Re: [wg-c] Pre-sold TLDs

[John Charles Broomfield <jbroom@manta.outremer.com>]

> > Chris, are you trying to tell us that IOD has finally thought that it
> would
> > be a good idea to accept a situation where ".web" gets inserted in the
> > legacy roots even if it doesn't belong to IOD, have it setup in a shared
> > registrar fashion, be ONE of those registrars and at day0 have something
> > along the lines of a round-robin system, and have all of it's existing
> > customers go through this process?
> 
> *SOMEONE* has to run the registry.

Agreed 100%

> If this happened, who would run the .web registry?

The money to be made is (should be?)on the registry side, so from that point
of view, one answer could be "who cares?". After all, what I *think* I'm
reading here is that you want IOD to be a registrar. Yes, I know you've said
many times that you want IOD to be a registry, but I don't think that the
way things are currently going would allow ANY company to be at the sime
time registrar AND registry. If you want IOD to be a registry too, then
you'd have to split into to separate entities *now* (not some 4 year -or
whatever- provision like what was given to NSI, which in any case I *think*
has already split completely -though I may be VERY wrong-).

In any case, to answer the question, we know -as you state- that *SOMEONE*
has to run the registry, and most of us understand that only *ONE* entity
can do so (anyone arguing differently for that still hasn't presented the
tech specs for their pipedream). Something which would fly today is putting
out a series of stringent requirements (much stricter than the registRAR
accreditation due to the fact that there can only be one) under which a
registry operator could be accredited and then tender it out for bids.
Presumably it would be done at the same time the other X (6-10?) gTLDs would
be tendered, and there would be a limit on how many gTLDs a single registry
operator could handle. Cheap, backend, no frills attached, just run the
registry. And, of course, open specs, agreed upon, with a provision to
escrow the data with one (or more) other registry operators, so that in the
event that one goes out of existance, the data continues flawlessly.
I think we would have learnt through proof of usage that one MAJOR
modification to be done to the current com/net/org sharing model is to have
a centralized whois database.

My *business* advice to IOD, go for it as a registrar. You've got lumps of
money to make there as you already have a bunch of customers just waiting to
go "live" (yes, you WOULD have a handful who wouldn't get their names in on
the first round-robin, but they are just the hopefuls who want to make it
rich with a "business.com" type scenario -nothing wrong with that- and
understand they were just gambling. In any case, they would still be visible
in your CURRENT servers, so they wouldn't be any worse off than they are
today).

Yours, John.