Selecting leadership, take 2

[Theodore Ts'o]

On Fri, Dec 06, 2002 at 09:29:09AM -0500, John C Klensin wrote:
> >...
> >Does draft-huston-pact list any problems that you disagree
> >with?  Does it suggest changes that you disagree with?
> >...

Yes, and yes.

I thought John's very thoughtful note on the subject very nicely
articulated my concerns with the pact I-D, and since I wanted to avoid
the AOL "me too" posting, I haven't said so publically.  But given
that there seems to be an attempt ram that proposal through,
short-circuiting a more thoughtful process (which I suppose is what
pact proposal is all about), let me register my own serious objections
to the pact proposal.

As Martin Luther once said, mankind is like drunk trying to ride home
at night.  First he falls off one side of the horse, then he falls off
the other....  There has been a concern raised that standards are
taking too long.  So the solution proposed is to move to a "Ready,
Fire, Aim" paradigm, which seems to go overboard in the opposite
direction by sacrificing quality controls in the hope that it will
speed up the standards process.

Yes, we need to make streamline the standards process.  At the same
time, I also hear people at the open-mike plenaries asking why we
can't do a better job enforcing architecture.  (Of course, we don't
agree on what's the most important part of architecture; for some,
it's internationalization; for others, it's security; for others, it's
the extermination of anything that violates the end-to-end principle;
for yet others, it's the avoidance of patent-encumbered technologies.)
But in any case, I think we get almost as many complaints from people
wanting to the IESG/IAB to stop protocols that violate some of their
pet peeves as we do from people want to see standards go out quickly.

So I'm quite convinced that the "damn the quality control torpedos,
full speed ahead", is *not* the right answer.  It's not the right
answer for the standards process, and it's certainly not the right way
to attempt to fix the IETF standards process.

						- Ted