Re: [wg-c] Private TLDs

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

> Kent Crispin wrote:
> 
> > Do you think there might be some demand for private TLDs?  Maybe
> > more than a few hundred?
> 
> The demand for private TLDs is not clear. But it is possible that such
> a demand will develop, and that it could pass into the thousands.
> I suspect that you will use this in an attempt to scare some of our
> more conservative colleagues into rejection of an open, registry-driven
> market for TLDs.
(...)
> The most reasonable way to anticipate the scale of demand for
> private TLDs would be to look at the number of large, privately-
> managed corporate networks. They run in the low thousands.
> Their number is declining with the rise of Internet and VPNs.

A while back, only established networks dared request their own SLD. Now
just about anyone who wants to even have the slightest "serious" look when
doing business "needs" their own domain name on their business card. (I'm
not going into debating whether it's a good thing or a bad thing. I'm just
stating that it is so, the proof being that there are 6million? .com names).
Entities requesting their own SLD started slow, but now it's really up to
speed (and accelerating).
Just about all estimations of growth of ANYTHING on the internet have been
generally 1-5 magnitudes of order too low, so why not not for companies
demanding their own private TLDs too?
Only those that have their own established network need a private TLD...
sure, same thing a few years ago for private SLDs...

Thing is that once one comapny gets their own TLD (say AOL with ".aol"),
then others will also want their own TLD, and will rightfully argue that if
X can have a privatly owned TLD, why can't they?

There's a rather large body of thought that includes developers of BIND that
says that the DNS can't handle the amount of TLDs that this would bring
(there is also a body of opinion that says this is b****hit, so take your
pick), but I won't go into that.

Let's imagine a scenario where every company can get a TLD if they request
it, and let's imagine that the situation progresses to be more or less what
we have today with ".com". Then we would cleraly need ONE organisation
dealing (as NSI deals today) with the allocation of those TLDs, or at least
a structure similar to a shared registry/registrar setup. Problem is that
(even if it didn't smash the DNS because of cache problems) the variety that
we are all arguing for when adding more TLDs would sudenly disappear. Today
we say that with only ".com" (and net/org) there is not enough. Well, if we
just chop ".com" off the end, we're back to the same problem...

Why risk the root if it doesn't bring any advantages with it?

Yours, John Broomfield.