[NA-Discuss] NA RALO response to DOC letter

[Eric Brunner-Williams]

I agree in part, and disagree in part, with John. Interlinear.

On 1/7/11 5:54 PM, John R. Levine wrote:
>> What should NA RALO's response be?
>>
>> For context, here's what Milton Mueller has to say:
>> http://blog.internetgovernance.org/blog/_archives/2010/12/3/4694980.html
>
> DOC is right, and Milt is wrong.  Milt is always wrong, so this shouldn't
> be a big surprise.

I've held that view since WG-C circa 1998, yet the Board's choice on 
VI last Fall was closer to the Milton Meuller, Avri Doria, Michael 
Palage "competition authority" model than to any other model advanced.

> Attempts to introduce "competition" into the DNS via adding TLDs have
> been, in case anyone has had their eyes shut for the past decade, a
> complete failure.  The only change that actually has made a difference is
> the registry/registrar split.

The divestiture of .org from VGRS to ISOC/AF was a competition policy 
win. It moved about 10% of the market from the legacy monopoly 
operator to some other entity.

Had the .net divestiture not be aborted, and had it gone to the same 
operator, then the market would be a duopoly.

A means as yet untried was the original SRS model of Crispin, Crocker, 
Gaetano & Langenbach, presented at IETF-43, 1999.

See 
www.icann.org/en/topics/new-gtlds/presentation-brunner-williams-19oct09-en.pdf

I presented this during the VI WG process. Had it been adopted, 
GoDaddy, holding 25% of the .com market, could become the 2nd largest 
registrar, reducing VGRS's backend revenues by a quarter, and the 
quartet of GoDaddy, TuCows, Enom and NetSol could become the 2nd, 3rd, 
4th and 6th largest registry backend providers (AF moving from 2nd 
place to 4th place), bringing VGRS backend revenues down to 49% of the 
.com backend.

To my mind this was, and remains, the best choice. Remove the clever 
fiction that registries are natural monopolies and allow registrants 
to select which backend operator of the ".com" namespace should be the 
data archive, with all backend operators of the ".com" namespace 
pushing updates to the .com zone publication function, for about 
$1/domain name year.

$5/unit/year under the hood leaving "Verisign" to publish the zonefile 
for $1/unit/year would at least transform the monopoly market into 
several competing verticals with "registrar-to-backend" slices of the 
.com market.

> Look at .INFO and .BIZ to see whether clones of existing TLDs add
> competition. Look at .AERO, .COOP, .JOBS, .TRAVEL, and .MUSEUM to see
> whether topical TLDs add competition. Look at .ASIA, .CAT, and the
> rebranded .LA to see whether geographic TLDs add competition. Or look at
> .PRO and .TEL and .MOBI to see whether service-specific TLDs add
> competition.

Agree in part, disagree in part.

None of the .info,.biz,.name,.pro and the later 
.asia,.jobs,.mobi,.tel,.travel have introduced competition, and 
without the .org revenues, Afilias would have exited the registry 
market, as RCOM did when it abandoned .pro, and as GNR did when it 
abandoned .name. Similarly, without the NANPA revenues Neu{Level,Star} 
would have exited the registry market after the initial failure of .biz.

However, it is too early to tell for .coop and .cat, and the errors of 
2001 that harmed .aero, .coop and .museum may still be reversible.

These sTLDs may obtain significant market share for their intended 
service populations (.aero probably excepted).

> There was a real issue about non-ASCII TLDs which appears to be addressed
> by the IDN fast track.  What's left?  At this point the only justification
> for adding thousands of TLDs is that ICANN always assumed they would, and
> they and their hangers on will make a lot of money.  If the argument is
> that new TLDs will add meaningful competition, nobody's offered a
> plausible explanation yet of how that will happen.

Agree in part, disagree in part.

The growth of the .cn ccTLD is an exception to the .com is the only 
commercial market claim, and within the .cn market are the  .中国,.公 
司,.网络 (CNNIC) and “.政务” and “.公益” (CONAC) segments.

It is premature to conclude that IDN TLDs will not displace a registry 
run by a private company located in Norther Virginia, within specific 
market segments.

After five year of operations, just over 50% of all .cat registrations 
have no "parallel" .com registration, supporting the theory that .cat 
has achieved parity with .com in the Catalan market.

VGRS intends to apply for between ten and twenty IDN forms of "com". 
The VGRS planners are willing to spend $4m to $8m in application fees 
alone to ensure that they are the early, and only, adopter of "com" in 
several IDN markets. Clearly they see risk of not being first, and 
some opportunity in being first.

Granted, this isn't the thousands of flowers advocated by Milton.

Eric