On 10/20/2013 09:54 PM, Joel M. Halpern wrote:
The obvious risk if one separates WG management from review is that
one could easily get the situation where the manager, in working with
the WG, says that the document needs X, Y, and not Z. And then the
reviewer says "needs Z". And there are more extreme versions of this.
Currently, the ability for the managing AD to say "no, this won't pass
muster" is part of his management tool. If he is not the reviewer, he
seems to have lost an important tool. I hope I am missing something
that would make this sort of approach workable.
Without having scanned every suggestion in this thread....
In my opinion, it is critical that it is clear who's got the final say
in whether a document is approved or not. At the moment, it's the IESG,
and I think it's likely that it should remain the IESG - if approval
authority isn't in the same chain of responsibility as management
authority, one has to find a way to deal with conflicts between the two
- and I think that's not a good way to go.
When I was on the IESG, and instituted the General Area Review Team, I
chose to trust reviewers to the degree that I would not read the
document unless my reviewer said I needed to. But the final decision on
the approval call was still mine; the reason I could do this was because
I trusted my reviewers.
Placing this amount of trust also means that it must be clear what
happens if the trust is broken - such a reviewer HAS to serve at the
pleasure of the AD.
If one wishes to shuffle around the responsibility differently, the
equations change. But splitting review from the IESG doesn't have to
mean splitting the authority for approval from the IESG.
Yours,
Joel
On 10/20/13 3:38 PM, Brian E Carpenter wrote:
On 21/10/2013 03:18, John C Klensin wrote:
...
Because working together with others on final document reviews
and providing a final verification that all necessary bits are
in place shouldn't require nearly the job-learning time that the
IESG/AD role does, one could consider asking members of that
review body to serve only a year or maybe six months, and
possibly relax the meeting attendance requirements,
I think we could totally remove attendance requirements. There
are cases where understanding a document requires background knowledge,
but if the document doesn't contain pointers to that knowledge, it's
a defect in the document, not in the reviewer. In fact, we can argue
that since the acid test for a document is whether a reader who has
zero IETF experience can understand it without ambiguity, at least
one reviewer should have limited background knowledge.
I don't think a time limit is the critical issue - it's how many
documents a given reviewer has to review per year, or to be more
precise, how many pages. Clearly the burden in pages/year on a
conscientious AD at the moment is silly.
either of
which should vastly increase the pool of plausible volunteers,
including some people we would very much like to have in those
jobs but who can't take on the IESG workload or long-term
commitment.
For an earlier and somewhat different take on this issue and a
variation on a proposal, see the long-expired
draft-klensin-stds-review-panel-00.
The high-order bit is whether we can separate the functions of
(a) steering the work of the IETF and (b) applying final quality
control to the documents. At the moment, these two jobs are bound
up with each other in the IESG.
Brian
--
Surveillance is pervasive. Go Dark.