--On Friday, 18 January, 2008 13:18 -0600 Eric Gray
<eric(_dot_)gray(_at_)ericsson(_dot_)com> wrote:
John,
Your description of the reasons for having the draft sub-
mission dead-lines may agree with original thinking that went
into setting them, initially. However, there were collateral
benefits that the new automated submission process helps to
improve - but does not eliminate.
For the people who participate in a fair number of working
groups in the IETF, requiring early posting allows for a
greater likelihood that they will be able to at least skim
each new draft sometime before setting up their laptop at the
beginning of each meeting in which that draft will be
discussed.
Eric,
To be clear, I was not advocating doing away with the cutoff. I
think cutoffs are a good idea, for exactly the reasons you
suggest and because they avoid any necessity for per-WG rules
about deadlines to prevent a particularly nasty way of gaming
the system. I just want to be sure that the intervals we have
been using are still appropriate.
Moreover, having a week longer grace period for subsequent
submissions also makes sense from this same perspective -
because it is usually the case that there is less new material
to absorb in a -01 or subsequent version than there was in a
-00 version. Not always, but usually. One exception is when a
draft becomes a working group draft - which means it becomes a
-00 version with virtually no change from a previous (often a
-03 or -04) version.
Just for purposes of discussion (I do not have any strong
positions on this other than a desire to have it reviewed), I
think it makes a lot of sense to require a long lead time on new
drafts in order that people have a way to consider new concepts,
whether they want to attend particular WGs, etc. Three weeks
does not strike me as unreasonable for that purpose. And the
draft-transition exception you mention is already part of the
rules.
However, in many WGs, we see a lot of work done in the weeks
before an IETF meeting, both before and after the current
posting deadline. I think it is in our interest to have WGs
looking at drafts that are as up-to-date as possible, consistent
with everyone in the WG having a reasonable opportunity to read
them before the actual meeting. So I'm not sure that two weeks
is optimal for revision drafts. Perhaps the tradeoffs would
work better at a week before, or perhaps at some other interval.
I suspect that setting the revised draft cutoff much shorter
than a week before the meeting would mean that some people would
not be able to review them before beginning to travel to the
meetings and that the Secretariat might not have time to
manually process whatever needed to be manually processed, but I
don't know.
I do not think it is either wise or useful to try to design the
details of this on the IETF list (nor do I think you were
suggesting that either) and hope my note does not set off a long
thread. I believe it is reasonable for us to ask the IESG to
think about the issue, make a decision and tell us. I'd prefer
to hear about their reasoning, but I can live without even that.
I just don't think it is reasonable to continue automatically
with the current cutoffs when the reason for setting those
particular cutoff values has largely disappeared.
And I wanted to mention the question on the IETF list precisely
because I hope that this is not an appropriate subject for an
RFC3933 experiment.
john
_______________________________________________
Ietf mailing list
Ietf(_at_)ietf(_dot_)org
https://www1.ietf.org/mailman/listinfo/ietf