--On Tuesday, 29 January, 2008 02:39 +0100 Frank Ellermann
<nobody(_at_)xyzzy(_dot_)claranet(_dot_)de> wrote:
I think procedural documents are different. The activity of
considering and discussing them is disruptive of the
community.
I guess that's because you feel they are very important.
Often I think about topics I consider as very important, "oh
no, I have no time for this, why must they screw *now* with
it, nothing is horribly wrong with how it is", etc.
No, it is for another reason entirely. My personal criterion
for when it is reasonable to approve, e.g., a standards-track
document is when that document represents rough consensus among
those who care, are materially concerned, and have sufficient
competence to have an opinion. That implies that, if there is a
piece of technical work about which I do not care and am not
concerned, it is reasonable for me to ignore it and keep working
on things I do care about.
Because process documents --especially ones that propose to make
fairly fundamental to 2026-- affect every one of us and
everything we do in the IETF, a Last Call on one should be
construed as a statement by the IESG that the topic is important
enough that we should all lay aside whatever technical work we
are doing and review it. That "laying aside" process is a big
enough deal (and disruptive enough of other work) that I believe
there should be at least some minimal evidence that the
community cares and believes that the proposed solution is
either the correct one or one on which we can reasonably build.
Incidentally, the reason I tend to oppose, or think we should be
very cautious about, process BOFs is a corollary of the above:
I'm uncomfortable having process decisions made by the subset of
the community who would rather spend their time in process BOFs
or at face to face process WG meetings, rather than the whole
community. At the risk of prompting a note from Spencer, I
would almost prefer that we tell people that, by registering for
an IETF meeting, they agree to run the risk of having to spend
some time in a process meeting, then select an appropriate
number of people at random and make them come. It would almost
certainly not work, but it would at least be far more
representative than what happens when we schedule a process BOF
against substantive WG sessions.
john
_______________________________________________
Ietf mailing list
Ietf(_at_)ietf(_dot_)org
https://www1.ietf.org/mailman/listinfo/ietf