Plenary decision?

[Theodore Ts'o]

On Fri, Jul 11, 2003 at 05:32:38AM -0700, Hallam-Baker, Phillip wrote:
> I seem to recall raising the issue in a plenary. At no time was I
> aware that the plenary was the decision maker for the issue in
> question.
> 
> This was I believe because of your reply 'There is a process", I
> remember that part of the reply because you neglected to say what
> the process was.
> 
> So here we are in this forum and I find out that oh, the plenary was the
> decision making body!

Phill,

First of all, thank you writing a much calmer and much more thoughtful
e-mail.

It may very well be that that the IETF has grown too large for our
earlier, more informal mechanisms of governance.  Just as in working
group chairs have the power to determine rough consensus, and are
charged with making good decisions about when there is rough
conensensus, and when people are raising technical objections of the
form, "it can't possibly work", versus "there are engineering
tradeoffs here, and I happen to strongly prefer door A over door B"
(where objections of the first kind must be taken care of as opposed
to objections of the second, which you try to reduce but ultimately is
not reduced down to zero because "rough consensus" doesn't mean
unanimity), at least in my view of the world, that's what the IESG is
charged with doing, on a larger scale for the entire IETF.  

This should be nothing new; it's in the TAO of the IETF, and in the
newcomer's briefing, but since you seem to be have some doubts about
what the process is, I thought it would be put it out explicitly.

For questions like whether or not we use non-ASCII documents
exclusively, currently, the process is the RFC Editor decides, with
strong input/direction coming from the IAB and the IESG.  On the other
hand, if they don't respect the community consensus, there are many
checks and balances, including (a) throwing the bums out, (b) the
appeals process, (c) simply going to other standards bodies.

That being said, the topic of non-ASCII RFC's comes up every couple of
years, and at every IETF plenary that I can remember, the number of
people who come up to mike who bring up advantages of ASCII and
disadvantages of other formats generally run about four or five to one
in favor of ASCII.  So I would argue that given the current decision
making process, the IESG did absolutely right thing.  Furthermore,
even if we believed that democracy was a magic wand that automatically
produced technically superior standards, it seems pretty clear that
ASCII would probably win out.

> If the plenary is a decision making body as you are implying then why does
> it keep no minutes?
> 
> It is possible that Harald is correct in his assertion, but I find no 
> record in the minutes because there arn't any.

It is not a decision making body, but it is the best way of measuring
the general consensus of the community.  Currently, we don't take
minutes, but rely on memories and the IESG to judge that consensus.

Your complaints about the IETF's about process, and who are the
decision-making bodies, and your accusations of ruling cliques I think
goes to the heart of the matter, which is fundamentally one about how
we govern ourselves, and whether we should be using a very formal
process, or an informal one.

I view this as an engineering tradeoff.  There are certainly
advantages in using a very formal, structured meeting format where
decisions are made via votes and minutes and Robert's Rule of Order.
There are also advantages to using a system of judging rough consensus
and an appeals process as a check and balance.  Both schemes also have
their disadvantages.

Some people feel much more comfortable with standards bodies that use
a very formal process.  Others have noted that in the past, bodies
that have gone down this past have sometimes gotten mired in
bureaucracies, and result in people (and large corporations) who are
really good at manipulated said process, but who are not necessarily
good at producing good technical results.  Certainly there was a time
when our lightweight decision-making process was compared favorably to
the process used by ISO and more formal standards making bodies.  It
is therefore ironic that one of the reasons why we're engaging in this
problem-statement process is because we seem to be falling prey to
many of the problems that we once caused a certain feeling of
superiority over these more formal standards bodies.

Is the answer to become more like these formal standards bodies?  That
seems to be what some people are suggesting.  I am hoping though that
there is something about our process of rough consensus and running
code which is unique and better than a heavy-weight system based on
process and votes.  That is, I hope that while we are fixing some very
real problems, that we don't end up throwing out the baby with the
bathwater.

On the other hand, it may be that that real problem is that smaller
bodies are just naturally more lightweight (regardless of what scheme
of polity they use), and that growth inevitably causes problems.  In
that case, the fact that some groups are trying out new standards
organizations might not necessarily be such a terrible thing.

						- Ted