At 12:11 09-07-2008, John C Klensin wrote:
Three things occur to me. Neither is a solution, but I believe
that they would be helpful:
(1) Unless there is a technical issue that affects the protocol
itself (not the form of the specification), IESG Statements and
notes on Protocol Action notices should be favored over IESG
modifications to, or disclaimers on, actual Standards Track RFC
text.
People won't see the IESG Statements when they read the RFC. Unless
there is a serious technical issue, I don't see why an IESG
disclaimer is needed in a Standard Track document. But if it helps
the IESG feel good ... :-)
(2) As was discussed in other contexts, any assertion of
substantive harm needs to be backed up by clear evidence and/or
community consensus. If neither is possible, the third
provision becomes even more important.
If IESG notes are to be taken seriously, then any assertion should be
backed by clear evidence, or if that's not possible, community consensus.
(3) Any time the IESG decides it is important to put one of
these statements into RFC text without clear and demonstrated
community (not just IESG consensus), all of the IESG members
supporting that move should be required to identify themselves
explicitly in the Protocol Action Notice or some other location
that is visible to the community and the Nomcom. The Nomcom is
the communities main way or judging, and protecting itself
against, behavior that might be considered petty or otherwise
inappropriate. ADs should be required to identify themselves
and take responsibility for their decisions, not to hide behind
"the IESG finds".
Either the IESG speaks as a collective body or else it's a group of
individuals where we might see dissenting arguments published. If
it's the first (speaking in general terms), then any dissenting
member should resign in case of serious disagreement. If it's the
second, then we have to rethink the notion of consensus.
For argument's sake, let's assume member X did something
inappropriate. As a collective it's for the group to see that it
doesn't happen as it reflects badly on the body as a whole. If that
cannot be avoided, then the group has to take action against member
X. Some groups may only be ready to do that in extreme
situations. Meanwhile, member X will continue causing damage. As
long as everything is tightly held within the group, there is no
transparency. If we apply the preceding sentence to groups regulated
by Nomcom, then there is no way for Nomcom to do the judging.
The comments about the subject line provides a view how things work
in the IETF. Participants can fight a decision. That works quite
well in Working Groups as the flame wars will attest. As we move
higher up the pecking chain, fighting a decision can have
consequences. Authors will back down from doing that unless the
stakes are worth it. Anywhere in the chain, we might see
participants turning off as the situation becomes frustrating or else
they'll settle for the lesser evil. The latter is more common as we
move up the chain as it's generally viewed as a lost battle beforehand.
Regards,
-sm