On 11/6/19 11:54 AM, Michael Richardson wrote:
Keith Moore<moore@xxxxxxxxxxxxxxxxxxxx> wrote:
>> WGs that regularly don't produce high quality work, perhaps, should
>> not be supported in the IETF?
> IMO most WGs should have a short (2-3 years) lifetime. So by the time it's
> clear that they do or don't produce high-quality work, it's time for them to
> be winding down anyway.
I would like to amend this. WGs should have a short (1-2) year iteration time.
That is, time from milestone being added to the charter to IESG Publication as PS.
I really think that WGs need to remain in existence longer in order to deal
with errata, to look at interop results, write operational documents, and
promote to IS.
I can see pros and cons with this. IMO one of the problems with
long-lived WGs is that they can keep creating more work for themselves
whether or not it is needed. Sure, they need to update their charters
to do so but I've rarely seen those efforts be rebuffed.
I also understand that the more WGs there are, the more load there is on
the ADs, even if many of those WGs are dormant.
So I think it might make more sense to have a specific long-term WG or
two per area to focus on maintenance, than to keep the narrowly-focused
WGs around indefinitely. Such a "maintenance" WG might also be tasked
with periodically reviewing each of the current standards-track
documents in that area, looking at the state of deployment and
suitability for purpose, and making recommendations for updating, status
change, new documents (e.g. usage profiles or applicability statements),
or deprecation as appropriate.
The list of AD candidates is effectively limited to the current list of WG chairs.
(I've said this before, multiple times)
I don't think this is actually true. It's closer to the set of people
who have chaired a WG at some time in the present or past. And I think
it would be fair to consider directorate members in some cases.
Since most of the ADs were awesome WG chairs, it's hardly surprising that
they revert to what they were good at, and basically just take on the WG
chair responsability for document quality.
While that might happen in some cases, I don't think that's a good
explanation for ADs taking on detailed review. ADs know that they're
in a unique position to uphold the quality of IETF standards. They see
the widely varying quality that WGs produce, and they know how important
it is to get some of that stuff right. (Some of it is relatively
inconsequential.) And in practice, an AD actually has fairly limited
power to make sure that a WG reliably produces good output. The AD can
in theory fire poor chairs, but this has the potential to be so
disruptive as to make things worse, and there may be a dearth of good
chairs to replace them. It's all well and good (and politically wise)
to get directorate members to do detailed reviews, but an AD doesn't
always have that luxury, and sometimes an AD has to dig fairly deeply
into a document in order to make sense of the (sometimes conflicting)
reviews and Last Call comments anyway. The buck really does stop with
the IESG and IMO they're the ones in the best positions to be making the
final decisions - partly because they are likely to have broad views,
and partly because about the only thing that gives them any "steering"
capability is the fact that they're the ones doing the final signoff.
Afterall: that's the skill that they demonstrated so well that it got them "promoted".
I believe that AD and WG chair require somewhat different skill sets.
If I were on nomcom I'd be looking for ADs who have demonstrated cat
herding expertise and are conversant in multiple technical subject
areas. A WG chair can be a specialist; an AD needs to be a generalist.
Keith
