Re: [gnso-vi-feb10] Another angle on allowing VI/CO

[Eric Brunner-Williams <ebw@xxxxxxxxxxxxxxxxxxxx>]

With the 15% cap, assuming .bankok was not satisfied with the local
ICANN accredited registrars, and any other ICANN accredited registrars
which passed the .bankok OT&E, perhaps because they are all lame, or
stupid, or ineffective, or charge high prices making .bankok less
competitive than it's plan requires with .asia and .south-east-asia
and ...

... then finding six other registries, where ever they happen to be,
in Saxony selling .ddr, in Africa selling .nile, ... , is sufficient
for .bankok to create, with those six other registries which are
unhappy-with-their-available-registrar-choices, a registrar these
seven registries manage, and may be happier with because this registrar.

With the cap at 5%, it takes 21, with the cap at 10%, it takes 10,
with the cap at 30%, it takes 3 plus 10% by a 4th party, with the cap
at 50%, it takes 2, and with the cap at 100%, it takes just 1.

The number of registry participants sharing the control of a registrar
that is as effective as, or more effective than, the other registrars,
is what we are picking when we pick a cap on ownership or control.

Of course, in theory, the registry shared registrar could be less
effective, but there's not a lot of motivation to create a worse-then
registrar as an alternative to existing registrars. And in theory, a
registry could make a 15% investment in a registrar 85% controlled by
an investment banker more interested in selling .com. Again, there's
not a lot of motivation to create a worse-then registrar as an
alternative to existing registrars.

So the question, as Alan just pointed out, is where in this range of
equal participants in the control of a registrar created to provide
the first registrar, or to provide competition to existing registrars,
for the same number of registries, is there either "more risk" or
"less known"?

What is the risk of adverse behavior by the registrar with 20
independent, equal votes forming its decision making process? How
about 10? How about 5? How about 2? And if only 1?

And, as Alan has pointed out, there are proposals that include, or
offer as an exclusive alternate, to ownership and control caps, other
exception mechanisms to Recommendation 19. Some use numbers, one uses
time. And here to, the issue with any cap on number or years is what
does a cap provide, as an answer to either the "more risk" or the
"less known" question.

Eric