RE: [wg-c] Re: IP/TM Concerns & New GTLDs

["Roeland M.J. Meyer" <rmeyer@mhsc.com>]

> Milton Mueller
> Sent: Sunday, August 01, 1999 10:39 AM
>
> Operationally, this is an open and shut case. 47 TLDs were
> added to the root in 1997, 31 TLDs
> were added to the root in 1996. That means that we already
> have experience, with less advanced
> processors and networking technology, with the addition of 78
> TLDs in two years.

> The IETF claims that you have to get up to 10,000 new TLDs
> before they say they start to worry
> about stability. The IETF is cautious to begin with, but
> let's be conservative and not go over
> 5,000, OK? Jon Postel himself proposed adding 150 in three years.

Current in-house market studies, conducted separately by IODesign, MHSC,
and (I believe) Iperdome, seem to indicate the current maximum market
size to be under 2000 gTLDs. The limiting factors are not technology,
rather they are the business processes and resources that are deemed to
be required for successful operation, as a business, for a new gTLD
registry. These factors also effect the maximum rate at which these TLDs
may be deployed. Each study assumed a different base-line of required
functionality and business stability. They were also remarkably similar.
Non-commercial enterprises, not having the resources of the commercial
entreprises, are even more restricted.

ICANN accreditation prolicy was not considered in those studies. IMHO,
in some ways they are arbitrarily draconian and in other ways they are
incomplete.

> There is *no* operational reason for concern about TLDs when
> we are in the range of 100-500.
> None.

Agreed, if one defines "operational" to be restricted to that of
"server" operations and data center operational guidelines. From the
business operational context, there are other issues. It is certain that
a quantity of under 10,000 new gTLDs will not effect the software
unduely. In fact, mostly those concerns are right on up there with
warnings that our Sun will collapse to a Red Dwarf someday. It is
equally certain that we will not have a gTLD growth-rate at anywhere
near one-tenth of the scale of the domain name growth, under
COM/NET/ORG. While the pent-up demand may be great, it is not that
great.

> Adding 100 does not change the status quo significantly.
> There are already 250.
>
> Kent Crispin wrote:
>
> > There are
> > strong operational arguments against immediate creation of a large
> > number of gTLDs -- adding one new gTLD will clearly not change the
> > status quo much; adding 100 will change the status quo a lot.  Given

It is argued that the status quo is exactly the reason that we are all
here. It needs to change, it will change, the remaining question is how
it will change. We, the Internet community, have been dead-locked, on
this issue, for over four years.

Deploying one new commercial TLD registry, at current estimates (I do
this for a living), would cost $2MUS to $4MUS, including data center and
staff, not including $2.9MUS for registry software
development/deployment. Operationally, costs are expected to be
approximately $1.5MUS annually. More importantly, the start-up time
requirement is a minimum of 8-10 calendar months.

Deploying 100 new TLDs, simultaneously,  would then consume, at
best-case, $490MUS over a period of 8 months. This is in the face of a
market that is proven to yield no more than $90MUS per annum, in gross
revenue. In fact, these business concerns are the main reason the market
studies have yielded such a low number of new gTLDs. It is simply not
possible to fund them all in such a short time period. Even the ccTLDs,
that we already have, are having trouble getting funded for full
deployment.

I hope that I have made it clear that technical considerations at the
least of the problems. Indeed, I consider them to be red herrings.

--------------------
Roeland M.J. Meyer, CEO
Morgan Hill Software Company, Inc.
http://www.mhsc.com/
mailto://rmeyer@mhsc.com
--------------------
Lead; Follow; Get out of the way.
     ... pick ONE!