On 11/6/19 7:43 PM, Brian E Carpenter wrote:
On 07-Nov-19 12:01, Keith Moore wrote:
On 11/6/19 5:54 PM, Brian E Carpenter wrote:
Here's a thought experiment.
Update the standards process such that the approval of Proposed Standard
RFCs, after an IETF last call including some specified cross-area review
requirements, is done by the WG consensus process with the consent of the AD .
I don't think a typical WG chair is in a good position to review things
from a broad perspective. The ADs are in a MUCH better position to do
that, precisely because they are exposed to everything that IETF does.
But it doesn't scale, or so the ADs are telling us.
I believe them :)
Or to put it differently: I don't think it scales so long as there is
essentially no pushback on chartering any WG for which there appears to
be community interest.
Note that there's not a contradiction there. The ADs really are in the
best position to review things from a broad perspective. AND with the
current number of working groups and number of drafts that they're
producing, there's too much work for the ADs to do.
But to me it looks like what doesn't scale is the community expectation
that IETF should take on every WG for which there seems to be public
interest, and those WGs should be able to produce as many drafts as they
wish, with no page limit on those drafts.
(I'm not saying that there can't be some middle ground, some
optimization of what's currently being done. But I do believe that the
organization needs to take a good hard look at its expectations for
output and for "fairness" in what IETF takes on, and think about what
the Internet needs from IETF rather than about what individuals and
companies want IETF to do for them.)
It's long been clear to me that there's an inherent limit to the extent
to which the IETF structure can scale, for several reasons other than
just AD workload. And to me that forces some hard questions about both
the volume and scope of IETF's work. IMO IETF should focus on
producing fewer but higher quality and more relevant RFCs, rather than
priding itself on the sheer number of RFCs produced.
And my thought
experiment wouldn't take ADs out of the loop; it would take them out
of the detailed review work.
It would also take away the tremendously valuable perspective that they
acquire as the result of doing (some of) that work. We would soon be
left without anyone who could really do a good job at broad review.
Or maybe the thing to do is to make directorates more formal and more
recognized, and expand their roles a bit. Maybe we should explicitly
look at (some?) directorates as the places to cultivate broad expertise
(others might still have narrower focus). Then have the directorates
tasked with not just reviewing documents and making recommendations
about them, but also other things - e.g. analyses of problem spaces and
tussles - that would inform AD decisions and also feed-forward to WGs.
The buck would still stop with IESG but they'd have more explicit support.
Also, the WG chairs are properly concerned with the specific perspective
of their WGs; they know where the hard battles were fought. Their WG
needs them to be in a position to defend the WG's work. To expect them
to do both that and the broad review would put them in a conflicted
position, and it's probably the broad review that would get shortchanged.
That's why the idea would be to make the WG chairs the *visible* approvers,
which IMHO would significantly change their incentives.
Yes, but then they wouldn't be in a good position to support the WGs.
A chair can't be both in the trenches in a WG AND be a reasonably
objective reviewer from a broad perspective. Either they're separate
people, or we just give up on our notion of rough consensus because
nobody would be in a good position to evaluate it.
Keith
