[ga] Re: .info LR2 process and failure of ICANN to heed warnings

["Richard Henderson" <richardhenderson@ntlworld.com>]

Hi Jim

I've done an analysis of 250 of the most popular names, to get a feel for
the way the names were distributed, and to see how names were distributed to
different registrants within one registrar's list.

My views to date are:

I actually think the execution of the randomised selection itself seemed OK.
There is a rationale and spread of names, once external factors (like the
length of a registrar's queue) is taken into account. So I don't tend to
think there was any problem at the actual randomisation and selection level.
(Having said that, I am not aware whether there was any independent watchdog
supervising theprocedure, and in the interests of openness and transparency
there certainly should have been.)

Turning to the names distributed to specific registrars, and the way
different registrars compiled their lists:

Here there were serious problems. You will have recognised yourself that,
although RegTek's conduct and treatment of the public were exemplary, you
only got 1 really top name out of your 172 successful registrations. The
same thing happened to many other registrars who opened their lists fully to
the public.

This was because of the very short and exclusive lists submitted by certain
registrars, which resulted in the "hi-jacking" of many of the very best
names. Clearly, if you only submitted 5 or 6 applications (like Signature
Domains) you were likely to get many of those in the first 5 or 6 rounds.
This is what happened - they got 3 very good names... for all I know they
only submitted 3 names, as they got no other names at all. In contrast, if
you submitted (say) 1500 names, your very best names were probably going to
emerge in round 70 or round 135 and if unlucky in round 1499. By then, of
course, most of the best names had been snapped up by people like Signature
Domains.

I refer to Signature Domains specifically because of their track record of
submitting short exclusive queues. In the .biz2B they only got 10
registrations and every single one was registered to their Business Partner,
Joshua Blacker. He admitted his part in this to me over the phone. It
totally disadvantaged registrars like you, who opened their lists fairly to
the public.

In LR2, the following registrars submitted very short lists or caused me
concern:

Signature Domains: 3 "top 250" names and 0 other names
ATech: 2 "top 250" names and only 4 other names - I also noted that ATech
was founded/owned by Jason Hendeles and he applied for himself for one of
those top names (adult.info)
AAAQ: submitted a very short list and got 11 "top 250" names and just 2
other names... however, to be fair to Jeff Field, they did open their list
to the public
Sitename: 17 "top 250" names and just 9 other names - I think they shifted
most of their applications to the GalComm list and saved Sitename for
special names - including, Moshe Fogel got domains.info - which I think was
poor judgement, an Afilias Director using a short (exclusive?) list to get a
premium name, when Afilias themselves share responsibility for the name
being hijacked the first time round. I don't think Moshe Fogel should have
done that.
Xin Net Corp: 7 "top 250" names - and just as in the .biz2B, all of their
registrations appear to have gone to Lin Zansong. I disagree with a
registrar "selling" a list to an individual, and excluding the Internet
public, because I think that is not a fair and open distribution of the DNS.
I warned Dan Halloran and Vint Cerf in advance that these things would
happen in LR2 after they occurred in .biz2B, but 68 days after sending the
e-mail (and repeated copies) Dan Halloran has yet to acknowledge receipt.
ICANN and Afilias chose to preside over this process and just let these
hings happen.
Nordnet: 11 "top 250" names and just 29 other registrations. This was a very
short list, and out of those 11 top registrations, 8 of them went to Francis
Piet himself at Nordnet. It appeared to me, that they made application as
unattractive as possible to reduce the size of their list, then filled it
with best names for themselves. You can distinguish their own registrations
from their customers, because customers had to apply for 10 years
registration, but Piet's names were only registered for 2 years.
Domain Investigator: The size of their list was fair, but it was made up of
what seems to me on my analysis to be ONE person's applications, namely Ross
LeBel (who also applied through many other registrars and did very well).
Basically, Domain Investigator's list seemed to be closed to the rest of the
public.
BondiLLC: 7 "top 250" names and just 26 other pretty good names. ALL these
names were registered to Marco Publishing Corporation ( Marc Ostrofsky?) who
appear to have "bought" the list. This was a re-run of .biz2B where Marco
Publishing again used BondiLLC as if it was its own personal registrar.
Again, I warned Dan Halloran and Vint Cerf that this would happen but they
did nothing.
EPAG: They submitted a normal-size list, but it was dominated (exclusively?)
by Alexander Schubert - the same issue arises: should registrars deal just
with one person, or should they be more open?
Multiple Lists: some registrars submitted multiple lists, which enebled them
to channel favoured/special names into one shorter list, and channel the
majority of names into a separate list. Moshe Fogel appears to have done
this with Sitename and GalComm although I stand to be corrected. Sitename
got 26 successes. Galcomm got 167. This fairly well indicates the relative
size of lists. The fact remains that Moshe Fogel had twice as many lists as
you or most other registrars, and is that fair? How can you compete fairly
if a competitor is able to do this and manipulate his lists perhaps?
Directi used at least 3 lists. Their main Directi list got 218 successes.
They also "bought" Global Media's list and submitted fewer names through
them, getting 86 successes. Finally, they used part of BookMyName's list.
I'm not criticising Bhavin for using his business instincts to do the best
for himself and his clients (which included me): but I feel ICANN and
Afilias should have realised that some kind of rules were needed to protect
other registrars like yourself.
Register.com were the most dramatic multi-list applicants. They submitted 4
lists. Their ordinary Register.com list was enormous, getting 890 successes.
They also used Virtual Internet's list for some of their clients and got 155
successes through them. They used Best Registrars and got 76 successes
through them. And they used Net Searchers and got 26 successes through hem,
usinng a much shorter and more exclusive list. The outcome for Register.com
was that they gained 19 "top 250" names (compared to 1 for RegTek). This is
a fair indication of the uneven playing field you were playing on.

My view is that Afilias had a duty to guarantee no undue advantage was
gained by individual registrars. It was disappointing to me that the Lubsens
also operated a single application interface but a dual list, with
DomainBank and DomainPro. As CEO of Afilias, I do not think Hal Lubsen
should have been associated with this kind of method, and I'd be very
curious to know what criteria he used to decide which names went on which
list (since all applications from either registrar were processed on the
same webpage).

So in conclusion, I believe ordinary, honest registrars like yourself were
cheated out of many of the best names (or rather, your customers were). I do
business with you because I value integrity, but what is to stop other
customers moving their business elsewhere if you failed to get a good
success rate with the best names? I would say your business was
disadvantaged by the process, and I would say that Afilias and ICANN knew in
advance that the process would be what I'd call corrupt, but chose to do
nothing about it.

Please don't misunderstand me. I did very well out of LR2. I got some great
names (though none through RegTek). So it's not sour grapes on my part. I
just believe that the DNS is a worldwide resource, and ICANN and the
Registries have a responsibility to the global internet community to
guarantee the fair distribution of names. Clearer, more definitive rules
should be drawn up in advance, and steps should be taken to stop individuals
hi-jacking the best names or "buying" a whole registrar's list. In my
opinion, every participating registrar should have an online interface open
to the internet public. And I think registrars should only be allowed to
submit one list. I don't blame the registrars who submitted more than one
list. I blame the Registry and ICANN for allowing it to happen.

It's 68 days since I sent Dan Halloran the first warning of these problems,
in the wake of .biz2B, and raised serious concerns and questions with him.
The fact that he has never even acknowledged my mail indicates the extent of
contempt in which members of the internet public can be held. It is a
shocking demonstration of the lack of openness and transparency that exists
within ICANN. The refusal to engage in dialogue or to respond to awkward
questions reflects their negative and defensive attitudes. It is very
disappointing, and Nancy Victory ought to consider their "promises" of being
more open and transparent very very carefully, when in fact they display
such a lack of openness.

In conclusion, I'd like to say thank you. You got me precisely 0 names in
LR2. But you acted with integrity and that is more important to me. I'd say
the same about Bob Connelly at PSI-Japan, who protected his Landrush 1
customers and stood by his commitment to them. When he resigned from the
Afilias Board last autumn, in protest at Afilias's dereliction of Landrush 1
customers and what he called the "Abomination" of Sunrise, he was
marginalised as a malcontent. At the time, Roland LaPlante was playing down
the whole shambles and denying the scale of the problem. In the event
Connelly was proved right and 25% of the names turned out to be false (and
there are still many others not dealt with).

Because there is growing pressure from other Governments for more say in the
running of the Internet, and because the US Govt retains oversight, when
ICANN acts without responsiveness and accommodates corruption (what action,
for example, have they taken against the registrars who committed fraud in
the .info Sunrise - none, they're still accredited and promoted by ICANN)...
it puts the US Government (as its sponsor) in a very poor light indeed. It
doesn't reflect well on the American people either. I feel Nancy Victory
should consider very carefully whether ICANN is truly committed to openness,
transparency and accountability - or whether that is all just words.

As we are talking about the distribution of a world resource, it is vital
(in the interests of registrars, and indeed in the interests of the US
Government) that procedure is seen to be honest, decent, and open.

In this .info fiasco, it has not been.

I believe some people should be called to account.

Best wishes

Richard Henderson



--
This message was passed to you via the ga-full@dnso.org list.
Send mail to majordomo@dnso.org to unsubscribe
("unsubscribe ga-full" in the body of the message).
Archives at http://www.dnso.org/archives.html