I agree with each of your concerns, and ask too for discussion on what was
brought up in your message.
At 09:02 AM 10/18/2004 -0400, John C Klensin wrote:
Summary: Four weeks? When we sometimes run only three months
Some years ago, the secretariat and IESG agreed on an I-D
posting deadline about a week before IETF began, in the hope of
getting all submitted drafts posted before WGs needed them for
review and discussion. Prior to that rule, the last drafts to
arrive either slipped through the cracks or were posted after we
had started meeting, which did no one any good.
As the load of incoming drafts increased, still with a
completely manual process, the posting deadline was shifted back
another week, to be two weeks before meetings began, and then a
rule was imposed (for which I fear I'm at least partially
responsible) requiring that initial-version drafts be posted yet
a week earlier -- three weeks out. The theory behind the latter
was the load continued to rise and that initial versions often
took longer to process and confirm than second and subsequent
versions, so it made sense to let the additional time burden
fall on them.
Such deadlines, considerably in advance of IETF meetings, are an
impediment to objectives we claim for the standards process --
opportunities for people to get as much work as possible done
outside the face to face meetings, and documents in hand that
are timely enough that people who do not attend meetings in
person can effectively express their comments.
Over the last few IETF meetings, processing has become more
automated, or the Secretariat has become more efficient in other
ways. The typical time to get an I-D posted other than in the
pre- and post-meeting rush has dropped to one working day and
has sometimes even been less. And, during the rush, the queue
has often cleared early enough that consideration of shortening
the deadlines/ lead time would be in order.
Instead, a new rule has apparently crept into the posting
deadlines, with no community discussion or announcement other
than in those deadline announcements. The rule, in this
meeting's form, is that
"As always, all initial submissions (-00) with a
filename beginning with "draft-ietf" must be approved by
the appropriate WG Chair before they can be processed or
announced. WG Chair approval must be received by
Monday, October 11 at 9:00 AM ET."
First of all, this isn't "as always". The rule requiring
explicit WG Chair approval is fairly recent. But, more
important, we now have a situation in which WG drafts --
presumably the most important documents for the face to face
meetings-- now require formal naming, authorization, and
approval a full four weeks before the first IETF meeting
sessions. Remembering that we have sometimes had meetings as
close as three months apart, but even with four months being the
nominal separation, this is a _big_ chunk of time. On the three
month schedule, and allowing a couple of weeks post-meetings for
things to stabilize, people to get caught up, and new
discussions to start, it could give a WG only six weeks to have
a discussion that could generate a new document for discussion
and agree on that document before cutoffs impose, at least,
names that make those documents harder to find and track.
As we continue to discuss problems and issues that get in the
way of our getting effective work done, it seems to me that this
is a new one that should be added to the list.
Also, in the context of administrative reorganization, I would
find it helpful, and others might too, to understand where this
new requirement came from:
(1) If from the Secretariat by unilateral action, it is
perhaps a symptom of difficulties with the Secretariat
that require some change in models.
(2) If from the IESG, it perhaps should be examined as a
procedural change made without an announcement to the
community and opportunity for comment -- precisely the
type of change that the "July14" draft was intended to
prevent in the future by providing a more efficient way
to get such changes made _with_ community involvement
and (at least default) authorization.
Finally, if four weeks is really necessary, I suggest that we
are in need of firm rules about minimum meeting spacing.
Ietf mailing list
Truth is not to be argued... it is to be presented
Ietf mailing list