At 13:23 19/05/2005, Hollenbeck, Scott wrote:
Ted and I have spent a lot of timing trying to assist working groups
that have been close to completing their work but have been unable to
finish for one reason or another. We've had some success in getting
work completed, but the amount of time we've spent doing clean-up work
has meant that we haven't been able to devote much time to strategic
thinking. We're hoping that one impact of the reduction will be that
we'll have more time to start looking forward.
From the two I participate to, I would suggest there is a more complex
problem. When a WG is started it is at the initiative of a small group.
Innovation cannot be shared by several persons without narrowing its
perspective. Even if the matter is broad, even if the IESG pays attention
to the charter to enlarge it, even if IAB provide guidance, the vision of
these individuals will prevail. Sometimes there is also the problem
identified by IAB's RFC 3869 about the commercial funding of R&D and the
need of more global vision and independent/larger funding by Govs.
But the matter is still more worrying, because the IETF is in fact the
addition of people interested in one or several WG and a new idea will
rarely excite people pursuing others. This means that recuitment is through
1. the less WGs the less recuitment
2. the more participants, the least knowledge/interest in Internet
architecture in the WG - and competence in the WG matters out of the scope
or of the agenda they pursue.
3. the more WGs, the more resulting confusion
4. the more confusion, the least interest from non-IETF specialists to join
the WG and the IETF.
I gave a few thinking about this and I came to the conclusion that the best
to get some real quality work was to slightly modify the working method to
keep open participation but to work on a filtered basis, stopping chatting
where the number of people prevails on the number of ideas. This can only
be organised outside of the IETF and result in external submissions. The
work is obviously not negligible, but I think it is necessary. Now the
problem is to see specialized TFs to emerge and to keep them coordinated
with the IETF. This can result in some slight modification of the Internet
standard process, formally including into the review these specialised TF.
IMHO rather than slowing the process, it could speed it up, because
parallel review by a larger number of specialised reviewevers/comments
periods we can rely on.
Ietf mailing list