Submitting Smith//Condorcet to the IETF

[Rob Lanphier]

For those who know how engaged I used to be in this list and wonder what
the heck happened, I can tell you it was because my job pretty much
devoured the time I would have spent participating here.  My work has
largely been irrelevant to the concerns of this group, but it just dawned
on me a very valuable connection between the two.

I've spent the last year and a half working for RealNetworks[1] as a sort
of computer industry "standards politician", working with the Internet
Engineering Task Force (IETF)[2] and the World Wide Web Consortium
(W3C)[3].  I've learned a lot about the workings of both and how there's
interesting politics that occur that have little to do with Capitol Hill.

Basically, both are a new breed of standards body that have been created
as a consequnce of the Internet.  They both operate on the basis of
consensus, with the IETF even having an unofficial (?) motto of "rough
consensus and running code".  Unfortunately, both have somewhat clumsy
ways of measuring consensus because, to date, there haven't been very good
ways of doing this.  Furthermore, many of the people who participate have
experience with other standards bodies where voting is used as a weapon to
slow things down.

The IETF has a process where one can submit a draft that is relavent and
useful to the Internet community, and they will make it available for 6
months while the draft is being discussed.  Depending on the process, this
draft can eventually be published as a "Request For Comments" (RFC), which
means that the IETF feels that this is a document that is at least good
enough to criticize (and are widely regarded as IETF standards, though
many are really FYI, and others are merely on track to become official
"Internet Standards" as approved by the IETF).  

This is where the work of the folks on this mailing list becomes useful. 
I think that what we've come up with Smith//Condorcet is something that is
useful in all forms of politics, including Internet standards politics,
and it's worth submitting to the appropriate working group.  We've already
gone through an IETF-like process in coming up with this method (rough
consensus among a self-appointed panel of interested parties), and so some
of the hard work is done.  Now it's a matter of describing that work in a
document, summarizing its importance to the Internet community, making a
case that something like this needs to be standardized, and moving forward
with it.

In looking at this, my hunch is that there may even be better than even
odds that something like this, positioned correctly, could actually
achieve publication as an RFC, which would be a *huge* step forward in
credibility, since the IETF is an internationally recognized body known
for putting out important standards.  This is obviously outside of the
normal problem domain of the IETF, but I think the appropriate working
group could be found to submit the draft to.

There would be *a lot* of work involved here, but this is one area of
politics where we wouldn't have to deal with the usual problem of
pointy-haired politicians who don't understand what we are talking about
and wouldn't have the patience and/or intelligence to sit down and
understand this.  Members of the IETF are usually very patient with folks
who present an intelligent, well-reasoned solution to a problem of import
to the IETF.  That's the key though "of import to the IETF".  This would
be no mean feat, and is the biggest hurdle by far.  I'm confident that if
the IETF could be convinced that this is an important problem to solve,
the right solution would emerge.

All that said, I'm still buried in work, so the chance of me having much
time for this is pretty close to zero.  However, I'd encourage folks to
think about this proposal and take a look at the IETF web site:

http://www.ietf.org

Specifically, look at the Tao of the IETF, which provides great
background into the IETF:
http://www.ietf.org/tao.html

One important thing about the IETF is that there isn't a lot of patience
for those that don't possess a deep understanding of the culture.  This
takes time to develop, and those that walk in with big ideas but don't
understand how things work will quickly be rejected.  Quiet study of the
organization is key.

This would have to be something for which a new working group was formed,
which would be extra work and extra reason for something like this not to
happen.  I've got several ideas on how to approach something like this,
but nothing in my head right now.

Anyway, I think the opportunity to make a difference here is much higher
than it is in many other areas, and thus is worth pursuing if the goal is
to actually see all of these theories that we've talked about used in
practice in a high profile way.  The potential for success here is really
great (relatively speaking).


Rob Lanphier
robla at eskimo.com
http://www.eskimo.com/~robla

Footnotes:
[1]  RealNetworks (http://www.real.com)
[2]  IETF (http://www.ietf.org) (For the interested, I worked on RTSP:
http://www.real.com/rtsp)
[3]  W3C (http://www.w3.org) (Also for the interested, I also worked on
SMIL: http://www.real.com/technology/smil)