ietf
[Top] [All Lists]

Re: Progressing I-Ds Immediately Before Meetings

2008-07-19 08:22:57
(reposting the substance of an earlier remark which appears to have
gotten lost)

...

So, Jari, let me make a suggestion.  Get the relevant draft
together, send it to the Secretariat, and ask (or tell) them to
manually post it and to do so RSN.  If they won't do it, you,
the IESG, and Ray need to have a discussion, but I can't imagine
that being necessary.  Especially because it has been some years
since this sort of exception was made, send a note to the WG,
copying the IETF list, indicating that you decided to apply good
sense and get the document posted in order to facilitate
discussions in Dublin and to avoid unnecessary delay in getting
a document wrapped up and published on which substantial
consensus already existed.

I have a more radical suggestion: Let's do away with the cutoff thing entirely,
replacing it if necessary with per-WG policies about what sort of advance
notice needs to be given for a document revision to receive consideration.

The reality is simply this: A given document is either stable or unstable. And
if it isn't stable, preventing people from posting a revision for a brief
period isn't going to magically make it stable. Really, all it does is insure
our repository contains stale data.

In many cases it doesn't even prevent consideration of a revised version by a
group. Over the years I've seen many postings along the lines of "revision
missed the cutoff so it isn't in the archive, you can find a copy of here" to
some list or other. And frankly this is preferable to the alternative of
having a bunch of slideware saying "this issue fixed in the revision that's
not posted yet".

Although to be fair, now that the gate opens once the week starts the
"revision by slideware" problem has become less common. What we see now
is revisions being posted the minute the gate opens and those revisions
being discussed during meetings. I guess this affords some "stability
protection" to the monday morning slot, but not much beyond that.

The rules we have now were designed for situation that no longer exists: One
where all drafts were processed manually, it was more difficult to find
alternate posting venues, and where the revision cycle was generally more
costly and as a result revisions occurred less often and were more substantial.
All of this has changed and we're saddled with a process that no longer makes
sense. It should be replaced by the application of common sense in the working
groups themselves.

I want to stress that I'm not recommending modifying the
non-document that defines the posting deadlines to try to
enumerate the exception cases, much less building a lot of extra
facilities into a future version of the posting tool.   I'm
suggesting that you have the authority (and, actually, the
responsibility) to get the document posted simply by the
application and assertion of good sense and that nothing more is
either necessary or desirable.

Well, we agree insofar as I think additional process rules for this are a
terrible idea. But while i applaud the general notion that the application of
common sense can take the place of all that, in this case the cost is going to
be pretty high: This sort of exception handling is going to take WG chair or AD
time, plus the mechanics of handling the exception will likely take secretariat
time. So I think the first application of common sense needs to be whether or
not any of this is necessary and whether the benefits of the draft cutoff are
worth the cost.

I imagine the WG would applaud.   Many of us, especially those
who have been increasingly concerned that the formation and
application of rules is replacing the application of good sense,
would also applaud.  I don't see any downsides other than
breaching the sanctity of The Rules, and I don't believe we
actually have Sacred Rules around here.

I'd much rather breach the sanctity of the rules by getting rid of some of them
entirely.

                                Ned
_______________________________________________
Ietf mailing list
Ietf(_at_)ietf(_dot_)org
https://www.ietf.org/mailman/listinfo/ietf