Speaking as an IESG member who may soon have to evaluate these
documents as proposed standards I would like to see the results of
a survey (not a vote, we don't do that :-) ) like the one that Steve
is running.
Fair enough. I do think there's ample reason to object to this vote on
procedural grounds. For one thing, chairs cannot call for a vote in any
procedural system I've ever heard of -- motions to vote have to come from the
floor. And this is especially true in one like the IETF that doesn't allow
voting in the first place! However, if the IESG (or even one member of the
IESG) think this is useful I won't bother to raise any of the procedural issues
I feel could legtimately be raised at this time. The IETF isn't much for
formality in these things anyhow, and frankly, neither am I.
I would like, however, for the period of the voting to be reduced. Three week
is far too long -- this effort is already in sudden death overtime. Had the
vote been called right after the meeting a three week period would have been
fine. In fact a last call is intended to be exactly this sort of survey. As
such, I think we've been "surveying" for three weeks already, and I think a 10
day period in which to actually vote is more than sufficient.
I therefore call on both the working group and the chair to act to reduce the
voting period to something more reasonable.
I have been monitoring this list for quite a while and do not remember
a time when there was quite so much 'discussion' on it. I find it
quite hard to judge the level of consensus present within the group
and would hope that Steve's survey will help to establish at least that
level of understanding. The level seems to be a bit over rough
at this point (30 grit?) even for the IETF. I would like to be shown to
be misunderstanding the tenor and content of the messages I've been
reading and the survey may just do that.
It is an unfortunate fact that this group had previously fallen into the
pattern of only discussing these documents at working group meetings, rather
than on the list. See my earlier note laying out the cronology of this effort
for details (or, if you don't believe me, take a look at the list archives
yourself). The fact that there's more discussion of MIME/PEM in the minutes of
the meetings than anywhere else in the list archives provides ample proof of
this.
Due to scheduling conflicts my involvement in PEM discussion prior to MIME/PEM
was fairly limited -- IETF meetings tended to schedule either the IETF-822 or
IETF-SMTP working group on top of the PEM working group meetings. (I actually
feel that many of the problems integrating these technologoies we're faced with
today traces back to these scheduling conflicts. The early MIME work could have
benefitted from additional PEM working group involvement, and vice versa.) As
such, I cannot adequately assess whether this pattern started with the MIME/PEM
work or is just the way this group handles some sorts of issues. Nevertheless,
up until recently this has been the way this group handled MIME/PEM.
But that's the past, and we're now faced with a very new phenomenon in this
group -- voluminous discussion of real MIME/PEM issues on the mailing list.
There have been more messages exchanged on MIME/PEM matters in the last month
than in the prior two years on the list. (Count them yourself and see.)
This aspect of recent events pleases me somewhat. I have never been comfortable
with the way this group has handled these documents. Specific constructive
criticism of the specifications has been shockingly limited in the past. (There
has been a lot of criticism, but a lot of it has been vague, generally
unhelpful, and often the obvious product of the grinding of one person's
personal axe.) Given the lack of the proper sort of feedback, I think the
specifications have stood up to recent intense scrutiny amazingly well. In
particular I feel that the security multipart specification is ready to move
forward at this time, since nobody has identified any significant issues with
it.
Depending on the handling of the key selector issue (I think there's a
consensus backing up the status quo, but I'm not completely sure about this),
the PEM-specific document may be ready to move soon as well.
(In response to a number of comments I've seen today, Yes, I am taking a
look at the documents on the virtual table and will give whatever feedback
I have to the authors if the comments are related to document style or
clarity and to the mailing list if the comments are related to technology.)
This is good. Having been on the receiving end of IESG document comments in the
past, I would like to make two particular requests: The first is to be as
specific as possible in any comments you make. Specification of actual sections
in the document that need work is essential, and specific prose changes would
be best. We're in last call here, and the time for general handwaving arguments
about the thrust of the specifications or their overall form or whatever is
long past.
Too often the IESG falls into the trap of thinking that specific targeted
criticism will be badly recieved by the document authors, and that generalities
are the only way to guard against offending authors of "sacred prose". Rest
assured that this is not the case with these documents -- they have passed
through too many hands and too many editors for there to be any sort of
negative response. Indeed, the major problem here is that the authors are sick
and tired of vituperative name-calling (as evidenced by the chair's repetition
of the anonymous characterization of these documents as "shit") that serves no
useful purpose.
My second request is that these matters be relayed to the document authors
directly. (You should feel free to post to the list or copy the chair or not as
you see fit, of course.) Normally I expect such feedback to come through the
chair, but here we are faced with the unfortunate fact that comments apparently
made by the IAB or the IESG about these specifications did not make it back to
the document authors so they could be addressed. I am considerably more upset
by this aspect of the handling of these matters than I am by the actual content
of the comments themselves. I therefore request that you deal directly with the
authors.
Enough said.
Ned