The LLC board should figure out who is to manage the RSE.

I think that's what philosophers call a 'category mistake'.

Obviously, if there is a formal contract for services between the LLC and the RSE, that contract will be administered by the LLC. But if the RSE is responsive to the community in one way or another, the LLC will not "manage" the RSE like a staff member or like an operational contractor. The RSE is a colleague, not a servant.
On Wed, Jul 01, 2020 at 09:41:45PM -0700, Christian Huitema wrote:

On 7/1/2020 7:00 PM, Michael StJohns wrote:

As I've mentioned countless (sure - go ahead and count them...) times,
my view of the role of RSE is one that's accountable to the community
(and through the various payment and contract mechanisms, to whoever
is paying them to the extent the contract or employment agreement or
blood oath allows) for the role we set them, but is not under the day
to day control (management/oversight/whining) of the community, nor
the IAB/RSOC/IESG/IETF or for that matter the LLC.     E.g. someone
like a CEO, Chief Scientist, CTO, Editor in Chief, Publisher, IAB
chair, IETF Chair, IRTF Chair, IAD?

I understand that this is your vision. Not everyone agrees.

Maybe you can give a concise definition of what you mean by
"accountability" that doesn't include the "control" aspect?

A lot of what the RSE actually does amounts to project management, such
as for example "managing the deployment of the new RFC format".

But that is practically on autopilot by now, isn't it?

Accountability here should be the classic accountability for project
management.

A mechanism for setting particular goals (update RFC 5620) and for
removing the RSE, should suffice.  That would give the RSE a great deal
of independence.  We're at this juncture because we didn't give the
previous RSE that independence.
On 2020-07-03, at 18:03, Ted Lemon <mellon@fugue.com> wrote:

What would be an example of an RFC, approved for publication, which would be the analog of an unconstitutional law?

Since we don’t have a constitution, that analogy is leaky.
We have pieces of foundational documents, e.g., for one aspect of the technology development we do, we have:

RFC 1984
RFC 3365 Section 6
RFC 7258
<draft-iab-for-the-users-04.txt> (not yet consensus)

These documents are all flawed and might have process issues (do we really have consensus?).  I probably should add RFC 1925, which doesn’t even purport to be a serious document.  On a more technical level, we also actively revisit some of our most deeply entrenched principles, say draft-iab-protocol-maintenance (and we don’t always succeed in writing up the newly emerging consensus).

So it seems we can’t hold the person with the veto power accountable to a specific document or set of documents.  That limits the value of the analogy.  But the fact that documents that the Bundespräsident considers unconstitutional (in process or content) do get approved by the Bundestag shows that these things happen even in much more formally controlled environments.
On 2020-07-03, at 18:03, Ted Lemon <mellon@fugue.com> wrote:

What would be an example of an RFC, approved for publication, =
which would be the analog of an unconstitutional law?

Since we don=E2=80=99t have a constitution, that analogy is leak=
y.
We have pieces of foundational documents, e.g., for one aspect of the techn=
ology development we do, we have:

RFC 1984
RFC 3365 Section 6
RFC 7258
<draft-iab-for-the-users-04.txt> (not yet consensus)

These documents are all flawed and might have process issues (do we really =
have consensus?).  I probably should add RFC 1925, which doesn=E2=80=99t ev=
en purport to be a serious document.  On a more technical level, we also ac=
tively revisit some of our most deeply entrenched principles, say draft-iab=
-protocol-maintenance (and we don=E2=80=99t always succeed in writing up th=
e newly emerging consensus).

So it seems we can=E2=80=99t hold the person with the veto power accountabl=
e to a specific document or set of documents.  That limits the value of the=
 analogy.  But the fact that documents that the Bundespr=C3=A4sident consid=
ers unconstitutional (in process or content) do get approved by the Bundest=
ag shows that these things happen even in much more formally controlled env=
ironments.