Scott W Brim wrote:
There are occasions when limiting the number of deployed solutions is
very good for the future of the Internet, and in those cases, pushing
for Foo even when Bar is just as good is quite legitimate.
Sure, but I think some of these things ("good", "legitimate")
are unknowable. In midcom we agreed to reuse an existing
protocol, went through the process of identifying objective
criteria to evaluate existing protocols against, did the
evaluation, and ended up with a protocol that's nearly
universally unpopular. It's left us with a midcom standard
that's unlikely to be deployed and leaving open the door
for a raft of alternative midcom protocols which may be
deployed by their inventors but are unlikely to be standardized
by the IETF because we've already got a midcom standard.
Intentions can be the best and methodology can be rigorous
(or a reasonable facsimile thereof) and still produce results
that are less-than-useful. My preference would be to put
more trust in working groups for this kind of decision, since
it's the people working on the stuff who will best understand
the tradeoffs between the inefficiencies in creating more
protocols to support and the inefficiencies in making protocols
do things they were not designed to do. The kind of decision
you're talking about is an economic one as much as (or more
than) a technical one, and I don't think the IESG is the
right place to put the locus of that kind of decision-making
even while I do agree they need to guide it.
Melinda
_______________________________________________
Ietf@xxxxxxxx
https://www1.ietf.org/mailman/listinfo/ietf