[wg-c] The limits of theoretical debate: time for experiments

[Milton Mueller <mueller@syr.edu>]

In Los Angeles, Paul Mockapetris "mocked" the new TLD debate by observing that if
this were a physics department we would have only theoretical physicists -- no
experiments or empiricists would be allowed. R Connelly made a similar point long
ago: "One test is worth three expert opinions!"

The point being, we have long ago reached the limit of what can be achieved through
theoretical debate. One third to one half of the wg believes there should be an open
and competitive marketplace for new TLDs, a marketplace in which a variety of models
can be tried out and consumers will dictate their success. Another third-to-half
believe that ICANN must specify the model (one that just happens to correspond to
the one they selected when participating in the gTLD-MoU).

We are not going to change each others' minds on this point. I suggest that we
accept that fact and try to find ways of moving forward. I would urge Kent and
others to disabuse themselves of the idea that clever debating tactics are somehow
going to snooker the people who advocate models different from theirs into accepting
the POC/Core model exclusively and in perpetuity. We all know what is at stake and
what is happening. The initial testbed will set the tone for the future. The people
who oppose for-profit registries now will oppose them in the future. No one is
fooled. No amount of noise generated on this list is going to evade or obscure the
fact that we don't agree on this.

We are now at the point where experiments are called for. I adopt an ecumenical
approach to these initial experiments. Some of the initial "testbed" TLDs should be
shared, some shouldn't. Some should be for-profit, some non-profit. Some should be
broadly generic TLDs, some narrowly defined chartered TLDs. If problems with any of
these models develop, then that can be noted and can guide future behavior.

In other words, time to put your money where your mouth is. If you want to debate
about for-profit v. non-profit, or shared v. proprietary, let's debate about real
facts, real experiments under controlled conditions.

The contractual arrangements that set up these tests can and will define temporary
and retractable licenses. The notion that permanent problems will be created through
these temporary, delimited licenses is a scare tactic. Not a single post has been
able to document any real problem with that. No, waving the bloody flag of NSI
doesn't count. My previous post completely discredited that argument. The problems
posed by giving NSI an open-ended, uncontested gTLD monopoly for 6 years and letting
it achieve a 75% market share are not in the slightest comparable to problems that
might be created by giving a TLD registry temporary rights to operate for a 6 month
evaluation period.

(Parenthetically, there is a real circularity to the parade of horribles associated
with NSI. They start with the premise that for-profit registries are bad. Then they
argue that the NSI agreements are terrible because they allowed a for-profit
registry to continue. In other words, for-profit registries are bad because...they
are bad.)

Kent Crispin wrote:

> But there are several ways to achieve competition.  That is the
> point of having competitive registrars.

We have the opportunity to test that way of achieving competition and comparing it
to the way I propose. I am confident that my proposals can withstand that test. Why
aren't you?

--MM