From: swb@employees.org (Scott W Brim)
Date: Tue, 27 Aug 2002 13:19:50 -0400
Subject: [xml2rfc] Re: nsgmls?
In-Reply-To: <008c01c24ded$bc25ffb0$6500000a@dev0196>; from mrose@dbc.mtview.ca.us on Tue, Aug 27, 2002 at 10:18:16AM -0700
References: <E17jf7k-000HpR-00@rip.psg.com> <r01050300-1015-E667C00DB9C811D6902C00039300CF5C@[193.157.66.10]> <20020827105931.A12703@localhost.localdomain> <008c01c24ded$bc25ffb0$6500000a@dev0196>
Message-ID: <20020827131950.A13861@localhost.localdomain>

On Tue, Aug 27, 2002 10:18:16AM -0700, Marshall Rose allegedly wrote:
> > It seems, though, that rfc262.dtd isn't formal enough for the sgml tools
> > I have here.  So it goes.  It's been interesting.
> 
> scott - this sort of troubles me in the sense that, afaik, rfc2629.dtd is a
> perfectly fine, conformant dtd.
> 
> what did nsgmls complain about, exactly?

I'll set it all up again and pass you the messages, offline.  If you
think it should work, I'm probably doing something wrong.


From: mrose@dbc.mtview.ca.us (Marshall Rose)
Date: Tue, 27 Aug 2002 10:18:16 -0700
Subject: [xml2rfc] Re: nsgmls?
References: <E17jf7k-000HpR-00@rip.psg.com> <r01050300-1015-E667C00DB9C811D6902C00039300CF5C@[193.157.66.10]> <20020827105931.A12703@localhost.localdomain>
Message-ID: <008c01c24ded$bc25ffb0$6500000a@dev0196>

> It seems, though, that rfc262.dtd isn't formal enough for the sgml tools
> I have here.  So it goes.  It's been interesting.

scott - this sort of troubles me in the sense that, afaik, rfc2629.dtd is a
perfectly fine, conformant dtd.

what did nsgmls complain about, exactly?

thanks,

/mtr




From: swb@employees.org (Scott W Brim)
Date: Tue, 27 Aug 2002 10:59:31 -0400
Subject: [xml2rfc] Re: nsgmls?
In-Reply-To: <r01050300-1015-E667C00DB9C811D6902C00039300CF5C@[193.157.66.10]>; from link@pobox.com on Tue, Aug 27, 2002 at 04:25:56PM +0200
References: <E17jf7k-000HpR-00@rip.psg.com> <r01050300-1015-E667C00DB9C811D6902C00039300CF5C@[193.157.66.10]>
Message-ID: <20020827105931.A12703@localhost.localdomain>

On Tue, Aug 27, 2002 04:25:56PM +0200, Terje Bless allegedly wrote:
> Find the catalog file it uses and associate the FPI or SI with
> /the/path/to/rfc2629.dtd, I would assume. On Red Hat Linux, the catalog and
> DTDs are kept in /usr/share/sgml/ somewhere and are named either "catalog"
> or "*.soc". The catalog should probably have an entry for:
> 
>   SYSTEM "rfc2629.dtd" "/real/path/to/rfc2629.dtd"
> 
> or the '<!DOCTYPE rfc SYSTEM "rfc2629.dtd">' bit could reference the file
> directly (iow, include a full path to rfc2629.dtd).
> 
> 
> With the caveat that I'm not familiar with 2629 so it may have
> peculiarities that I'm not aware of which makes this not work.

Thanks for the clues.  On RH 7.3, the main catalog is /etc/sgml/catalog.
Subsidiary catalogs are kept there and in subdirectories under
/usr/share/sgml.  DTDs are kept in subdirectories of /usr/share/sgml.
It seems, though, that rfc262.dtd isn't formal enough for the sgml tools
I have here.  So it goes.  It's been interesting.

...Scott


From: link@pobox.com (Terje Bless)
Date: Tue, 27 Aug 2002 16:25:56 +0200
Subject: [xml2rfc] Re: nsgmls?
In-Reply-To: <E17jf7k-000HpR-00@rip.psg.com>
Message-ID: <r01050300-1015-E667C00DB9C811D6902C00039300CF5C@[193.157.66.10]>

Randy Bush <randy@psg.com> wrote:

>i have become a fan of tidy, on freebsd /usr/ports/www/tidy, it
>diagnoses and cleans up html and xml, and prettyprints it as well.

But it's a lint and not a validator; especially not in the SGML/XML sense.


>Emacs psgml mode wants to use nsgmls as an XML verifier (on linux).
>[...] what do I have to do to make nsgmls aware of rfc2629.dtd?

Find the catalog file it uses and associate the FPI or SI with
/the/path/to/rfc2629.dtd, I would assume. On Red Hat Linux, the catalog and
DTDs are kept in /usr/share/sgml/ somewhere and are named either "catalog"
or "*.soc". The catalog should probably have an entry for:

  SYSTEM "rfc2629.dtd" "/real/path/to/rfc2629.dtd"

or the '<!DOCTYPE rfc SYSTEM "rfc2629.dtd">' bit could reference the file
directly (iow, include a full path to rfc2629.dtd).


With the caveat that I'm not familiar with 2629 so it may have
peculiarities that I'm not aware of which makes this not work.


-- 
"Hath no man's dagger here a point for me?"   - Leonato, Governor of Messina.
                   See Project Gutenberg <URL:http://promo.net/pg/> for more.


From: randy@psg.com (Randy Bush)
Date: Tue, 27 Aug 2002 05:07:09 -0700
Subject: [xml2rfc] nsgmls?
References: <20020827060102.A5033@localhost.localdomain>
Message-ID: <E17jf7k-000HpR-00@rip.psg.com>

i have become a fan of tidy, on freebsd /usr/ports/www/tidy, it
diagnoses and cleans up html and xml, and prettyprints it as
well.

randy



From: swb@employees.org (Scott W Brim)
Date: Tue, 27 Aug 2002 06:01:02 -0400
Subject: [xml2rfc] nsgmls?
Message-ID: <20020827060102.A5033@localhost.localdomain>

Emacs psgml mode wants to use nsgmls as an XML verifier (on linux).  All
I know about XML I learned from Marshall -- what do I have to do to make
nsgmls aware of rfc2629.dtd?  Alternatively, what do you use to easily
find problems with your XML on *n*x?  I'd even be willing to try out a
different editing environment.

Thanks ... Scott


From: mcr@sandelman.ottawa.on.ca (Michael Richardson)
Date: Mon, 26 Aug 2002 18:49:54 -0400
Subject: [xml2rfc] references
In-Reply-To: Your message of "Mon, 26 Aug 2002 12:13:53 PDT." <20020826121353.303f6a5d.mrose@dbc.mtview.ca.us>
Message-ID: <200208262249.g7QMnsvA016351@marajade.sandelman.ottawa.on.ca>

-----BEGIN PGP SIGNED MESSAGE-----


>>>>> "Marshall" == Marshall Rose <mrose@dbc.mtview.ca.us> writes:
    >> where killtodeepcontents.pl is on the above web site as well, and
    >> rfc_pg.c is the standard thing used for processing nroff output.

    Marshall> hi. did you know that the current xml2rfc has a tocdepth PI,
    Marshall> e.g.,
    
    Marshall> <?rfc tocdepth='2' ?>
    
  I did not.
  I'll make use of it.

]       ON HUMILITY: to err is human. To moo, bovine.           |  firewalls  [
]   Michael Richardson, Sandelman Software Works, Ottawa, ON    |net architect[
] mcr@sandelman.ottawa.on.ca http://www.sandelman.ottawa.on.ca/ |device driver[
] panic("Just another Debian GNU/Linux using, kernel hacking, security guy"); [


-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.7 (GNU/Linux)
Comment: Finger me for keys

iQCVAwUBPWqwkIqHRg3pndX9AQG5IAQAybjvYuRutak2dLS6wGoxtr1IQ8+ODcb+
1UDWPjYf0slkVyBFDR3XfKrsceofPC8PuLYZdIJ/rQUBiQGzCN0PU4vnIEESmrQm
B/E8oX2y6iDYoP+0DQLji0lbmzw7yTI+ClTmUO8fB0McAGG8VjFULfzaf6uIpRBl
orFXljkLShQ=
=KZnS
-----END PGP SIGNATURE-----


From: randy@psg.com (Randy Bush)
Date: Mon, 26 Aug 2002 13:25:21 -0700
Subject: [xml2rfc] double space after e.g. i.e.
References: <E17jBpA-0005x3-00@rip.psg.com> <20020826093432.108af095.mrose+internet.xml2rfc@dbc.mtview.ca.us>
Message-ID: <E17jQQK-000GUN-00@rip.psg.com>

>>    E.g.  a flozzie may floozie ...
>>        ^^
> well, this is a hard one to get right.

agreed.  fyi, emacs M-Q seems to retain whatever one puts after a
period.

> the original version of xml2rfc just put one space after "."
> because that worked for everything. then it was noted that the
> rfc editor prefers that sentences have two spaces after them.

as do i, so i really appreciate xml2rfc doing the double space.

> to avoid the problem, i do this "e.g.," "cf.," etc.

hmmm.  that's proper.

> if someone has a counter-example, let me know.

well, there is the case of ... in the middle of a sentence.

> the alternative is to add some rather odd heuristics, e.g.,
> 	if you see <not-a-letter><letter>". " then it should have one space
> 	if you see "cf. " then it should have one space
> 	otherwise, if you see ". ", it should have two spaces

see elipses above.

i promise to use the comma and shut up. :-)

randy



From: mrose@dbc.mtview.ca.us (Marshall Rose)
Date: Mon, 26 Aug 2002 12:13:53 -0700
Subject: [xml2rfc] references
In-Reply-To: <200208261822.g7QIMqtu015096@marajade.sandelman.ottawa.on.ca>
References: <20020826080628.A25398@localhost.localdomain> <200208261822.g7QIMqtu015096@marajade.sandelman.ottawa.on.ca>
Message-ID: <20020826121353.303f6a5d.mrose@dbc.mtview.ca.us>

> where killtodeepcontents.pl is on the above web site as well, and rfc_pg.c
> is the standard thing used for processing nroff output.

hi. did you know that the current xml2rfc has a tocdepth PI, e.g.,
    
    <?rfc tocdepth='2' ?>
    
/mtr


From: mcr@sandelman.ottawa.on.ca (Michael Richardson)
Date: Mon, 26 Aug 2002 14:22:50 -0400
Subject: [xml2rfc] references
In-Reply-To: Your message of "Mon, 26 Aug 2002 08:06:28 EDT." <20020826080628.A25398@localhost.localdomain>
Message-ID: <200208261822.g7QIMqtu015096@marajade.sandelman.ottawa.on.ca>

 
If there is any value, I've been posting the XML for
the document I've been working on at
http://www.sandelman.ca/SSW/freeswan/oeid/. It is not a short
document, and it may serve as a good reference for others.

I produce the .txt files with:

rfc_pg: utils/rfc_pg.c

draft-richardson-ipsec-opportunistic.txt: draft-richardson-ipsec-opportunistic.nr rfc_pg
        perl utils/killtoodeepcontents.pl draft-richardson-ipsec-opportunistic.nr \
        | nroff -ms \
        | sed -e 's/FORMFEED\[Page/        ^L\[Page/' | ./rfc_pg -n5 >draft-richardson-ipsec-opportunistic.txt

draft-richardson-ipsec-opportunistic.nr: src/draft-richardson-ipsec-opportunistic.xml
        XML_LIBRARY=$(XML_LIBRARY):./src xml2rfc xml2nroff $? $@

src/draft-richardson-ipsec-opportunistic.html: src/draft-richardson-ipsec-opportunistic.xml
        XML_LIBRARY=$(XML_LIBRARY):./src xml2rfc xml2html $? $@


where killtodeepcontents.pl is on the above web site as well, and rfc_pg.c
is the standard thing used for processing nroff output.

]       ON HUMILITY: to err is human. To moo, bovine.           |  firewalls  [
]   Michael Richardson, Sandelman Software Works, Ottawa, ON    |net architect[
] mcr@sandelman.ottawa.on.ca http://www.sandelman.ottawa.on.ca/ |device driver[
] panic("Just another Debian GNU/Linux using, kernel hacking, security guy"); [


From: mrose+internet.xml2rfc@dbc.mtview.ca.us (Marshall Rose)
Date: Mon, 26 Aug 2002 09:34:32 -0700
Subject: [xml2rfc] double space after e.g. i.e.
In-Reply-To: <E17jBpA-0005x3-00@rip.psg.com>
References: <E17jBpA-0005x3-00@rip.psg.com>
Message-ID: <20020826093432.108af095.mrose+internet.xml2rfc@dbc.mtview.ca.us>

> xml2rfc seems to double space after abbreviations.  in noticed this
> in the following result
> 
>    E.g.  a flozzie may floozie ...
>        ^^

well, this is a hard one to get right.

the original version of xml2rfc just put one space after "." because that worked for everything. then it was noted that the rfc editor prefers that sentences have two spaces after them.

to avoid the problem, i do this "e.g.," "cf.," etc. i have yet to meet an abbreviation ending with a period that can't also take a comma.

if someone has a counter-example, let me know.

the alternative is to add some rather odd heuristics, e.g.,

	if you see <not-a-letter><letter>". " then it should have one space
	if you see "cf. " then it should have one space
	otherwise, if you see ". ", it should have two spaces

/mtr


From: mrose+internet.xml2rfc@dbc.mtview.ca.us (Marshall Rose)
Date: Mon, 26 Aug 2002 08:30:28 -0700
Subject: [xml2rfc] references
In-Reply-To: <E17jIa1-000K0l-00@rip.psg.com>
References: <E17jIa1-000K0l-00@rip.psg.com>
Message-ID: <20020826083028.20a7c9bb.mrose+internet.xml2rfc@dbc.mtview.ca.us>

> how do i sub-divide <references> into normative and non-normative?

there's an example of that under "Helpful Hints" on http://xml.resource.org/

/mtr


From: swb@employees.org (Scott W Brim)
Date: Mon, 26 Aug 2002 08:06:28 -0400
Subject: [xml2rfc] references
In-Reply-To: <E17jIa1-000K0l-00@rip.psg.com>; from randy@psg.com on Mon, Aug 26, 2002 at 05:02:49AM -0700
References: <E17jIa1-000K0l-00@rip.psg.com>
Message-ID: <20020826080628.A25398@localhost.localdomain>

On Mon, Aug 26, 2002 05:02:49AM -0700, Randy Bush allegedly wrote:
> how do i sub-divide <references> into normative and non-normative?

<references> can have title="Normative References", etc.


From: randy@psg.com (Randy Bush)
Date: Mon, 26 Aug 2002 05:02:49 -0700
Subject: [xml2rfc] references
Message-ID: <E17jIa1-000K0l-00@rip.psg.com>

how do i sub-divide <references> into normative and non-normative?

randy



From: randy@psg.com (Randy Bush)
Date: Mon, 26 Aug 2002 04:10:52 -0700
Subject: [xml2rfc] expiration date calculation
References: <E17jBw8-0006LR-00@rip.psg.com> <20020825223330.3a6dc0c4.mrose+internet.xml2rfc@dbc.mtview.ca.us> <E17jCcS-0008bF-00@rip.psg.com> <20020825224739.40460df8.mrose+internet.xml2rfc@dbc.mtview.ca.us>
Message-ID: <E17jHlj-000Hwu-00@rip.psg.com>

aha!

six months != year/2

now that i can understand.  thanks.

randy



From: mrose+internet.xml2rfc@dbc.mtview.ca.us (Marshall Rose)
Date: Sun, 25 Aug 2002 22:47:39 -0700
Subject: [xml2rfc] expiration date calculation
In-Reply-To: <E17jCcS-0008bF-00@rip.psg.com>
References: <E17jBw8-0006LR-00@rip.psg.com> <20020825223330.3a6dc0c4.mrose+internet.xml2rfc@dbc.mtview.ca.us> <E17jCcS-0008bF-00@rip.psg.com>
Message-ID: <20020825224739.40460df8.mrose+internet.xml2rfc@dbc.mtview.ca.us>

> but feb 23rd is not six months later than august 25th.  feb 25th is.
> though i could see argument to use feb 24th, giving the document a
> six month life.  but feb 23rd i just don't understand.
> 
> btw, i freely admit this is picayune pedantry.  but it's the kind of
> thing a picayune and pedantic nerd notices.

the current algorithm is:
    
    start with the current time in seconds
    add 182*86400 seconds
    add 43200 seconds
    
that's half a year.
    
now it is true that half a year isn't six months...
    
/mtr


From: randy@psg.com (Randy Bush)
Date: Sun, 25 Aug 2002 22:40:56 -0700
Subject: [xml2rfc] expiration date calculation
References: <E17jBw8-0006LR-00@rip.psg.com> <20020825223330.3a6dc0c4.mrose+internet.xml2rfc@dbc.mtview.ca.us>
Message-ID: <E17jCcS-0008bF-00@rip.psg.com>

but feb 23rd is not six months later than august 25th.  feb 25th is.
though i could see argument to use feb 24th, giving the document a
six month life.  but feb 23rd i just don't understand.

btw, i freely admit this is picayune pedantry.  but it's the kind of
thing a picayune and pedantic nerd notices.

randy



From: mrose+internet.xml2rfc@dbc.mtview.ca.us (Marshall Rose)
Date: Sun, 25 Aug 2002 22:33:30 -0700
Subject: [xml2rfc] expiration date calculation
In-Reply-To: <E17jBw8-0006LR-00@rip.psg.com>
References: <E17jBw8-0006LR-00@rip.psg.com>
Message-ID: <20020825223330.3a6dc0c4.mrose+internet.xml2rfc@dbc.mtview.ca.us>

> i had been presuming that the expiration date was calculated
> cleverly to not occur on weekends or something.  but i decided
> to check when i noticed this document

okay, why do you care if it's a weekend or a holiday. it's just 6 months later, right?

/mtr


From: randy@psg.com (Randy Bush)
Date: Sun, 25 Aug 2002 21:57:13 -0700
Subject: [xml2rfc] expiration date calculation
Message-ID: <E17jBw8-0006LR-00@rip.psg.com>

i had been presuming that the expiration date was calculated
cleverly to not occur on weekends or something.  but i decided
to check when i noticed this document

Network Working Group                                 R. Bush
Internet-Draft                                            IIJ
Expires: February 23, 2003                    August 25, 2002

and that's a sunday!

   February 2003
Su Mo Tu We Th Fr Sa
                   1
 2  3  4  5  6  7  8
 9 10 11 12 13 14 15
16 17 18 19 20 21 22
23 24 25 26 27 28

could someone please explain?

randy



From: randy@psg.com (Randy Bush)
Date: Sun, 25 Aug 2002 21:50:00 -0700
Subject: [xml2rfc] double space after e.g. i.e.
Message-ID: <E17jBpA-0005x3-00@rip.psg.com>

xml2rfc seems to double space after abbreviations.  in noticed this
in the following result

   E.g.  a flozzie may floozie ...
       ^^

randy



From: randy@psg.com (Randy Bush)
Date: Wed, 21 Aug 2002 10:51:52 -0700
Subject: [xml2rfc] anchors away - the newless cluebie saga continues
References: <E17hYtx-000OT6-00@rip.psg.com> <20020821132822.A11344@localhost.localdomain>
Message-ID: <E17hZe4-000Phe-00@rip.psg.com>

>>     <reference anchor="RFC2065">
>> 	<?rfc include="reference.RFC.2065" ?>
>> 	</reference>
> the <reference anchor=... > and </reference> are already in the include
> file.  Just leave them out.

thanks!

randy



From: remco@vandemeent.net (Remco van de Meent)
Date: Wed, 21 Aug 2002 19:40:52 +0200
Subject: [xml2rfc] anchors away - the newless cluebie saga continues
In-Reply-To: <E17hYtx-000OT6-00@rip.psg.com>
References: <E17hYtx-000OT6-00@rip.psg.com>
Message-ID: <20020821174052.GA23363@frml.snt.utwente.nl>

Randy Bush wrote:
> in the <middle>, i have
> 
>     Secure DNS RFC <xref target="RFC2065"></xref>.
> 
> then, in the <back><references> i have
> 
>     <reference anchor="RFC2065">
> 	<?rfc include="reference.RFC.2065" ?>
> 	</reference>
> 
> and i get the diagnostic
> 
>     anchor attribute already in use: RFC2065

You can leave out both the <reference ...> and </reference> tags, as
they are already present in the included file. And that also solves
your problem :)


regards,
Remco.
-- 
:: remco@vandemeent.net
:: http://remco.vandemeent.net/


From: mrose+internet.xml2rfc@dbc.mtview.ca.us (Marshall Rose)
Date: Wed, 21 Aug 2002 10:32:06 -0700
Subject: [xml2rfc] anchors away - the newless cluebie saga continues
In-Reply-To: <E17hYtx-000OT6-00@rip.psg.com>
References: <E17hYtx-000OT6-00@rip.psg.com>
Message-ID: <20020821103206.1153cb0e.mrose+internet.xml2rfc@dbc.mtview.ca.us>

> in the <middle>, i have
> 
>     Secure DNS RFC <xref target="RFC2065"></xref>.
> 
> then, in the <back><references> i have
> 
>     <reference anchor="RFC2065">
> 	<?rfc include="reference.RFC.2065" ?>
> 	</reference>
> 
> and i get the diagnostic
> 
>     anchor attribute already in use: RFC2065

that's becasue the reference already defines the anchor. (hint: look at
the reference.RFC.2065.xml file!)
    
just have:
    
<rfc>
    ...
<back>
    <references>
        <?rfc include='reference.RFC.2065' ?>
    </references>
</back>
</rfc>

the design principles for 2629/xml2rfc include trying to limit the
amount of typein...
    
/mtr


From: mrose+internet.xml2rfc@dbc.mtview.ca.us (Marshall Rose)
Date: Wed, 21 Aug 2002 10:29:48 -0700
Subject: [xml2rfc] .xml2rfc.rc
In-Reply-To: <E17hYdx-000O0U-00@rip.psg.com>
References: <E17hL1M-0002qw-00@rip.psg.com> <20020821003856.7f01605f.mrose+internet.xml2rfc@dbc.mtview.ca.us> <E17hYdx-000O0U-00@rip.psg.com>
Message-ID: <20020821102948.7a8e3fef.mrose+internet.xml2rfc@dbc.mtview.ca.us>

> and if i want to add multiple?  follow with
>     append library $sep~/refs/rfcs

well, if you want to have multiple directories, then look at the
original three lines:
    
    foreach bibxmlD [lsort -dictionary [glob -nocomplain ~/rfcs/bibxml/*]] {
        append library $sep$bibxmlD
    }

this says add everything you find in ~/rfcs/bibxml/ to the search
path. (obviously, put only directories there...)
    
so, just change it to whatever you want.
    
/mtr


From: swb@employees.org (Scott W Brim)
Date: Wed, 21 Aug 2002 13:28:22 -0400
Subject: [xml2rfc] anchors away - the newless cluebie saga continues
In-Reply-To: <E17hYtx-000OT6-00@rip.psg.com>; from randy@psg.com on Wed, Aug 21, 2002 at 10:04:14AM -0700
References: <E17hYtx-000OT6-00@rip.psg.com>
Message-ID: <20020821132822.A11344@localhost.localdomain>

On Wed, Aug 21, 2002 10:04:14AM -0700, Randy Bush allegedly wrote:
> in the <middle>, i have
> 
>     Secure DNS RFC <xref target="RFC2065"></xref>.
> 
> then, in the <back><references> i have
> 
>     <reference anchor="RFC2065">
> 	<?rfc include="reference.RFC.2065" ?>
> 	</reference>
> 
> and i get the diagnostic
> 
>     anchor attribute already in use: RFC2065
> 
> randy

the <reference anchor=... > and </reference> are already in the include
file.  Just leave them out.


From: randy@psg.com (Randy Bush)
Date: Wed, 21 Aug 2002 10:04:14 -0700
Subject: [xml2rfc] anchors away - the newless cluebie saga continues
Message-ID: <E17hYtx-000OT6-00@rip.psg.com>

in the <middle>, i have

    Secure DNS RFC <xref target="RFC2065"></xref>.

then, in the <back><references> i have

    <reference anchor="RFC2065">
	<?rfc include="reference.RFC.2065" ?>
	</reference>

and i get the diagnostic

    anchor attribute already in use: RFC2065

randy



From: randy@psg.com (Randy Bush)
Date: Wed, 21 Aug 2002 09:47:41 -0700
Subject: [xml2rfc] .xml2rfc.rc
References: <E17hL1M-0002qw-00@rip.psg.com> <20020821003856.7f01605f.mrose+internet.xml2rfc@dbc.mtview.ca.us>
Message-ID: <E17hYdx-000O0U-00@rip.psg.com>

> actually, it looks for the .rc file in the same directory as the input
> file.

that's new and different.  thanks for that clue

> you don't source .xml2rfc.rc directly

i figured that out in the shower this morning.  no, i do not have a
waterproof laptop (which may explain a lot:-).

>> secondly, how can i hack it to suck up the ~/refs/ids/* directory
>> as well?
> since you're starting with the .rcfile that comes with the software,
> replace these three lines:
>     foreach bibxmlD [lsort -dictionary [glob -nocomplain ~/rfcs/bibxml/*]] {
>         append library $sep$bibxmlD
>     }
> with this line
>     append library $sep~/refs/ids
> that adds one directory, ~/refs/ids/ to your search path for include
> files.

and if i want to add multiple?  follow with
    append library $sep~/refs/rfcs

randy



From: mrose+internet.xml2rfc@dbc.mtview.ca.us (Marshall Rose)
Date: Wed, 21 Aug 2002 00:38:56 -0700
Subject: [xml2rfc] .xml2rfc.rc
In-Reply-To: <E17hL1M-0002qw-00@rip.psg.com>
References: <E17hL1M-0002qw-00@rip.psg.com>
Message-ID: <20020821003856.7f01605f.mrose+internet.xml2rfc@dbc.mtview.ca.us>

> first, will the xml2rfc tcl script always find ~/.xml2rfc.rc?  i
> presume so.

actually, it looks for the .rc file in the same directory as the input
file.

    
> but, if i manually run a slightly modified version of the one in
> the distribution (appended), i get
> 
>     % tclsh8.3 .xml2rfc.rc
>     can't read "inputD": no such variable
> 	while executing
>     "file nativename $inputD"
> 	invoked from within
>     "set nativeD [file nativename $inputD]"
> 	(file ".xml2rfc.rc" line 16)
    
well, to begin, you don't source .xml2rfc.rc directly, you let xml2rfc
do that at the right time (when it will set inputD)
    

    
> secondly, how can i hack it to suck up the ~/refs/ids/* directory
> as well?

since you're starting with the .rcfile that comes with the software,
replace these three lines:
    
    foreach bibxmlD [lsort -dictionary [glob -nocomplain ~/rfcs/bibxml/*]] {
        append library $sep$bibxmlD
    }

with this line
    
    append library $sep~/refs/ids
    
that adds one directory, ~/refs/ids/ to your search path for include
files.
    
/mtr


From: mrose+internet.xml2rfc@dbc.mtview.ca.us (Marshall Rose)
Date: Tue, 20 Aug 2002 23:47:54 -0700
Subject: [xml2rfc] url prob in ref when converting to html
In-Reply-To: <20020820213940.A8750@localhost.localdomain>
References: <E17hIuR-000Pd5-00@rip.psg.com> <042601c248ac$89127c20$6500000a@dev0196> <20020820213940.A8750@localhost.localdomain>
Message-ID: <20020820234754.65c3e735.mrose+internet.xml2rfc@dbc.mtview.ca.us>

> Why no vaidation pass locally?  It would obviously be useful.

useful, yes.

however, validation is currently done via a separate service written in java. given all the complaints i got a year about people not wanting to install tcl to run xml2rfc, i'm just not up for the conversations i'll have with folks stepping them through how to install a jvm, a whole bunch of packages, etc., etc.

on the plus side, i recently wrote a limited validator in tcl for another project. maybe i'll put that in the next release...

/mtr


From: randy@psg.com (Randy Bush)
Date: Tue, 20 Aug 2002 19:14:56 -0700
Subject: [xml2rfc] .xml2rfc.rc
Message-ID: <E17hL1M-0002qw-00@rip.psg.com>

first, will the xml2rfc tcl script always find ~/.xml2rfc.rc?  i
presume so.

but, if i manually run a slightly modified version of the one in
the distribution (appended), i get

    % tclsh8.3 .xml2rfc.rc
    can't read "inputD": no such variable
	while executing
    "file nativename $inputD"
	invoked from within
    "set nativeD [file nativename $inputD]"
	(file ".xml2rfc.rc" line 16)

secondly, how can i hack it to suck up the ~/refs/ids/* directory
as well?

randy, obviously not a tcl hacker

---

global env tcl_platform

if {![string compare $tcl_platform(platform) windows]} {
    set sep ";"
} else {
    set sep ":"
}

if {[catch { set env(XML_LIBRARY) } library]} {
    set library ""
    foreach bibxmlD [lsort -dictionary [glob -nocomplain ~/refs/rfc/*]] {
        append library $sep$bibxmlD
    }
}

set nativeD [file nativename $inputD]
if {[lsearch [split $library $sep] $nativeD] < 0} {
    set library "$nativeD$sep$library"
}

set env(XML_LIBRARY) $library



From: swb@employees.org (Scott W Brim)
Date: Tue, 20 Aug 2002 21:39:40 -0400
Subject: [xml2rfc] url prob in ref when converting to html
In-Reply-To: <042601c248ac$89127c20$6500000a@dev0196>; from mrose@dbc.mtview.ca.us on Tue, Aug 20, 2002 at 05:48:58PM -0700
References: <E17hIuR-000Pd5-00@rip.psg.com> <042601c248ac$89127c20$6500000a@dev0196>
Message-ID: <20020820213940.A8750@localhost.localdomain>

On Tue, Aug 20, 2002 05:48:58PM -0700, Marshall Rose allegedly wrote:
> the reason i say this is that when xml2rfc runs locally, it doesn't do a
> validation pass. if a validation pass were made, it would complain that the
> <author/> doesn't contain an <organization/>.

Why no vaidation pass locally?  It would obviously be useful.

Thanks ... Scott


From: mrose@dbc.mtview.ca.us (Marshall Rose)
Date: Tue, 20 Aug 2002 17:58:09 -0700
Subject: [xml2rfc] url prob in ref when converting to html
References: <E17hIuR-000Pd5-00@rip.psg.com><042601c248ac$89127c20$6500000a@dev0196> <E17hJkE-0000v8-00@rip.psg.com>
Message-ID: <045601c248ad$d1dd7990$6500000a@dev0196>

> thankd

no worries. enjoy!

/mtr




From: randy@psg.com (Randy Bush)
Date: Tue, 20 Aug 2002 17:53:10 -0700
Subject: [xml2rfc] url prob in ref when converting to html
References: <E17hIuR-000Pd5-00@rip.psg.com> <042601c248ac$89127c20$6500000a@dev0196>
Message-ID: <E17hJkE-0000v8-00@rip.psg.com>

>> [0] - btw, how do i get xml2rfc to produce an html output?
> % ln -s xml2rfc.tcl xml2html.tcl

aha!

> i'm also assuming that the reason you got that particular error is that
> when you selected a file to convert, you accidentally selected the text
> file rather than the html file.

<blush>

moi do something that stoopid?  nah.  never.

<blush^2>  <sigh>

thankd

randy



From: mrose@dbc.mtview.ca.us (Marshall Rose)
Date: Tue, 20 Aug 2002 17:48:58 -0700
Subject: [xml2rfc] url prob in ref when converting to html
References: <E17hIuR-000Pd5-00@rip.psg.com>
Message-ID: <042601c248ac$89127c20$6500000a@dev0196>

first, the short answer:

> [0] - btw, how do i get xml2rfc to produce an html output?

what i do is this:

    % cd ~/bin/
    % ln -s xml2rfc.tcl xml2html.tcl

then i can just say

    xml2html.tcl foo.xml

and it will produce

    foo.html

(assuming of course that ~/bin/ is in your $PATH and you've done a "chmod
+x")


now,  the longer answer:

> <reference anchor="Kerc83"
target="http://www.cl.cam.ac.uk/~fapp2/kerckhoffs">
>     <front>
>     <title>La Cryptographie Militaire</title>
>     <author initials="A." surname="Kerckhoffs" fullname="A. Kerckhoffs">
>     </author>
>     <date year="1883" />
>     </front>
>     </reference>
>
> produces what i wanted for text
>
>    [Kerc83]   Kerckhoffs, A., "La Cryptographie Militaire", 1883,
>               <http://www.cl.cam.ac.uk/~fapp2/kerckhoffs>.
>
> but the html (via the web page converter [0]) gets me (my backslashes)
>
>     <title>You lose</title>
>     <h1>Unable to Convert File</h1><pre>element \
>     http://www.cl.cam.ac.uk/~fapp2/kerckhoffs remains unclosed around line
101
>
>     Context:
> &lt;http://www.cl.cam.ac.uk/~fapp2/kerckhoffs&gt;
>     </pre><hr><address>Apache/1.3.20 Server at xml.resource.org \
>     Port 80</address>
>
>     <html>
>     <body >
>     </body>
>     </html>
>
> huh?

i'm assuming that what's going on here is that you are running xml2rfc
locally to produce text, but using the xml2rfc service to produce html. i'm
also assuming that the reason you got that particular error is that when you
selected a file to convert, you accidentally selected the text file rather
than the html file. just a theory. (you can check it by putting your .txt
file in an editor and looking at the line number where

                 <http://www.cl.cam.ac.uk/~fapp2/kerckhoffs>.

appears. i bet it's around line 101.

here's why i think that:

the reason i say this is that when xml2rfc runs locally, it doesn't do a
validation pass. if a validation pass were made, it would complain that the
<author/> doesn't contain an <organization/>.

when i take that reference and give it to the service (regardless of whether
i ask for text or html), i get this error:

    Unable to Validate File
    [Error] README.xml:299:15: The content of element type "author" is
incomplete,
    it must match "(organization,address?)".


anyway, you should now be able to generate html locally (the short
answer)...

enjoy!

/mtr




From: randy@psg.com (Randy Bush)
Date: Tue, 20 Aug 2002 16:59:39 -0700
Subject: [xml2rfc] url prob in ref when converting to html
Message-ID: <E17hIuR-000Pd5-00@rip.psg.com>

<reference anchor="Kerc83" target="http://www.cl.cam.ac.uk/~fapp2/kerckhoffs">
    <front>
    <title>La Cryptographie Militaire</title>
    <author initials="A." surname="Kerckhoffs" fullname="A. Kerckhoffs">
    </author>
    <date year="1883" />
    </front>
    </reference>

produces what i wanted for text

   [Kerc83]   Kerckhoffs, A., "La Cryptographie Militaire", 1883,
              <http://www.cl.cam.ac.uk/~fapp2/kerckhoffs>.

but the html (via the web page converter [0]) gets me (my backslashes)

    <title>You lose</title>
    <h1>Unable to Convert File</h1><pre>element \
    http://www.cl.cam.ac.uk/~fapp2/kerckhoffs remains unclosed around line 101

    Context: 
	&lt;http://www.cl.cam.ac.uk/~fapp2/kerckhoffs&gt;
    </pre><hr><address>Apache/1.3.20 Server at xml.resource.org \
    Port 80</address>

    <html>
    <body >
    </body>
    </html>

huh?

randy

---

[0] - btw, how do i get xml2rfc to produce an html output?



From: Michael.Tuexen@icn.siemens.de (Michael Tuexen)
Date: Tue, 20 Aug 2002 09:48:50 +0200
Subject: [xml2rfc] raft- instead of draft-
References: <F92978223B01D311ACFF0008C71EE19D015DF901@MCHH225E> <20020819105219.46c96e0b.mrose+internet.xml2rfc@dbc.mtview.ca.us>
Message-ID: <3D61F462.C2ECDB5@icn.siemens.de>

Seems to work again for me to. Maybe a temporary problem.
Thank for the help.
Best regards
Michael

Marshall Rose wrote:
> 
> > Dear all,
> >
> > there seems to be a problem with the convert to .txt, .html and
> > so on function. When submitting a .xml file I get
> >
> > ERROR
> >
> > The requested URL could not be retrieved
> >
> >
> >
> > While trying to retrieve the URL: http://xml.resource.org/cgi-bin/xml2rfc.cgi
> >
> > The following error was encountered:
> >
> >      Zero Sized Reply
> >
> > Squid did not receive any data for this request.
> >
> > Any hints?
> 
> looks fine to me.
> 
> /mtr


From: mrose+internet.xml2rfc@dbc.mtview.ca.us (Marshall Rose)
Date: Mon, 19 Aug 2002 10:52:19 -0700
Subject: [xml2rfc] raft- instead of draft-
In-Reply-To: <F92978223B01D311ACFF0008C71EE19D015DF901@MCHH225E>
References: <F92978223B01D311ACFF0008C71EE19D015DF901@MCHH225E>
Message-ID: <20020819105219.46c96e0b.mrose+internet.xml2rfc@dbc.mtview.ca.us>

> Dear all,
> 
> there seems to be a problem with the convert to .txt, .html and
> so on function. When submitting a .xml file I get
> 
> ERROR
> 
> The requested URL could not be retrieved
> 
> 
> 
> While trying to retrieve the URL: http://xml.resource.org/cgi-bin/xml2rfc.cgi 
> 
> The following error was encountered: 
> 
>      Zero Sized Reply 
> 
> Squid did not receive any data for this request. 
> 
> Any hints?

looks fine to me.

/mtr


From: michael.tuexen@siemens.com (Tuexen Michael)
Date: Mon, 19 Aug 2002 10:07:51 +0200
Subject: [xml2rfc] raft- instead of draft-
Message-ID: <F92978223B01D311ACFF0008C71EE19D015DF901@MCHH225E>

Dear all,

there seems to be a problem with the convert to .txt, .html and
so on function. When submitting a .xml file I get

ERROR

The requested URL could not be retrieved



While trying to retrieve the URL: http://xml.resource.org/cgi-bin/xml2rfc.cgi 

The following error was encountered: 

     Zero Sized Reply 

Squid did not receive any data for this request. 

Any hints?

Best regards
Michael

Michael Tuexen    Tel.:   +49 89 722 47210
Siemens AG        Fax:    +49 89 722 48212
ICN WN CC SE 7    e-Mail: Michael.Tuexen@icn.siemens.de
 

> -----Original Message-----
> From: Marshall Rose [mailto:mrose+internet.xml2rfc@dbc.mtview.ca.us]
> Sent: Friday, August 16, 2002 7:58 PM
> To: xml2rfc@lists.xml.resource.org
> Cc: Michael.Tuexen@icn.siemens.de
> Subject: Re: [xml2rfc] raft- instead of draft-
> 
> 
> > I'm using the ID database available at xml.resource.org. The
> > title of all ID start with raft- instead of draft-.
> > 
> > Could someome fix that? That would be great.
> 
> oops. the datasource got a little jittery. fixed.
> 
> /mtr
> 


From: mrose+internet.xml2rfc@dbc.mtview.ca.us (Marshall Rose)
Date: Fri, 16 Aug 2002 10:57:35 -0700
Subject: [xml2rfc] raft- instead of draft-
In-Reply-To: <F92978223B01D311ACFF0008C71EE19D015DF8FA@MCHH225E>
References: <F92978223B01D311ACFF0008C71EE19D015DF8FA@MCHH225E>
Message-ID: <20020816105735.496fa204.mrose+internet.xml2rfc@dbc.mtview.ca.us>

> I'm using the ID database available at xml.resource.org. The
> title of all ID start with raft- instead of draft-.
> 
> Could someome fix that? That would be great.

oops. the datasource got a little jittery. fixed.

/mtr


From: Michael.Tuexen@icn.siemens.de (Tuexen Michael)
Date: Fri, 16 Aug 2002 10:48:20 +0200
Subject: [xml2rfc] raft- instead of draft-
Message-ID: <F92978223B01D311ACFF0008C71EE19D015DF8FA@MCHH225E>

Dear all,

I'm using the ID database available at xml.resource.org. The
title of all ID start with raft- instead of draft-.

Could someome fix that? That would be great.

Best regards
Michael

Michael Tuexen    Tel.:   +49 89 722 47210
Siemens AG        Fax:    +49 89 722 48212
ICN WN CC SE 7    e-Mail: Michael.Tuexen@icn.siemens.de
 


From: jabley@automagic.org (Joe Abley)
Date: Thu, 1 Aug 2002 18:04:30 -0400
Subject: [xml2rfc] can't read "counter(firstxref)": no such element in array
In-Reply-To: <20020801194115.GC94747@felix.automagic.org>
Message-ID: <A75867CE-A59A-11D6-9AEA-00039312C852@automagic.org>

On Thursday, August 1, 2002, at 03:41 , Joe Abley wrote:

> xml2txt is complaining:
>
>   can't read "counter(firstxref)": no such element in array
>
> when I attempt to render a document which contains no xref elements.
> Inserting a dummy xref element avoids the error. This seems like a bug;
> is it?

For the benefit of the archive, this does indeed appear to be a bug and 
Marshall has identified a fix.


Joe



From: jabley@automagic.org (Joe Abley)
Date: Thu, 1 Aug 2002 12:41:15 -0700
Subject: [xml2rfc] can't read "counter(firstxref)": no such element in array
Message-ID: <20020801194115.GC94747@felix.automagic.org>

xml2txt is complaining:

  can't read "counter(firstxref)": no such element in array

when I attempt to render a document which contains no xref elements.
Inserting a dummy xref element avoids the error. This seems like a bug;
is it?

This is xml2rfc 1.13 running on Mac OS 10.1.5 with the shipped tclsh8.3.


Joe


From: mrose@dbc.mtview.ca.us (Marshall Rose)
Date: Thu, 1 Aug 2002 08:58:04 -0700
Subject: [xml2rfc] non-ascii names in xml2rfc, the sequel
References: <00c801c235ca$9d1f1100$0701000a@FATORA><JIEGINCHMLABHJBIGKBCAEJBFAAA.julian.reschke@gmx.de> <20020728223020.23d49898.mrose+internet.xml2rfc@dbc.mtview.ca.us>
Message-ID: <016601c23974$38ea1650$6500000a@dev0196>

here is my current thinking. comments are welcome.

to support authors who want localized names when rendering to a format that
supports that,
an optional element will be added to the <author/> element, i.e.,

    <!ELEMENT author      (organization,address?,unictext?)>

    <!ELEMENT unictext    (organization,address?)>
    <!ATTLIST unictext
              initials        %ATEXT;            #IMPLIED
              surname     %ATEXT;            #IMPLIED
              fullname     %ATEXT;            #IMPLIED>

 which allows CDATA and #PCDATA to contain unicode characters. The xml:lang
attribute may also be used to provide additional information.

The <unictext/> element is consulted only if the output format supports
unicode.

comments?

/mtr