DT members say "no"

[Bob Stewart <[e-mail]>]

>There's a simple solution to this problem.  Careful discussion followed by
>humming/show of hands/whatever, at the WG meeting.  If the change is hummed
>in after detailed debate, then that's it.  If the problem is that the DT
>folks have good arguments against, such that no changes get hummed in,
>well, good show for the careful work by DT, no?

That's exactly what we've been doing in SNMPv2.  We still get the occasional
accusation of cronyism and railroading, not to mention issues that keep
returning and returning and returning.  Certain types of favorite solutions
always have proponents, because they just seem so right, whether they are or
not.  I believe we just have to live with this.

Working group chairs have to go out of their way to be fair, but when there is
disagreement, someone will win and someone will lose.  A small group who's
worked long and hard and done a good job of understanding the issues will win
most of the time, giving the appearance of favoritism, when in fact they won
because they worked harder and were more right and more convincing.

Starting from a comprehensive design is good.  Having strong proponents of
that design is good.  It helps maintain focus and ward off too much
ornamentation by the well-meaning but wrong.  Unfortunately, it will also
sometimes ward off repair by the right but less convincing.  All things
considered I'd rather have that than design by committee.

Much of this comes down to the persuasive abilities of individuals.  The best
prepared, best presenters will win most often.  Even if you're right, you have
to be able to present your arguments convincingly.  Process will not fix that.
Working groups will be dominated by a small number of people, whether they're
a recognized subgroup or not.  It's the way groups work.

Process will replace trust, good intentions, and community spirit only at the
cost of quality and efficiency.

	Bob