Re: [ALSC-Forum] The Road System

[Sandy Harris]

Bret Fausett wrote:

> So what does scale? Our existing governments already have in place effective
> means of representation.

**Some** of them have, but there are quite a few dictatorships, oligarchies,
theocracies, ...

> If ICANN can't scale to allow effective, elected
> representation from the user community, then perhaps it ought to abandon the
> idea of "self-regulation" altogether and return the decision-making to the
> governments of the world, where there is already an established
> infrastructure for elected representation.

In many countries, there is no such infrastructure. Even in democratic nations,
large parts of gov't may not understand the net or may be involved in various
actions contrary to the interests of netizens.

> Make no mistake, I would prefer to see the ICANN ideal of self-regulation
> and self-representation succeed, but in a study where "no question is
> off-limits," you ought to consider the ramifications of failing to find an
> appropriate representation structure for the user community. If ICANN is
> incapable of scaling to meet that challenge, one very good option for the
> ALSC to consider is that the technical coordination functions assumed by
> ICANN be transferred back to the government.

The last major gov't-backed effort at technical co-ordination was the ISO
(International Standards Organisation) work on standards for networking,
their "Open Systems Interconnection" project, which never really got beyond
an attempt to specify a "Reference Model" for networking.

That was a disaster. See Padlipsky's "Elements of Networking Style" for
a detailed critique. Some chapters are available as RFCs 871 to 875.

So I'd say we have hard evidence that, at least for technical issues,
the IETF model ("rough consensus and running code", working groups
based on mailing lists, membership open to anyone, "leave your corporate
affiliations at the door" and focus on technical merits, ...) is the
only demonstrated way to design and build an Internet. Also, that large
standards organisations dominated by gov'ts and phone companies can get
it horribly wrong.

So, to the extent that the issues are technical, I don't think there's
any question that an IETF-like approach is the way to go. The question
is whether such an approach can work for the political and business
issues and, if not, what is the alternative.

Another problem is how do you assign number of representatives?

If you do it by national population, I think China and India have a majority.
Canada and California get about equal representation.

If you do it by number of people on the net, then the US, Japan and Europe
dominate for now, but various other countries may take over in a few years.
Again, Canada and California get roughly equal representation. 

If you do it on technical criteria -- contributions to net.development,
backbone bandwidth, etc. -- then it's the US, Japan and Europe, perhaps
permanently. Canada gets a small fraction of California's representation.