okay, this is getting way too long, and starting to get repetitive and
even personal, so I'm going to summarize:
1. A review structure that works for conferences doesn't necessarily
work for IETF.
Conferences and IETF have different goals. IETF tries to do
engineering and produce technical consensus behind particular
solutions. Conferences try to stimulate thought and discussion.
IETF tries to get closure; while in a conference, some amount of
controversy is even desirable. Conference paper reviews are just
trying to keep the signal-to-noise ratio high; IETF reviews are trying
to see if the design is complete, appropriate design choices were made,
and if the various protocols can peacefully coexist in the network.
Conferences and IETF also draw from different constituencies,
and IETFers may have more constraints on their time than those
who attend conferences and submit papers to conferences.
2. IESG's scaling problems are a direct result of low-quality output
from working groups, and we can't do much to address that problem
by changing how IESG works.
3. I don't think we can make IESG significantly larger, I don't think
we can dispense with final document review and keep document quality
up, and I don't think that additional reviewers can signficantly
relieve IESG of the need to do final review. I do think that
additional reviewers could be very valuable in giving WGs feedback from
early drafts, keeping them on the right track, and keeping IESG
informed about the status of the WGs. I also think that a document
that has enjoyed such review and feedback throughout its life cycle
will be much easier for IESG to review, and that (without any changes
to IESG's organization or process) it will be harder for IESG to reject
such documents without sound technical justification.
4. Broad based (not just "cross-area") review is essential and IETF
doesn't have adequate structures to do this at present. WGs generally
need to have short lifetimes and to stay focused on deliverables,
while issues of conflict between concerns exist for years after a WG
has finished.
5. These days, most IETFers are narrowly focused. What this means
is that we have to actively cultivate a broad view among people
entrusted to do broad review. IESG turns out to (accidentally) be a
mechanism for cultivating broad views, but it's not the only possible
mechanism for doing so.
I suspect we would agree that the ADs alone aren't sufficient. But I
think the primary benefit of additional reviewers is in early review
rather than late review - I think we want the early review both to give
timely feedback and correction to WGs and to give the ADs confidence
that the WGs have stayed on track and done their homework, thus making
the late reviews easier.
---
and one specific response to something you wrote:
We don't believe in kings, and IMO, the IESG have too much king-like
power in the current structure.
Start giving IESG real rough consensus (rather than consensus by
exclusion and/or exhaustion) and real running code (or even
better, convincing analysis that the protocol will work well in the
wild), and they won't feel the need to rule by edict. IESG appears to
rule by edict because WGs demand that IESG provide them with very
concrete feedback. Simply saying "you failed to provide security" or
"you failed to address the concerns of this other group that you'll
harm their interoperability" doesn't work - either the WG will balk or
they'll sit on the document for months not being willing to fix the
problems. So ADs try to come up with good technical solutions in order
to get closure. If the WG likes the solution all is well; if they
don't like the solution they complain about ADs who act like kings.
But mostly, WGs get what they deserve. ADs like nothing better than
to be presented with well-written, focused documents from groups that
have obviously done their background work and given it appropriate
consideration in their designs.
Keith
_______________________________________________
Ietf mailing list
Ietf(_at_)ietf(_dot_)org
https://www1.ietf.org/mailman/listinfo/ietf