From: Brian Lloyd <brian(_at_)lloyd(_dot_)com>
Neither "consensus" nor "democracy" by themselves produce good work. It is
possible for the overwhelming majority (consensus) or bare majority
(democracy) to choose the mediocre over the good. At least with consensus
a few can have a more significant effect.
There's a lot of confusion over what "consensus" means.
It's not simply unanimity or a modified form of voting,
but a process that allows all voices to be heard and
enables groups to come to decisions that most accurately
reflect the views and values of the group members. It
specifically tries to prevent a majority (or loud minority,
as so often happens in the IETF) from using their weight to
force bad decisions through. Used properly it allows groups
to come to better decisions than they would otherwise.
It's slow and tedious and requires investment in the process
from all participants, and is particularly ill-suited, I think,
for technical work.
So, as I see it, it is the few rather than the many, who do the good
work. TCP works as well as it does because of the work of handful of
people, not the democracy of the majority or the consensus of the
population of the IETF.
It's been my experience that even in the most popular (what
a concept) working groups only a handful of people actively
participate, anyway. The primary disadvantages, to my mind,
of the size of the body pool in working groups have been 1)
it's often difficult to get a seat in meetings, 2) mailing
list management becomes more time-consuming, and 3) more time
is wasted dealing with questions from people who haven't read
the documents or working group charters. I don't see any of
this as particularly onerous.
It seems as if the real question here is how to handle
controversial working group proposals. The OPES proposal
has a fair amount of support behind it and quite a bit of
objection to it. In a consensus-oriented decision-making
framework everybody with an opinion would work together to
find some mutually acceptable (not loved - acceptable)
accomodation, whether it's sending the work off to another
standards body or modifying the charter and having the
work done in the IETF. That hasn't been the ways things
have worked during my short time with the IETF - noise is
made and the IESG goes off to think about it, work directly
with interested parties, and then make a decision. Maybe
it's not that meaningful for us to be talking about consensus
or direct voting when what we've really got is a republic.
Melinda