On Aug 31, 2011, at 2:36 AM, Jari Arkko wrote:
Keith, thank you for the feedback. Some responses inline:
1. Fix the broken IESG voting system before you try to establish more
decision criteria.
I do agree with your general thinking here. The way that you describe the
different positions is what I personally try to achieve in my IESG reviews.
But I think you are overemphasizing the role of the vote designations. Again,
I try to do this already, though maybe not always succeeding. In general, if
the Area Directors are not doing their job of stopping bad stuff and moving
good stuff along, they don't need a new voting system, they need to grow a
spine.
Elaborating a bit more about this:
We have plenty of cases where a DISCUSS has been left standing, and today
that acts as your "NO" vote. (Obviously, in many cases this was an error, or
lack of effort. I'm quilty of this for sure, even right now I have a queue of
DISCUSSes and their proposed resolutions that I need to go check.) But I
think a DISCUSS should stand, if there is a serious issue and it is not
rectified.
There are corner cases where a single AD has an opinion that is not shared by
the rest of the community. For those cases we have an override procedure.
This has been never been invoked, but that's probably because it gets pretty
ugly way before that -- there are often heated discussions between ADs about
discusses.
We have the ABSTAIN vote, which some ADs use to vote "NO", often together
with other ADs who feel the same. There's never been a case where this would
have blocked a document from proceeding, as we've never collected the
necessary 1/3 number of ADs to vote ABSTAIN for any single document. My
conclusion is that ABSTAIN stands for NO-OBJECTION in practical terms. I
don't recommend its use...
The biggest problem with the current voting system (other than misleading
labels, which do cause real problems of their own) is the presumption that the
document should go forward no matter how few IESG members read the document.
So No Objection votes from ADs who didn't read the document count as Yes votes,
but there's also a presumption in the rules (as well as pressure from other ADs
who want to get documents off the agenda) to clear Discuss votes in favor of
moving a document forward whether or not the identified issues have been
adequately addressed.
(One thing that I didn't mention that also needs to be fixed if it's still the
case is the presumption that the responsible AD votes Yes for the document. I
don't know what the tools do now, but this Yes vote used to be automatically
filled-in.)
2. Don't overconstrain the use of DISCUSS. In particular, don't ever create
a situation where a reviewer can't cite a problem with a document,
regardless of whether that problem has previously been enumerated.
I agree, and that's why the guidelines I posted are just that -- guidelines.
They are not binding rules, they leave room for a judgment call.
3. I take serious issue with the statement in the draft that IESG reviews
are "reviews of last resort" and the implication that WG reviews are
sufficient. In numerous situations this has not been the case.
Of course. But I don't see a conflict between a "review of last resort" and
having the last resort find issues. I wish we'd find less issues, but at
least I still view the IESG as the final check, that should catch issues if
others have not. Its not a "last resort" in the sense that it would not be
invoked; we do review all documents very carefully. Its a last resort only in
the sense that there is an expectation that previous stages should have
produced a quality result without issues.
That is not a valid expectation, in either theory or practice. That's my
point. Arguably when a poor quality document gets to IESG, it's a failure on
the part of the WG to do due diligence. But the problem is actually deeper
than that - it's partially structural (in that IETF partitions almost all work
into narrowly focused WGs who don't represent a sufficiently broad spectrum of
interests), and partially due to a failure to consistently apply good
engineering principles across all of IETF.
IESG's assuming that the WG has produced a quality result basically works to
mask the other problems with IETF's way of doing work. But even if WGs
generally did produce high quality results without issues (which I don't think
is the case now), IESG review should still not presume that they do. There
will always be some failures at the WG level, and the IESG's job is to try to
catch those.
Keith
_______________________________________________
Ietf mailing list
Ietf(_at_)ietf(_dot_)org
https://www.ietf.org/mailman/listinfo/ietf