--On Monday, 09 May, 2005 13:56 +0200 Harald Tveit Alvestrand
My immediate reaction is "who were the available candidates
In contentious groups, the requirement list is roughly (not in
- Knows enough of the technology to understand the issues
- Knows enough about the IETF process to steer the group
- Respected enough by the groups of people involved
- Not strongly identified with any of the camps of contention
- Has time enough available (and an employer or family that
will allow them to spend that time)
- Has the personal qualities needed to get people to come to
- Is known enough to the AD that he/she is comfortable working
with that person
- ....and I'm sure there are things I've forgotten.....
And since the intersection of all those qualities is
frequently the null set, chair candidates tend to be lacking
in one or more of these qualities. In the cases cited, the
"time enough available" may be the factor that changed - I
don't know the specifics.
I see two possibilities when a chair fails to work out:
- The AD made a bad choice, and there was someone else who
could have done a better job. Solution: AD needs help in
- There was no better candidate at the time (all the other
candidates being more obviously the wrong person for the job).
Solution: The chairs need help in calling for help earlier
when they're unable to perform, and the AD needs to be more
proactive in replacing chairs who aren't doing what they
should (again going back to the candidate pool).
I suspect that the latter happens more often than the former -
but as I said, I don't know those specific cases.
I would add one, which is the consequence of your "known enough
to the AD..." observation. There is a completely natural
tendency, whether it causes this problem or others, for the ADs
to keep going back to the same well of people who have known
abilities, especially abilities to handle this sort of
situation. To the extent to which that happens, it suggests
(i) We aren't doing enough to develop leadership and spread the
responsibilities around. In many cases, if all of those
criteria cannot be met, it would be far better to drop the
"known enough to the AD" and/or "knows enough about the IETF
process" criteria requirements and focus on the others plus an
apparent willingness to learn, then add in someone who meets
those requirements but whose role is advise and mentor the
Chair(s) --and do so with some authority and a close
relationship with the AD-- than to focus on those criteria.
(ii) We need to encourage more interactions between sitting ADs
and those who are not already WG Chairs, IESG members, or IAB
members. The current patterns of ADs tending to spend all of an
IETF week together, with WG Chairs, and monitoring WGs are not
conducive to ADs getting to know people who are not already part
of the leadership structure well enough to meet that criterion.
(iii) Without getting into whether the problem of overlong AD
tenures has been fixed or whether the shorter terms today are
just a temporary aberration, very long tenures on the IESG,
rather than having people regularly return to the trenches and
get first-hand experience, also, IMO, tends to aggravate the
problem of the AD not knowing a wide enough range of possible WG
Ietf mailing list