ISTM that what you mean is >something different. No. Every followup MUST have a References header, so if software sees a References header it knows it is a followup (or something to be treated like a followup). So it can adjust its display accordingly (e.g. by threading); it does not need to see the "Re: ". From the software writer's POV, "Re: " is just some awkward characters that he has to detect and ignore in order to get his threading (well, some forms of threading) to work properly. >"3.2.1.3. Mail-Copies-To Has now been removed from all our drafts. >If it is permissible for a mail article to contain a newsgroups line >(I've seen it), then ISTM that a mail reading agent should give clear >warning. The intention is that the meaning of a Newsgroups header, if it is seen in an email message, is to indicate that it was also posted to that newsgroup (but replies to that email message should not include such a header unless they are posted to the newsgroup as well). >"3.2.1.4. References >... > carefully put their > by precursors.]" >"there" ! Oops! But it was not in text that was meant to go into the final version. >"3.2.2.1. Quoting and Attributions >... > SHOULD be so dintinguished ..." - distinguished. Genuine Oops! that one. >... > " The followup agent SHOULD also precede the quoted content by an > "attribution line" (however, ..." >I'd like to see a terminological amendment, to remove any taint of a >suggestion that the attribution is a single physical line. Point taken. s/"attribution line"/"attribution"/ and consequential changes elsewhere. >" The attribution MAY contain also a single (the one > from which the followup is being made), the precursor's message > identifier and/or the precursor's Date and Time." >'single' .. usually; but it can be convenient to change that manually to >indicate the degree of cross-posting. Sure. You edit anything manually as you want. But nobody wants to see more than one in the attribution, and I have never seen a user agent that put more than one (in fact they often don't put any). >'Date and Time' - IMHO, the Zone of the date & time is needed and should >be mentioned; and there should be consideration of whether this should >be a direct copy of the original Date:, whether it may be translated >into the responding poster's (or ISO) notation, whether it may be >adjusted to GMT or responder's zone. I do dislike being attributed as >having written on such as 3/21/05, or during AM/PM. I think normal practice is to use the same date-time format as in the Date header. Does anyone know of a system that changes the zone from whatever was actually in the Date header of the precursor? Basically, we are trying to document current best practice here rather than invent anything new. >" o The various fields may be separated by arbitrary text " > ^ brief OK. "arbitrary (but brief) text". Yes, we have all seen some horrors that people have inserted in there, and there is not much we can do to stop them. But no harm in giving a hint and hoping they take it :-( . >"3.3.2. Presentation of Articles >3.3.2.1. Threading > 4. Construct a tree in " ... >ISTM that it *might* be useful for a reading agent to display an >indication, for example, of the difference in tree-depth between that >article and the most previously read article in that thread, presumably >as a signed number, possibly background-coloured for sign. Of course, >all reading-agents should make thread-structure capable of being fully >seen; but it's not necessarily obvious when going from article to >article. Just a thought. I think this is an area where newsreader implementors should be encouraged to experiment. If you don't like what your newsreqder does, then you should go out and buy a better one. I think the purpose of that whole section was to point out various possibilities, each of which has disadvantages if used on its own, and to encourage experimentation with hybrids. There are some agents out there that seem to do a pretty good job, and others that don't. >"3.3.2.2. Killfiles > Moreover, articles > crossposted to many newsgroups SHOULD be considered to have been read > once they have been seen in any of those groups." >But "many" is not the right word, surely. Cross-posting can be to 2 >groups, and 2 < a few < many. s/many/several/ >"3.3.3. Interpretation of Bodies >... > Tab (US-ASCII 9) SHOULD be interpreted as sufficient horizontal white > space to reach the next of a set of fixed positions (customarily set > at every 8th character)." >"... after every 8th character)." OK. >"4.1. Construction of Headers > According to [USEPRO], an injecting agent MAY add other headers not > already provided by the poster, but SHOULD NOT alter, delete, or > reorder any existing header." >It might be useful to attempt to post a copy of that to certain >moderated newsgroups, sic, as comp.lang.asm.x86 :-( . And, perhaps, to >strengthen SHOULD NOT, if possible. No. Moderation comes before injection. Usepro makes it clear that moderators have some discretion to establish a "house style". In a well-managed hierarchy, there will be charters and moderation rules and means to enforce them. OTOH, some charters overdo it with pages and pages of rules governing the moderation process (and you and I know which group I am talking about :-( ). Our drafts need to stand well back from such issues. >"6.1.1. The 'newgroup' and 'mvgroup' Control Messages" >... >And 'rmgroup' ???? Actually no. That section was concerned with the detailed contents of newgroup etc messages, and rmgroup messages don't have any details - it's all or nothing :-) . >"9.2. Construction of Bodies > Posters SHOULD avoid using control characters and escape sequences > except for tab (US-ASCII 9), formfeed (US-ASCII 12) and, possibly, > backspace (US-ASCII 8), for reasons already explained in section > 3.3.3." >I thought that CR & LF were control characters too. Sure, but the requirements for CRLF are well documented in Usefor (via RFC 2822). >"12. Contact Address >Editor > Charles. H. Lindsey" > ^ One wonders what the dot is for. One does indeed! >P.S. I have discovered that the document bears a non-functional E-mail >address; sending the above to :- >" Email: chl@clw.cs.man.ac.uk" >gave >"A message that you sent could not be delivered to one or more of its >recipients. Ah! Manchester have finally pulled the plug on my old email address. -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3B2CaSE068794 for ; Sun, 10 Apr 2005 19:12:36 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3B2CaIv068793 for ietf-usefor-skb; Sun, 10 Apr 2005 19:12:36 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtp809.mail.ukl.yahoo.com (smtp809.mail.ukl.yahoo.com [217.12.12.199]) by above.proper.com (8.12.11/8.12.9) with SMTP id j3B2CZ32068768 for ; Sun, 10 Apr 2005 19:12:36 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from unknown (HELO host81-144-75-36.midband.mdip.bt.net) (ietf-usefor@imc.org@81.144.75.36 with poptime) by smtp809.mail.ukl.yahoo.com with SMTP; 11 Apr 2005 02:12:24 -0000 Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3B2CCG09328 for ietf-usefor@imc.org; Mon, 11 Apr 2005 03:12:12 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20648 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: Broken Mesage-ID syntax (was: Issues outstanding) Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: <4255153E.3080500@isode.com> <42594879.3B92@xyzzy.claranet.de> Date: Sun, 10 Apr 2005 23:19:09 GMT Lines: 74 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In <42594879.3B92@xyzzy.claranet.de> Frank Ellermann writes: >Alexey Melnikov wrote: >> NO-WS-CTL have been removed. Can people check that the new >> syntax is Ok? >It is not. The "pseudo-consensus" between Charles and me was >to replace msg-id-core by msg-id everywhere, and to fix the >dot-atom-text issue with leading / trailing dots plus dot-dot. Yes, I asked Ken to make that msg-id change, but he hasn't done it yet. The dot-atom-text issue is noted, but not fixed yet. >Two (Charles + me) is a bad number, could somebody like you >please volunteer as third for a proper "rough consensus" about >the last open msg-id point, a better name for the "LHS @ RHS" >productions ? I've interpreted an unrelated article from Bruce >elsewhere as support for "unique @ domain" productions, but >that's obviously dubious, You can count me in for getting rid of , but you can count me out for any name change for and . >> I suspect that some minor issues raised by Frank are yet to >> be addressed. >The msg-id and dot stuff isn't minor, the former is important, >the latter is erroneous. Here's the latest ABNF for a msg-id: >| msg-id = "<" unique "@" mdomain ">" >| unique = dot-atom-text / ( DQUOTE unique-quote DQUOTE ) >| unique-quote = ( "." [unique-part] ) / >| ( [unique-part] "." ) / >| ( [unique-part] unique-literal [unique-part] ) >| unique-part = 1*( atext / "." / unique-literal ) >| unique-literal = "(" / ")" / "," / ; all specials, minus ">", >| "[" / "]" / "@" / ; minus DQUOTE, minus "\", >| ":" / ";" / "<" / ; minus single ".", plus: >| ".." / "\\" / ( "\" DQUOTE ) >| mdomain = dot-atom-text / ("[" address-literal "]") >| address-literal = 1*( %d33-61 / ; printable ASCII minus >| %d63-90 / ; ">", "[", "\", "]" >| %d94-126 / ; plus "\[", "\\, "\]" >| "\[" / "\\" / "\]" ) Well not with those names :-( . But it is still not technically correct. It prevents the msg-id <"foo.bar.baz"@example.com> as required (because under RFC 2822 that is indistinguishable from ), but it also prevents <"@.@"@example.com> and there is no reason to do that (ugly though it may be). BTW, the "[[Adjacent dots should not be allowed]]" remark in the draft is not a correct description of the problem. -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3AFe3J2023250 for ; Sun, 10 Apr 2005 08:40:03 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3AFe3ws023249 for ietf-usefor-skb; Sun, 10 Apr 2005 08:40:03 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from ciao.gmane.org (main.gmane.org [80.91.229.2]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3AFe0Cj023235 for ; Sun, 10 Apr 2005 08:40:01 -0700 (PDT) (envelope-from usenet-format@gmane.org) Received: from list by ciao.gmane.org with local (Exim 4.43) id 1DKeUi-0000pl-70 for ietf-usefor@imc.org; Sun, 10 Apr 2005 17:37:04 +0200 Received: from 212.82.251.245 ([212.82.251.245]) by main.gmane.org with esmtp (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Sun, 10 Apr 2005 17:37:04 +0200 Received: from nobody by 212.82.251.245 with local (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Sun, 10 Apr 2005 17:37:04 +0200 X-Injected-Via-Gmane: http://gmane.org/ To: ietf-usefor@imc.org From: Frank Ellermann Subject: Broken Mesage-ID syntax (was: Issues outstanding) Date: Sun, 10 Apr 2005 17:38:33 +0200 Organization: Lines: 53 Message-ID: <42594879.3B92@xyzzy.claranet.de> References: <4255153E.3080500@isode.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Complaints-To: usenet@sea.gmane.org X-Gmane-NNTP-Posting-Host: 212.82.251.245 X-Mailer: Mozilla 3.0 (OS/2; U) Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Alexey Melnikov wrote: > NO-WS-CTL have been removed. Can people check that the new > syntax is Ok? It is not. The "pseudo-consensus" between Charles and me was to replace msg-id-core by msg-id everywhere, and to fix the dot-atom-text issue with leading / trailing dots plus dot-dot. Two (Charles + me) is a bad number, could somebody like you please volunteer as third for a proper "rough consensus" about the last open msg-id point, a better name for the "LHS @ RHS" productions ? I've interpreted an unrelated article from Bruce elsewhere as support for "unique @ domain" productions, but that's obviously dubious, > I suspect that some minor issues raised by Frank are yet to > be addressed. The msg-id and dot stuff isn't minor, the former is important, the latter is erroneous. Here's the latest ABNF for a msg-id: | msg-id = "<" unique "@" mdomain ">" | unique = dot-atom-text / ( DQUOTE unique-quote DQUOTE ) | unique-quote = ( "." [unique-part] ) / | ( [unique-part] "." ) / | ( [unique-part] unique-literal [unique-part] ) | unique-part = 1*( atext / "." / unique-literal ) | unique-literal = "(" / ")" / "," / ; all specials, minus ">", | "[" / "]" / "@" / ; minus DQUOTE, minus "\", | ":" / ";" / "<" / ; minus single ".", plus: | ".." / "\\" / ( "\" DQUOTE ) | mdomain = dot-atom-text / ("[" address-literal "]") | address-literal = 1*( %d33-61 / ; printable ASCII minus | %d63-90 / ; ">", "[", "\", "]" | %d94-126 / ; plus "\[", "\\, "\]" | "\[" / "\\" / "\]" ) Some weeks later the "no-pseudo-consensus" arrived at this: msg-id = "<" id-local "@" id-domain ">" msg-id = "<" msg-local "@" msg-domain ">" msg-id = "<" unique "@" mdomain ">" There it ended with another "pseudo-consensus" to wait for the next draft and a ruling from the chair (= you). The latter was of course a joke, but maybe you can toss a coin - in that case don't forget Charles' msg-id = "<" id-left "@" id-right ">" Bye, Frank Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j3A2CTAx040128 for ; Sat, 9 Apr 2005 19:12:29 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j3A2CTGt040127 for ietf-usefor-skb; Sat, 9 Apr 2005 19:12:29 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtp814.mail.ukl.yahoo.com (smtp814.mail.ukl.yahoo.com [217.12.12.204]) by above.proper.com (8.12.11/8.12.9) with SMTP id j3A2CR8F040115 for ; Sat, 9 Apr 2005 19:12:28 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from unknown (HELO host81-144-72-58.midband.mdip.bt.net) (ietf-usefor@imc.org@81.144.72.58 with poptime) by smtp814.mail.ukl.yahoo.com with SMTP; 10 Apr 2005 02:12:21 -0000 Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j3A2CA802775 for ietf-usefor@imc.org; Sun, 10 Apr 2005 03:12:10 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20643 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Path header delimiters Message-ID: X-Newsreader: NN version 6.5.2 (NOV) Date: Sat, 9 Apr 2005 21:06:31 GMT Lines: 138 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Our Chair has ruled that this matter it still open, so here is a summary of where I think we are at. Many years ago, when this WG was first formed, we were concerned that malefactors regularly preloaded the Path header with all sorts of sites (usually culled for some other article), in an attempt to disguise the point at which it had been injected. We therefore proposed to modify the Path header so that relaying agents would be able to make the following assertions when they add a new path-identity to the Path-header: #1 I am the injecting site. #2 I have checked the identity of the previous site, and I believe the path-identity inserted by that site to be correct. #3 I have checked the identity of the previous site, and I do not believe the path-identity claimed by that site; here is what I believe to be the true identity of that site. #4 I have made no checks on the identity of the previous site. The Path header is to News as the sequence of Received headers is to Email, and Received headers currently convey much the same information, and very useful they are when trying to trace the origin of some spam or other malefaction. I think we are still agreed that such a feature in the Path header is desirable - the point at issue is how to convey this information in a backwards compatible manner. The original suggestion was to introduce the delimiters '%', '/', and '?' in addition to the usual '!', it being apparently the case that RFC 1036 already allowed all these characters, and many more, to be used as delimiters, and it appeared that existing implementations did indeed accept them. However, RFC 1036 is somewhat vague on exactly what the allowed set of characters was. A few months ago, Bruce Lilly pointed out that maybe some implementations were not so accomodating, and pointed to the implementation of BNews (written by the author of RFC 1036) which actually accepted a rather small set. Following from that, various alternative schemes were proposed, as follows. In the following running example injector.com always uses #1 new-site.com always uses #2 or #3 good-site.com always uses #2 or #3 old-site always uses #4 dodgy.com was a bogus identity actually inserted by mallet.com A. Current draft: ----------------- Uses '%' for #1, '/' for #2, '?' for #3 amd '!' for #4 Path: good-site.com/mallet.com?dodgy.com!old-site.com! new-site.com/injector.com%not-for-mail B. Henry's proposal : ------------------------------------------------------------------ Uses '@' for #1, ',' for #2, ' ' for #3 amd '!' for #4, since it is clear from RFC 1036 that all of those are intended to be usable as delimiters. Path: good-site.com,mallet.com dodgy.com!old-site.com! new-site.com,injector.com@not-for-mail Observe that the ' ' delimiter turns up rather conveniently as a separator between the correct and bogus identities of mallet.com. One would need to discuss whether FWS as well as SP should delimit this case. C. The Diablo scheme -------------------- I still have not been able to find documentation on this, but from observed instances it appears to work as follows: Path: good-site.com!mallet.com.MISMATCH!dodgy.com!old-site.com! new-site.com!injector.com.POSTED!not-for-mail I see 2 problems with this one: 1: Any site which peers with injector.com (e.g. new-site.com) would normally scan the received Path for occurrences of "injector.com", and would send the article back to injector.com if it was not found (which, of course, it isn't here because it recorded itself as "injector.com.POSTED"). 2: It provides no distinction betwen cases #2 and #4, which rather defeats the object of the whole exercise. D. Another possible scheme -------------------------- If you want to avoid all delimiters other than '!', and to overcome the problems with the Diablo scheme, then here is one which relies on special keywords "M", "MISMATCH" and "POSTED" in places where the current syntax would expect a path-identity. Path: good-site.com!M!mallet.com!MISMATCH!dodgy.com!old-site.com! new-site.com!M!injector.com!POSTED!not-for-mail It makes the Path a little longer, but not unacceptably so, and assumes that those keywords will never represent real sites. E. A variant on scheme D ------------------------ Instead of the keyword 'M' to indicate case #2, just place two '#' delimiters in succession, giving: Path: good-site.com!!mallet.com!MISMATCH!dodgy.com!old-site.com! new-site.com!!injector.com!POSTED!not-for-mail This avoids any delimiter other than '!', but it assumes that two delimiters in succession will not cause any trouble. But it is clear that RFC 1036 permits such usage, because it uses two delimiters in succession in some of its own examples. My own view is that we should adopt scheme E. It is much easier to recognize what is meant, rather than having to remember exactly what '%', '/', and '?' mean. In the discussion since I posted that last list of outstanding issues, Frank Ellerman agreed that we should go with this scheme (having carefully examined those examples in RFC 1036), and nobody suggested anything different. So that scheme is the front runner at the moment. I could easily propose modifications to the existing texts to incorporate that scheme, but I don't want to embark on that unless we are agreed it is the way to go. So does anyone else want to comment, and can I assume that, if there are no objections, we go with scheme E? -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j39KAE6o022796 for ; Sat, 9 Apr 2005 13:10:14 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j39KAEbO022795 for ietf-usefor-skb; Sat, 9 Apr 2005 13:10:14 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtp801.mail.ukl.yahoo.com (smtp801.mail.ukl.yahoo.com [217.12.12.138]) by above.proper.com (8.12.11/8.12.9) with SMTP id j39KADWn022779 for ; Sat, 9 Apr 2005 13:10:13 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from unknown (HELO host81-144-73-19.midband.mdip.bt.net) (ietf-usefor@imc.org@81.144.73.19 with poptime) by smtp801.mail.ukl.yahoo.com with SMTP; 9 Apr 2005 20:10:06 -0000 Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j39K9o229793 for ietf-usefor@imc.org; Sat, 9 Apr 2005 21:09:50 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20642 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Differences between RFC 2822 and Usefor Message-ID: X-Newsreader: NN version 6.5.2 (NOV) Date: Sat, 9 Apr 2005 20:00:42 GMT Lines: 318 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: I published a list of differences from RFC 2822 as they stood in the old draft-13 last July (and it has been on our web site ever since). Most of those differences were agreed years ago when this WG was first formed and have been in the drafts ever since. I don't recall any disagreement with the list when I published it. There have been a few things changed since then (like the MIME-style parameters have gone), and also things are described somewhat differently in USEFOR, so I have updated it. The new texts follows (and it will shortly be put on the website), and after that there are the diffs from the previous one. Note that many of these agreed changes have not made it into the new USEFOR yet, and so I have marked the missing ones with "**". Differences between RFC 2822 and Usefor. ---------------------------------------- 08 April 2005 NOTE: Some items have been removed from the previous version, hence some discontinuities in the listing below. A '**' before an item indicates that the restriction in question has not yet been incorporated into the USEFOR document. 1. Differences enforced syntactically. -------------------------------------- 1.1 There is a REQUIRED SP (not even WSP) after the ':' in each header (even if the header has no content). 1.2 An MUST have at least one character. Thus the Subject (for example) of an article cannot be empty. 1.3**In the References-header, there is an obligatory CFWS (currently restricted to FWS by 2.1.1 below) in between each msg-id. NOTE: all the above were in RFC 1036, and the first is also required by NNTP. 1.5**Body lines are restricted to 998 characters plus CRLF (as in RFC 2822). However, all agents SHOULD, and relaying agents MUST, process lines of arbitrary length. 1.8 All the "obsolete" syntax in RFC 2822 is NOT REQUIRED (though it MAY be recognized). There are two small exceptions, listed below. 1.8.1 In the Date-header, the obsolete "UT" and "GMT" forms of zone MUST be recognized (because of their current widespread use), but MUST NOT be generated. 1.8.2**The (mis-named) "obsolete" syntax for phrase from RFC 2822, which allows for 'John D. Smith' to remain unquoted, is retained (but renamed as extended-phrase). NOTE: NNTP imposes similar restrictions. 1.11 Message-ID 1.11.1 No CFWS allowed (only FWS, which in practice means only non-folded WSP). 1.11.2 No quoted-pairs in msg-id, except for '\\' and '\"' in id-left '\[', '\]' and '\\' in id-right 1.11.3 No quoted-strings in id-left unless they contain one of the "specials". 1.11.4 No '>' anywhere within a msg-id, even within a quoted-string (to conform with RFC 1036). 1.11.5 All these restrictions on msg-id apply also to the References- and Supersedes-headers. 1.11.6 No control characters in msg-id. 2. Differences enforced by verbiage. ------------------------------------ 2.1 The following MUST be accepted, but SHOULD NOT be generated (yet): 2.1.1**Comments, except after a mailbox (where there is a now-deprecated convention for indicating the mailbox owner), or at the end of a date-time (which conventionally indicates the timezone). They are, however, freely allowed in headers that are newly defined in Usefor. 2.1.2**Extended-phrases (see 1.8.2 above). 2.2 The content of the first line of a header MUST NOT consist of WSP only (though such SHOULD be accepted). Observe that continuation lines of headers also MUST NOT consist of WSP only, as in RFC 2822. 2.3 Headers with empty content are deprecated (but if present that SP after the ':" is still required). 2.4 All agents MUST support header lines up to 998 octets, but there is no RECOMENDED limit of 78 characters as in RFC 2822. There is mention of a purely advisory limit of 79 (with a reference to USEAGE). 2.5 Relaying agents MUST NOT refold headers in transit. 2.6 There must not be more than one header with a given header-name, except where explicitly sanctioned by the appropriate standard. In particular, there MUST NOT be more than one Keywords-header. 2.7 The length of a msg-id MUST NOT exceed 250 octets. 2.8**The body of an article SHOULD NOT be empty. 2.9 (was 1.6) RFC 2047 and RFC 2231 are fully integrated into the Netnews. 2.10 (was 1.7) All the Content-* MIME headers are considered to be incorporated into Netnews and MUST be accepted in articles at, least to the extent required by RFC 2049. 2.11 A References header MUST be provided for followups (as opposed to SHOULD be provided for replies in RFC 2822). 3. Rules specific to Netnews headers. ------------------------------------- 3.2 Comments (but not FWS) are forbidden in the Newsgroups-, Distribution- Path- and Followup-To-headers (also see 1.11.1 above for Message-ID). 3.3**WSP and folding in Newsgroup- and Followup-To-headers MUST be accepted, but SHOULD NOT be generated (yet). NOTE: The effect of all these differences still preserves the property that the articles that Usefor permits to be generated form a proper subset of the articles that are required to be acceptable to RFC 2822. *** rfc2822-diffs.old Thu Jul 1 23:24:57 2004 --- rfc2822-diffs.txt Fri Apr 8 22:05:05 2005 *************** *** 3,6 **** ! 01 July 2004 1. Differences enforced syntactically. --- 3,12 ---- ! 08 April 2005 + NOTE: Some items have been removed from the previous version, hence some + discontinuities in the listing below. + + A '**' before an item indicates that the restriction in question has not yet + been incorporated into the USEFOR document. + 1. Differences enforced syntactically. *************** *** 11,13 **** ! 1.2 An 'unstructured' MUST have at least one character. Thus the Subject (for example) of an article cannot be empty. --- 17,19 ---- ! 1.2 An MUST have at least one character. Thus the Subject (for example) of an article cannot be empty. *************** *** 14,16 **** ! 1.3 In the References-header, there is an obligatory CFWS (currently restricted to FWS by 2.1.1 below) in between each msg-id. --- 20,22 ---- ! 1.3**In the References-header, there is an obligatory CFWS (currently restricted to FWS by 2.1.1 below) in between each msg-id. *************** *** 20,26 **** ! 1.4 The allowed characters in a header-name are restricted to ALPHA, ! DIGIT and embedded '-'. However, agents SHOULD accept all printables ! except SP and ':'. ! ! 1.5 Body lines are restricted to 998 characters plus CRLF (as in RFC 2822). However, all agents SHOULD, and relaying agents MUST, process --- 26,28 ---- ! 1.5**Body lines are restricted to 998 characters plus CRLF (as in RFC 2822). However, all agents SHOULD, and relaying agents MUST, process *************** *** 28,38 **** - 1.6 RFC 2047 and RFC 2231 are fully integrated into the syntax. Thus - encoded-words are explicitly included within 'unstructured', - 'ccontent' and 'phrase'. Moreover, it is the RFC 2231 version of - encoded-word that is used. - - 1.7 All the Content-* MIME headers are considered to be incorporated into - the syntax (i.e. they are to be accepted in articles, though not all - of them are required to have their semantic intentions implemented). - 1.8 All the "obsolete" syntax in RFC 2822 is NOT REQUIRED (though it MAY --- 30,31 ---- *************** *** 44,46 **** ! 1.8.2 The (mis-named) "obsolete" syntax for phrase from RFC 2822, which allows for 'John D. Smith' to remain unquoted, is retained (but --- 37,39 ---- ! 1.8.2**The (mis-named) "obsolete" syntax for phrase from RFC 2822, which allows for 'John D. Smith' to remain unquoted, is retained (but *************** *** 68,70 **** --- 61,65 ---- + 1.11.6 No control characters in msg-id. + 2. Differences enforced by verbiage. *************** *** 74,83 **** ! 2.1.1 Comments, except after a mailbox (where there is a now-deprecated convention for indicating the mailbox owner), or at the end of a ! date-time (which conventionally indicates the timezone). ! BUG: they ought to be freely allowed in headers that are newly ! defined in Usefor (and indeed the text positively encourages them ! in some of those places). ! 2.1.2 Extended-phrases (see 1.8.2 above). --- 69,76 ---- ! 2.1.1**Comments, except after a mailbox (where there is a now-deprecated convention for indicating the mailbox owner), or at the end of a ! date-time (which conventionally indicates the timezone). They are, ! however, freely allowed in headers that are newly defined in Usefor. ! 2.1.2**Extended-phrases (see 1.8.2 above). *************** *** 88,91 **** 2.3 Headers with empty content are deprecated (but if present that SP ! after the ':" is still required). Injecting agents SHOULD delete such ! headers, but other agents MUST propagate them. --- 81,83 ---- 2.3 Headers with empty content are deprecated (but if present that SP ! after the ':" is still required). *************** *** 103,107 **** ! 2.8 The body of an article SHOULD NOT be empty. 3. Rules specific to Netnews headers. --- 95,107 ---- ! 2.8**The body of an article SHOULD NOT be empty. + 2.9 (was 1.6) RFC 2047 and RFC 2231 are fully integrated into the Netnews. + 2.10 (was 1.7) All the Content-* MIME headers are considered to be + incorporated into Netnews and MUST be accepted in articles at, least + to the extent required by RFC 2049. + + 2.11 A References header MUST be provided for followups (as opposed to SHOULD + be provided for replies in RFC 2822). + 3. Rules specific to Netnews headers. *************** *** 109,118 **** - 3.1 All structured headers have MIME-style extension-parameters, with - x-attributes or to be defined in future standards. Some have explicit - parameters defined in this standard. However, this does not apply to - headers which are taken from RFC 2822 or other mail standards, nor to - the Mail-Copies-To, Complaints-to and Supersedes-header defined in - this standard. Nevertheless, such parameters SHOULD be recognized - (and ignored) in all headers. - 3.2 Comments (but not FWS) are forbidden in the Newsgroups-, --- 109,110 ---- *************** *** 121,128 **** ! 3.3 The following MUST be accepted, but SHOULD NOT be generated (yet): - 3.3.1 MIME-style parameters in headers defined prior to this standard. - - 3.3.2 WSP and folding in Newsgroup- and Followup-To-headers. - NOTE: The effect of all these differences still preserves the property --- 113,117 ---- ! 3.3**WSP and folding in Newsgroup- and Followup-To-headers MUST be ! accepted, but SHOULD NOT be generated (yet). NOTE: The effect of all these differences still preserves the property -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j39KACH9022787 for ; Sat, 9 Apr 2005 13:10:12 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j39KACFG022786 for ietf-usefor-skb; Sat, 9 Apr 2005 13:10:12 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtp801.mail.ukl.yahoo.com (smtp801.mail.ukl.yahoo.com [217.12.12.138]) by above.proper.com (8.12.11/8.12.9) with SMTP id j39KABRT022778 for ; Sat, 9 Apr 2005 13:10:12 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from unknown (HELO host81-144-73-19.midband.mdip.bt.net) (ietf-usefor@imc.org@81.144.73.19 with poptime) by smtp801.mail.ukl.yahoo.com with SMTP; 9 Apr 2005 20:10:05 -0000 Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j39K9na29787 for ietf-usefor@imc.org; Sat, 9 Apr 2005 21:09:49 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20641 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: Babble from Frank Ellermann Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: <4255153E.3080500@isode.com> <4256206D.6512@xyzzy.claranet.de> <425756AE.2232@xyzzy.claranet.de> Date: Sat, 9 Apr 2005 19:33:52 GMT Lines: 48 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In <425756AE.2232@xyzzy.claranet.de> Frank Ellermann writes: >Charles Lindsey wrote: >>> it's a valid RfC 2822 mail header field. >> It is indeed, but since it is in breach of RFC 2047 >It shouldn't, I've sent it in two lines, 2nd line 76 characters: >Subject: > =?UTF-8*de-DE-1996?B?RG9uYXVkYW1wZnNjaGlmZmZhaHJ0c2thcGl0w6Ruc23DvHR6ZQ==?= >....5...10....5...20....5...30....5...40....5...50....5...60....5...70....5. Ah! It was all on one line when I received it, so I didn't understand what your problem was. >http://article.gmane.org/gmane.ietf.usenet.format:28154:raw >says that "something" between me and GmaNe "fixed" this, but >generally my UA is innocent if it comes to attempts of being >smart. >> Subject: =?UTF-8*de-DE-?B?RG9uYXVkYW1wZnNjaGlmZmZh?= >> =?UTF-8*de-DE-?B?aHJ0c2thcGl0w6Ruc23DvHR6ZQ==?= >Sure, you and me know this, but I'm talking about existing MUAs, >mail2news gateways, injection agents, servers, and newsreaders. Well if the sending agent cannot figure how to do it that way, then it will just have to leave it on one long line. Then it becomes the reading agent's problem :-) . >> BTW, your message achieved a Spammassassin score of 3 :-( . >1st UTF-8, 2nd B64, what was the 3rd ? SpamAssassin hits were FORGED_RCVD_HELO RCVD_BY_IP RCVD_NUMERIC_HELO SUBJECT_EXCESS_BASE64 SUBJECT_NOVOWEL -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j394uqGc031776 for ; Fri, 8 Apr 2005 21:56:52 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j394uqKx031775 for ietf-usefor-skb; Fri, 8 Apr 2005 21:56:52 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from ciao.gmane.org (main.gmane.org [80.91.229.2]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j394upSL031769 for ; Fri, 8 Apr 2005 21:56:51 -0700 (PDT) (envelope-from usenet-format@gmane.org) Received: from list by ciao.gmane.org with local (Exim 4.43) id 1DK7yz-0003nh-QC for ietf-usefor@imc.org; Sat, 09 Apr 2005 06:54:09 +0200 Received: from du-001-228.access.de.clara.net ([212.82.227.228]) by main.gmane.org with esmtp (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Sat, 09 Apr 2005 06:54:09 +0200 Received: from nobody by du-001-228.access.de.clara.net with local (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Sat, 09 Apr 2005 06:54:09 +0200 X-Injected-Via-Gmane: http://gmane.org/ To: ietf-usefor@imc.org From: Frank Ellermann Subject: Re: =?UTF-8*de-DE-1996?B?RG9uYXVkYW1wZnNjaGlmZmZhaHJ0c2thcGl0w6Ruc23DvHR6ZQ==?= Date: Sat, 09 Apr 2005 06:55:19 +0200 Organization: Lines: 9 Message-ID: <42576037.190F@xyzzy.claranet.de> References: <4255153E.3080500@isode.com> <4256206D.6512@xyzzy.claranet.de> <425756AE.2232@xyzzy.claranet.de> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Complaints-To: usenet@sea.gmane.org X-Gmane-NNTP-Posting-Host: du-001-228.access.de.clara.net X-Mailer: Mozilla 3.0 (OS/2; U) Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: I've just tested it with two news servers, and both rejected a "Subject:" SP CRLF SP "stuff" CRLF So "fixing" the FWS issue by removing the separate MUST about non-empty header body lines Usefor-03 would be a very bad idea. Maybe I should a post an I-D "Subject: Re: considered effective at least" mentioning this case. Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j394Oqd8030346 for ; Fri, 8 Apr 2005 21:24:52 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j394Oq04030345 for ietf-usefor-skb; Fri, 8 Apr 2005 21:24:52 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from ciao.gmane.org (main.gmane.org [80.91.229.2]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j394OnCL030339 for ; Fri, 8 Apr 2005 21:24:50 -0700 (PDT) (envelope-from usenet-format@gmane.org) Received: from list by ciao.gmane.org with local (Exim 4.43) id 1DK7Tz-0002KG-W0 for ietf-usefor@imc.org; Sat, 09 Apr 2005 06:22:08 +0200 Received: from du-001-228.access.de.clara.net ([212.82.227.228]) by main.gmane.org with esmtp (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Sat, 09 Apr 2005 06:22:07 +0200 Received: from nobody by du-001-228.access.de.clara.net with local (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Sat, 09 Apr 2005 06:22:07 +0200 X-Injected-Via-Gmane: http://gmane.org/ To: ietf-usefor@imc.org From: Frank Ellermann Subject: =?UTF-8*de-DE-1996?B?RG9uYXVkYW1wZnNjaGlmZmZhaHJ0c2thcGl0w6Ruc23DvHR6ZQ==?= Date: Sat, 09 Apr 2005 06:14:38 +0200 Organization: Lines: 25 Message-ID: <425756AE.2232@xyzzy.claranet.de> References: <4255153E.3080500@isode.com> <4256206D.6512@xyzzy.claranet.de> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Complaints-To: usenet@sea.gmane.org X-Gmane-NNTP-Posting-Host: du-001-228.access.de.clara.net X-Mailer: Mozilla 3.0 (OS/2; U) Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Charles Lindsey wrote: >> it's a valid RfC 2822 mail header field. > It is indeed, but since it is in breach of RFC 2047 It shouldn't, I've sent it in two lines, 2nd line 76 characters: Subject: =?UTF-8*de-DE-1996?B?RG9uYXVkYW1wZnNjaGlmZmZhaHJ0c2thcGl0w6Ruc23DvHR6ZQ==?= ....5...10....5...20....5...30....5...40....5...50....5...60....5...70....5. http://article.gmane.org/gmane.ietf.usenet.format:28154:raw says that "something" between me and GmaNe "fixed" this, but generally my UA is innocent if it comes to attempts of being smart. > Subject: =?UTF-8*de-DE-?B?RG9uYXVkYW1wZnNjaGlmZmZh?= > =?UTF-8*de-DE-?B?aHJ0c2thcGl0w6Ruc23DvHR6ZQ==?= Sure, you and me know this, but I'm talking about existing MUAs, mail2news gateways, injection agents, servers, and newsreaders. > BTW, your message achieved a Spammassassin score of 3 :-( . 1st UTF-8, 2nd B64, what was the 3rd ? Bye, Frank Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j392CYMg017602 for ; Fri, 8 Apr 2005 19:12:34 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j392CY3O017601 for ietf-usefor-skb; Fri, 8 Apr 2005 19:12:34 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtp809.mail.ukl.yahoo.com (smtp809.mail.ukl.yahoo.com [217.12.12.199]) by above.proper.com (8.12.11/8.12.9) with SMTP id j392CXwS017584 for ; Fri, 8 Apr 2005 19:12:33 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from unknown (HELO host81-144-66-5.midband.mdip.bt.net) (ietf-usefor@imc.org@81.144.66.5 with poptime) by smtp809.mail.ukl.yahoo.com with SMTP; 9 Apr 2005 02:12:26 -0000 Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j392CDr25532 for ietf-usefor@imc.org; Sat, 9 Apr 2005 03:12:13 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20638 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: Issues outstanding Message-ID: X-Newsreader: NN version 6.5.2 (NOV) References: <4255153E.3080500@isode.com> Date: Fri, 8 Apr 2005 21:32:34 GMT Lines: 188 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In <4255153E.3080500@isode.com> Alexey Melnikov writes: >Ken has just published a new revision of USEFOR (03). Some comments on >the changes and how they relate to the list of issues below: There is still an awful lot of stuff that is meant to be covered in USEFOR, but still isn't. But that is for another thread. >Charles Lindsey wrote: >>>We are coming to the point where there is little more that can be done on >>>the documents we are supposed to be producing without deciding how various >>>outstanding issues are to be resolved. >> >>1. Complaints-To >> >>I think the conclusion we reached on this was to have a >>'mail-complaints-to' parameter in the Injection-Info header with an >> for its parameter. And we decided not to have any provision >>for URLs at this time, though a url-complaints-to parameter could be added >>as a future extension if there was a demand for it. >> >>If that is agreed, then this issue is CLOSED, >> >This change is done as well as description of different parameters. Yes, but not quite as agreed, because the token he has given for the parameter is 'complaints-to', rather than 'mail-complaints-to' as we agreed. That will not satisfy the people who wanted to leave room for a 'url-complaints-to' parameter as a future extension. >>except for deciding whether >>multiple
s meant you were supposed to reply to ALL of them, or to >>ANY ONE of them. Input on this is still needed. >> >This minor issue is still open. As currently written, the of that parameter is an , which by default means "send to them all" (cf. the Reply-To header). >My personal opinion is that if there are multiple email addresses they >all should be treated as equal and mailing to ANY of them should suffice. That could be achieved by adding some wording to say so. I am easy either way, so we need to hear more opinions. >>2. Path header delimiters >> >> >This is still open. >[...] Shame! I though we had more-or less agreed on that one. OK, I shall start another thread to discuss that. >>3. Mail-Copies-To and Posted-And Mailed >> >>>1. Include them as in draft-13 >>>2a. Defer them to a future document (standards-track) >>>2b. Defer them to a future document (experimental) >>>3. Drop them entirely >> >This issue is closed now: the headers will not appear in the USEFOR >document. The choice between 2a/2b/3 is up to the WG. Actually, I have now come to the conclusion that this problem is far worse with mailing lists than it is on Usenet, so it might be better to fix it there, maybe based on Mail-Followups-To with some extra features to incorporate News. Indeed, there was some discussion on the ietf-822 list about Mail-Followups-To as a possible solution, with many in favour but two diehards implacably opposed (for completely opposite reasons). Ours is not the only list that suffers from long discussions with no decision at the end :-( . (Oddly, I noticed that John Stanley was in favour of Posted-And-Mailed - he is quite right, but if Mail- Copies-To has to wait, then Posted-And-Mailed must wait too.) I shall now remove all mention of both these headers from USEPRO. >>4. Terminology for followups. >> >>This one is still OPEN. There are two alternative texts in USEPRO, but the >>matching alternative texts for USEFOR are not in place yet (I hope Ken is >>working on them). So I am happy to let this one be for now. There is no >>technical issue involved - just a question of how to define things. >> >I don't believe that anything in the USEFOR should be changed, so this >issue concerns the USEPRO document at best. Could you please explain your reasoning here? ISTM that any header described in USEFOR needs the following information, as appropriate: 1. A brief statement of what the header is supposed to achieve. 2. Its syntax. 3. Its semantics (i.e. what information is conveyed by its various syntactic parts). 4. Any restrictions or requirements on when it is to be used (e.g. it is "mandatory", or it MUST/SHOULD [NOT] be present if such and such other circumstances pertain). Following #4, there has always been a statement like the following associated with this header: A followup MUST have a References-header, and an article that is not a followup MUST NOT have a References-header. It is particularly important to say that here, because it is a change from RFC 2822, where the word used is only SHOULD. Now the precise wording of that varies according to exactly how the term "followup" is defined (and that is what this issue is all about). But both sides to this argument are agreed that that "MUST" needs to be said, and this is (AFAICS) the only place where it could be said. Yes, USEPRO describes in detail how to construct this header in the particular case of responses/replies to earlier articles, but there are other applications for it also, for example multipart FAQs and message/partial (and again it is common ground that these are legitimate applications). >>5. Review Injection-Info syntax (this might be related to Complaints-To) >> >The updated USEFOR draft now includes description of different parameters. >If people want an alternative syntax, please speak up now! Well nobody has spoken on this for a long time. I think the important thing is that Russ declared that he could live with this syntax, and I think his opinion is important. But I think we also need some deprecatory remarks to discourage the wilder (and unnecessary) extremes of RFC 2231 (and please can we use that lovely word "gibbous" in them :-) ). It is nice to see all the other Injection-Info parameters fully described. I have a few niggles about the precise details, and there are issues regarding the way they have been introduced syntactically, but that too is for another thread. >>6. Remove filename parameter from the Archive header. >> >In the latest USEFOR draft the filename parameter was replaced by a >generic parameters. Fine! >>7. FWS issue in headers. >> >>Frank was very keen to introduce *FWS rather than *CFWS or *FWS in various >>headers to cope with the rule that folding should not result in empty lines, >>or even in lines with empty content. It was established, however, that the >>present verbiage covering this issue would still be needed because it was not >>possible to solve all such cases syntactically. I argued that there was no >>point in changing only those cases where it would work, thereby introducing >>differences from RFC 2822. Note that this issue involves no technical change - >>just the method of description. >> >>Frank received no other support, and I propose to do nothing. If Ken wants to >>make these changes to USEFOR, then so be it. I regard this one as CLOSED. >> >I tend to agree. >>8. Define a Message-ID compatible with NNTP, get rid of NO-WS-CTL. >> >>We agreed to get rid of NO-WS-CTL (it would have been incompatible with >>the new NNTP draft), but our Chair rules that further departures from RFC >>2822 were not to be allowed. So I think this is CLOSED. >NO-WS-CTL have been removed. Can people check that the new syntax is Ok? >I suspect that some minor issues raised by Frank are yet to be addressed. Yes, there is a placeholder "[[Adjacent dots should not be allowed]]" for Frank's problem, but we still need some syntax to plug in there. We also need some wording drawing attention to this extra departure from RFC 2822, and referring to [NNTP] which necessitated it. -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j392CW9M017594 for ; Fri, 8 Apr 2005 19:12:32 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j392CWUt017592 for ietf-usefor-skb; Fri, 8 Apr 2005 19:12:32 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from smtp809.mail.ukl.yahoo.com (smtp809.mail.ukl.yahoo.com [217.12.12.199]) by above.proper.com (8.12.11/8.12.9) with SMTP id j392CVxE017583 for ; Fri, 8 Apr 2005 19:12:32 -0700 (PDT) (envelope-from news@clerew.man.ac.uk) Received: from unknown (HELO host81-144-66-5.midband.mdip.bt.net) (ietf-usefor@imc.org@81.144.66.5 with poptime) by smtp809.mail.ukl.yahoo.com with SMTP; 9 Apr 2005 02:12:25 -0000 Received: (from news@localhost) by clerew.man.ac.uk (8.11.7+Sun/8.11.7) id j392CCE25526 for ietf-usefor@imc.org; Sat, 9 Apr 2005 03:12:12 +0100 (BST) To: ietf-usefor@imc.org Xref: clerew local.usefor:20637 Newsgroups: local.usefor Path: clerew!chl From: "Charles Lindsey" Subject: Re: =?UTF-8*de-DE-?B?RG9uYXVkYW1wZnNjaGlmZmZhaHJ0c2thcGl0w6Ruc23DvHR6ZQ==?= Content-Type: text/plain; charset=iso-8859-1 Message-ID: Content-Transfer-Encoding: 8bit X-Newsreader: NN version 6.5.2 (NOV) References: <4255153E.3080500@isode.com> <4256206D.6512@xyzzy.claranet.de> Mime-Version: 1.0 Date: Fri, 8 Apr 2005 20:11:05 GMT Lines: 55 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: In <4256206D.6512@xyzzy.claranet.de> Frank Ellermann writes: >Alexey Melnikov wrote: >>> Frank received no other support, and I propose to do >>> nothing. If Ken wants to make these changes to USEFOR, then >>> so be it. I regard this one as CLOSED. >> I tend to agree. >One last attempt, the subject of this article is a German fun >word: Donaudampfschifffahrtskapitänsmütze. The 3 "fff" are a >case of de-DE-1996, I use RfC 2231 and one B64 UTF-8 word. >It is too long for RfC2047, therefore it's folded with a FWS. >This violates a MUST in Usefor-03, but it's a valid RfC 2822 >mail header field. It is indeed, but since it is in breach of RFC 2047, a conforming reading agent SHOULD render it exactly as received ('=?...?=' and all). However, some kind agents _might_ try to decode it for you. The proper way to create that header is as: Subject: =?UTF-8*de-DE-?B?RG9uYXVkYW1wZnNjaGlmZmZh?= =?UTF-8*de-DE-?B?aHJ0c2thcGl0w6Ruc23DvHR6ZQ==?= That works because, if you have two encoded words with nothing but FWS between them, then the decoder is required to ignore the FWS. And you have to be careful where you make the split, because each encoded-word must encode an integral number of octets (so after a multiple of 4 for base64, and watch out for '=xy' in Q-P). And on top of that you must not split any multi-octet character between the encoded words (which could be tricky in UTF-8, where the number of octets per character is variable). But any decent MUA ought to be able to do it. I tried it in Opera, and it produced: Subject: =?utf-8?Q?Donaudampfschifffahrtskapit?= =?utf-8?Q?=C3=A4nsm=C3=BCtze?= OK, it used Q-P rather than Base 64, but it was quite happy to display the original word without any spaces in the middle when read back. BTW, your message achieved a Spammassassin score of 3 :-( . -- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5 Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j386FFAu004709 for ; Thu, 7 Apr 2005 23:15:15 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j386FF1P004708 for ietf-usefor-skb; Thu, 7 Apr 2005 23:15:15 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from ciao.gmane.org (main.gmane.org [80.91.229.2]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j386FDj7004686 for ; Thu, 7 Apr 2005 23:15:14 -0700 (PDT) (envelope-from usenet-format@gmane.org) Received: from list by ciao.gmane.org with local (Exim 4.43) id 1DJmjT-0000VH-4G for ietf-usefor@imc.org; Fri, 08 Apr 2005 08:12:43 +0200 Received: from 212.82.251.127 ([212.82.251.127]) by main.gmane.org with esmtp (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Fri, 08 Apr 2005 08:12:43 +0200 Received: from nobody by 212.82.251.127 with local (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Fri, 08 Apr 2005 08:12:43 +0200 X-Injected-Via-Gmane: http://gmane.org/ To: ietf-usefor@imc.org From: Frank Ellermann Subject: =?UTF-8*de-DE-1996?B?RG9uYXVkYW1wZnNjaGlmZmZhaHJ0c2thcGl0w6Ruc23DvHR6ZQ==?= Date: Fri, 08 Apr 2005 08:10:53 +0200 Organization: Lines: 23 Message-ID: <4256206D.6512@xyzzy.claranet.de> References: <4255153E.3080500@isode.com> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: quoted-printable X-Complaints-To: usenet@sea.gmane.org X-Gmane-NNTP-Posting-Host: 212.82.251.127 X-Mailer: Mozilla 3.0 (OS/2; U) Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Alexey Melnikov wrote: >> Frank received no other support, and I propose to do >> nothing. If Ken wants to make these changes to USEFOR, then >> so be it. I regard this one as CLOSED. > I tend to agree. One last attempt, the subject of this article is a German fun word: Donaudampfschifffahrtskapit=E4nsm=FCtze. The 3 "fff" are a case of de-DE-1996, I use RfC 2231 and one B64 UTF-8 word. It is too long for RfC2047, therefore it's folded with a FWS. This violates a MUST in Usefor-03, but it's a valid RfC 2822 mail header field. So now what, shoot my UA (in theory, in practice I edited the subject manually) ? Let some "injection agent" fix it ? Or reject it ? What about mail2news gateways ? = Bye, Frank Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j37BBewT032549 for ; Thu, 7 Apr 2005 04:11:40 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j37BBeGO032548 for ietf-usefor-skb; Thu, 7 Apr 2005 04:11:40 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from rufus.isode.com (rufus.isode.com [62.3.217.251]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j37BBcKY032526 for ; Thu, 7 Apr 2005 04:11:39 -0700 (PDT) (envelope-from alexey.melnikov-usefor@isode.com) Received: from [192.168.0.7] ([62.3.217.253]) by rufus.isode.com via TCP (internal) with ESMTPA; Thu, 7 Apr 2005 12:11:33 +0100 Message-ID: <4255153E.3080500@isode.com> Date: Thu, 07 Apr 2005 12:10:54 +0100 From: Alexey Melnikov User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.7.2) Gecko/20040804 Netscape/7.2 (ax) X-Accept-Language: en-us, en MIME-Version: 1.0 To: ietf-usefor@imc.org Subject: Re: Issues outstanding References: In-Reply-To: Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: Ken has just published a new revision of USEFOR (03). Some comments on the changes and how they relate to the list of issues below: Charles Lindsey wrote: >Some while back I posted a list of issues (and our Chair added some more). >We have now reached the point where we cannot continue working on our >drafts until these are resolved. > >So here is the list again, with my comments on where we are at on each >one. SO PLEASE CAN WE HAVE INPUT ON THESE, especially on the ones which >still appear to be OPEN? > > >>We are coming to the point where there is little more that can be done on >>the documents we are supposed to be producing without deciding how various >>outstanding issues are to be resolved. >> >> > >1. Complaints-To > > > >>I published a list of 4 options (and invited other options) in >>http://www.imc.org/ietf-usefor/mail-archive/msg00151.html. Only three >>people have expressed any preference amongst them. I think #4 is dead, and >>#2 is the one most people could live with (but on such a small sample, >>that is hardly meaningful). >> >> >>#2 was to do it in Injection-Info rather than in a Complaints-To header, >>and to provide only a mail-complaints-to parameter (which would leave open >>the option to provide a separate url-complaints parameter as a future >>extension). >> > >I think the conclusion we reached on this was to have a >'mail-complaints-to' parameter in the Injection-Info header with an > for its parameter. And we decided not to have any provision >for URLs at this time, though a url-complaints-to parameter could be added >as a future extension if there was a demand for it. > >If that is agreed, then this issue is CLOSED, > This change is done as well as description of different parameters. >except for deciding whether >multiple
s meant you were supposed to reply to ALL of them, or to >ANY ONE of them. Input on this is still needed. > This minor issue is still open. My personal opinion is that if there are multiple email addresses they all should be treated as equal and mailing to ANY of them should suffice. > >2. Path header delimiters > > This is still open. [...] > > >3. Mail-Copies-To and Posted-And Mailed > > > >>Available options appear to be: >> >>1. Include them as in draft-13 >>2a. Defer them to a future document (standards-track) >>2b. Defer them to a future document (experimental) >>3. Drop them entirely >> >> >>Earlier discussions were inconclusive. I gather our Chair prefers #2 (a or >>b), but he has made no definitive pronouncement. >> >> > >I think all that the discussion established was that it was as much effort to >remove them (from USEPRO) as to add them (to USEFOR). It is still not clear >(to me) what the objection to keeping them is, and I see no merit at all in >#2b (since these headers are in moderately common use, and the "experiment" >has, in effect, been done). > >So this issue is still OPEN. > This issue is closed now: the headers will not appear in the USEFOR document. The choice between 2a/2b/3 is up to the WG. > > >4. Terminology for followups. > > > >>1. A followup is a response, and MUST have a References header. A part of >> a multi-part FAQ (or anything similar) is not a followup, but it MAY >> nevertheless have a References header. >> >>2. A followup is a response, or a part of a multi-part FAQ (or anything >> similar). A followup MUST have a References header, and anything else >> MUST NOT have one. >> >>It has been established that there is no technical difference between >>these formulations. It is just a matter of wording, so a simple majority >>for one of the other should settle the matter. >> >>There are alternative definitions in USEPRO, but no corresponding wordings >>for the References header in USEFOR yet, so maybe we should wait until >>there are. >> >> > >This one is still OPEN. There are two alternative texts in USEPRO, but the >matching alternative texts for USEFOR are not in place yet (I hope Ken is >working on them). So I am happy to let this one be for now. There is no >technical issue involved - just a question of how to define things. > I don't believe that anything in the USEFOR should be changed, so this issue concerns the USEPRO document at best. > > >5. Review Injection-Info syntax (this might be related to Complaints-To) > >I invited proposals from anyone who wanted to pursue this. I received none, so >I think this one is CLOSED. > The updated USEFOR draft now includes description of different parameters. If people want an alternative syntax, please speak up now! > >We all agree that RFC 2231 is ugly, but most of it is quite unnecessary in >Netnews. I would be happy for this to be pointed out in USEFOR with suitably >discouraging wording. > >6. Remove filename parameter from the Archive header. > >I think we concluded that the filename-parameter (and perhaps other >parameters) might well be useful in the future, but there was no need to >define them now. Therefore, we should just keep provision for MIME-style >parameters in this header (so software would be required to ignore such >parameters for now), but leave the definition of any actual parameters for >future extensions. > In the latest USEFOR draft the filename parameter was replaced by a generic parameters. > > >7. FWS issue in headers. > >Frank was very keen to introduce *FWS rather than *CFWS or *FWS in various >headers to cope with the rule that folding should not result in empty lines, >or even in lines with empty content. It was established, however, that the >present verbiage covering this issue would still be needed because it was not >possible to solve all such cases syntactically. I argued that there was no >point in changing only those cases where it would work, thereby introducing >differences from RFC 2822. Note that this issue involves no technical change - >just the method of description. > >Frank received no other support, and I propose to do nothing. If Ken wants to >make these changes to USEFOR, then so be it. I regard this one as CLOSED. > I tend to agree. > > >8. Define a Message-ID compatible with NNTP, get rid of NO-WS-CTL. > >We agreed to get rid of NO-WS-CTL (it would have been incompatible with >the new NNTP draft), but our Chair rules that further departures from RFC >2822 were not to be allowed. So I think this is CLOSED. > > > > NO-WS-CTL have been removed. Can people check that the new syntax is Ok? I suspect that some minor issues raised by Frank are yet to be addressed. > So could people who disagree with the ones I have marked CLOSED please >speak up, and otherwise will our Chair please confirm that they are >CLOSED. > >And please may we have discussion of the ones still OPEN, especially the >Path header one. > > > Alexey Received: from above.proper.com (localhost.vpnc.org [127.0.0.1]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j36JZWPx053318 for ; Wed, 6 Apr 2005 12:35:32 -0700 (PDT) (envelope-from owner-ietf-usefor@mail.imc.org) Received: (from majordom@localhost) by above.proper.com (8.12.11/8.12.9/Submit) id j36JZWbw053317 for ietf-usefor-skb; Wed, 6 Apr 2005 12:35:32 -0700 (PDT) X-Authentication-Warning: above.proper.com: majordom set sender to owner-ietf-usefor@mail.imc.org using -f Received: from ietf.org (odin.ietf.org [132.151.1.176]) by above.proper.com (8.12.11/8.12.9) with ESMTP id j36JZUaM053311 for ; Wed, 6 Apr 2005 12:35:31 -0700 (PDT) (envelope-from dinaras@cnri.reston.va.us) Received: from CNRI.Reston.VA.US (localhost [127.0.0.1]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id PAA20760; Wed, 6 Apr 2005 15:35:27 -0400 (EDT) Message-Id: <200504061935.PAA20760@ietf.org> Mime-Version: 1.0 Content-Type: Multipart/Mixed; Boundary="NextPart" To: i-d-announce@ietf.org Cc: ietf-usefor@imc.org From: Internet-Drafts@ietf.org Subject: I-D ACTION:draft-ietf-usefor-usefor-03.txt Date: Wed, 06 Apr 2005 15:35:27 -0400 Sender: owner-ietf-usefor@mail.imc.org Precedence: bulk List-Archive: List-Unsubscribe: List-ID: --NextPart A New Internet-Draft is available from the on-line Internet-Drafts directories. This draft is a work item of the Usenet Article Standard Update Working Group of the IETF. Title : News Article Format Author(s) : C. Lindsey, et al. Filename : draft-ietf-usefor-usefor-03.txt Pages : 31 Date : 2005-4-6 This document specifies the syntax of network news (Netnews) articles in the context of the 'Internet Message Format' (RFC 2822) and 'Multipurpose Internet Mail Extensions (MIME)' (RFC 2045). This document supersedes RFC 1036, updating it to reflect current practice and incorporating incremental changes specified in other documents. A URL for this Internet-Draft is: http://www.ietf.org/internet-drafts/draft-ietf-usefor-usefor-03.txt To remove yourself from the I-D Announcement list, send a message to i-d-announce-request@ietf.org with the word unsubscribe in the body of the message. You can also visit https://www1.ietf.org/mailman/listinfo/I-D-announce to change your subscription settings. Internet-Drafts are also available by anonymous FTP. Login with the username "anonymous" and a password of your e-mail address. After logging in, type "cd internet-drafts" and then "get draft-ietf-usefor-usefor-03.txt". A list of Internet-Drafts directories can be found in http://www.ietf.org/shadow.html or ftp://ftp.ietf.org/ietf/1shadow-sites.txt Internet-Drafts can also be obtained by e-mail. Send a message to: mailserv@ietf.org. In the body type: "FILE /internet-drafts/draft-ietf-usefor-usefor-03.txt". NOTE: The mail server at ietf.org can return the document in MIME-encoded form by using the "mpack" utility. To use this feature, insert the command "ENCODING mime" before the "FILE" command. To decode the response(s), you will need "munpack" or a MIME-compliant mail reader. Different MIME-compliant mail readers exhibit different behavior, especially when dealing with "multipart" MIME messages (i.e. documents which have been split up into multiple messages), so check your local documentation on how to manipulate these messages. Below is the data which will enable a MIME compliant mail reader implementation to automatically retrieve the ASCII version of the Internet-Draft. --NextPart Content-Type: Multipart/Alternative; Boundary="OtherAccess" --OtherAccess Content-Type: Message/External-body; access-type="mail-server"; server="mailserv@ietf.org" Content-Type: text/plain Content-ID: <2005-4-6160717.I-D@ietf.org> ENCODING mime FILE /internet-drafts/draft-ietf-usefor-usefor-03.txt --OtherAccess Content-Type: Message/External-body; name="draft-ietf-usefor-usefor-03.txt"; site="ftp.ietf.org"; access-type="anon-ftp"; directory="internet-drafts" Content-Type: text/plain Content-ID: <2005-4-6160717.I-D@ietf.org> --OtherAccess-- --NextPart--