Hi Wes,

My concerns may not be identical to Bill's.

I support the suggestion that the co-chairs do a consensus check on
the statements Bill mentioned.

The co-chair's - now Tony is the sole chair - rely on our support for
publishing the draft as an RRG RFC, and will test that with an online
poll.  I think we all know the potential problems of of such
arrangements, but its the best which can be done in a self-selected
group such as this, with on formal membership.  So I think it is right
that serious questions about the nature of the RFC be discussed, with
a consensus test initiated by the chair if more than a few people
express support for something contrary to the current plans of the chair.

I support Bill's view that it is best the chair initiate any such
consensus test.  If someone else initiates a poll, then the results
are of debatable value, since not everyone would be motivated to
participate who would be motivated to participate in a proper
consensus check poll initiated by the chair.

I don't think it would be disastrous for the draft to be finalised as
it is, to be the "RRG Recommendation", which is ~95% useful material
on 15 proposed architectures, and ~5% the co-chairs' recommendation.
By "useful" I mean informative and helpful for people in the near term
and longer term future who want to understand the alternatives and do
their own evaluations.

The recommendation is quite brief and not very well explained or
argued.  Its value is that it is the carefully weighed judgement of
two people with long experience in this field.  However, I think its
value  is limited because the recommendation is so brief, because
there is little or nothing on the goals they hope to achieve, and
because in the absence of them detailing their goals, the current
version of the Design Goals is old and likewise brief and inadequate.
 (I plan to write my own "alternative RRG Design Goals" I-D - looking
at this massive global IT system upgrade more comprehensively and in
more detail than is favored by the chair.)

The value of the co-chairs' recommendation will diminish over time as
new architectures are developed and as those which were discussed in
this phase of the RRG are improved, developed, or found to be wanting
and abandoned.  The value of the summaries etc. will also diminish
over time, but they tell the reader much more about the various
architectures strengths and weaknesses than the co-chair's recommendation.

I think it would be better to put their recommendation in a different
RFC from the RFC which contains the material on the 15 proposed
architectures, and call the latter RFC the "RRG Report" or "Summaries
and Critiques of RRG Candidate Architectures" or whatever.

The authorship of these two items is quite different.  The co-chairs
alone are the authors of their recommendation.  The summaries and
critiques etc. are the work of the people who actually designed and
proposed new architectures, and of some of the people who wrote
critiques of these architectures.

The bulk of the current draft is not a recommendation - it is a
summary of the work done by many, but unfortunately not all, of the
most active and constructive members of the group over the last 3 or
so years.

The recommendation does not come from the RRG as a group.  It only
formally comes from the RRG by way of the co-chairs deciding to use
their defined ability to do things as they wish.  That's fine, but the
recommendation does not come from vast majority of RRG participants.
Nor does it have the support of the vast majority of the RRG
participants.  To bundle the RRG group's major single collective work,
which is not a recommendation, under the title "RRG Recommendation"
does not seem right to me.

Furthermore, given the apparent misunderstandings the co-chairs seem
to have had about the candidate architectures in March, when they
decided on their recommendation:

  http://www.ietf.org/mail-archive/web/rrg/current/msg06373.html

I would argue that their recommendation is not based very much on the
main body of the current draft - the summaries, critiques and
counterpoints of the 15 candidate architectures.

This is a second reason for separating the recommendation into a
different document from these summaries etc.

Although the draft is perfectly clear that the recommendation is from
the co-chairs alone, I think its placement after all the summary etc.
material gives an impression that it was well informed by that
material.  It think such an impression is not as true as it would
ideally be.

  - Robin
_______________________________________________
rrg mailing list
[email protected]
http://www.irtf.org/mailman/listinfo/rrg

Reply via email to