Re: [gnso-vi-feb10] RE: Proposed draft message to ICANN staff

[Eric Brunner-Williams <ebw@xxxxxxxxxxxxxxxxxxxx>]

Volker,

I wrote you, Jean-Christophe, Michele, Stephane, on the 14th, about
this very presumption:

> ... It will make no economic sense
> for a new gTLD registry to violate their obligations if it will cost
> them their TLD at the end of the day.

At Cairo, Marcus Faure and I met with Doug Brent. We (CORE) observed
that while the application cost penalized applicants that present no
complications to an evaluator and subsidized applications which do
present complications, the greatest single threat to registry success
was the recurring cost, then $75k, annually, for contract management
costs. The recurring cost was subsequently reduced to $25k.

ICANN's record of registry contract enforcement concerns me. At the
Rome meeting the SiteFinder "registry service" had been in production
for months, breaking NXDOMAIN checks globally, wrecking spam filtering
based upon NXDOMAIN returns and causing ISPs to deploy patches to
their caching nameservers to correct this problem, introducing an
alternate authority for the COM and NET namespaces. I came "this
close" to pounding on my table with my Birkenstocks a la Nikita
Khrushchev while telling Vint and Paul and their entourage doing the
Tuesday Constituency rounds that they screwed up letting Verisign
break the net for profit for months.

The .travel registry hasn't been shut down.

And for better or for worse, we've reduced ICANN's per-registry
oversight budget from $75k to $25k.

When is the "end of the day" for an applicant which intends to violate
its obligations?

Is it before or after it has captured the registrar and the registry
margins for the first 100,000 names? This is the very point I made in
a note earlier to the list, the ROI is pretty good for the first
quarter or two of operations, and much less attractive there after. If
100,000 registrants are going to bring breach of contract actions
against the registry for violating its obligations, why should ICANN
accept the registry's contract liabilities?

My prior questions remain unanswered. For the list, they are if the
Registry Constituency resists any attempt by others to create
liability within the registry contract that is in excess of US law,
the subject of the PDDRP discussion on the temporary drafting group
mailing list, then what remedy is there other than prevailing in an
antitrust claim in a competent court?

Prior note below the ==== bar.

Cheers,
Eric

===================

Volker, Jean-Christophe, Michele, Stephane,

The proposal by Jean-Christophe Vignes (EuroDNS), Michele Neylon
(Blacknight Solutions), Stéphane Van Gelder (INDOM) and yourself
(Key-Systems) appears to rely upon a complaints process to detect
problems that may arise from either the full financial separation
between registrar and registry entity failing or being breached
intentionally, or the functional separation, and you provided the
example of information firewalls, failing or being breached intentionally.

If I have understood you all correctly, I want to point out the
discussion on the temporary drafting group legal list. In that list, a
proposal to define "registry abuse" as "willful blindness" is being
discussed. The proposal is from the IOC and other intellectual
property holders, and NeuStar is arguing that the standard which is
proposed goes beyond US law and creates new liabilities for registry
operators.

My understanding (para 1, above) is that your proposal is cross
ownership in is allowed in both the registry contract, in excess of
the general range allowed in the 2001 and later contracts, and the
registrar contract.

I don't understand the presence of "SRSU" in your condensed version,
not because the type doesn't exist, but because I don't see how full
financial separation between registrar and registry entity or
functional separation exists interestingly, particularly this "even to
the exclusion of other registrars (sale of domain names does not apply)".

I also don't understand the "SRMU" point, again, not because the type
doesn't exist, because you propose "very restricted to prevent
circumvention of the registrar system", but because I don't see how,
unless the word "allocation" rather than "sale", distinguishes this
case from the existing standard or community-based types.

So assuming that your proposal is not just for various SR subtypes,
and is for standard and community-based types, my limited
understanding of your proposal is that financial or functional merger
between the cross owning registrar and registry entities may be the
basis for for some complaint, if discovered by a third party lacking
access to the merged financial accounts or functions, and if the
complaint is correct, some penalty will be imposed.

So far, so good, ignoring all my concerns about SRxx not existing yet,
or my concern that a limit on cross ownership prevents the need to
find or prove or punish financial or functional merger between the
cross owning entities.

However, assuming that ICANN is no different as an enforcer of
contracts in the future, and knowing that NeuStar, and possibly all
but the sponsored registries, are asserting that ICANN SHOULD NOT
create any liability for conduct, such as the "willful blindness"
conduct, which is in excess of US law, how do you four imagine the
successful complaint and penalty process will take place?

Fundamentally, how is your proposal any different, assuming that the
generic registry interests carry the PDDRP issue in the tdg-legal
list, and no additional liability is created by ICANN for "registry
abuse", from existing US antitrust law?

If the complaining party can obtain a favorable ruling in a court for
an antitrust claim, then the court determines the penalty.

Is that what you are proposing, that where a third party, with legal
standing to bring an antitrust case in a competent court, that
completely defines the protections your policy proposal will provide,
for merged interest or function, to the detriment of registrants,
registrars, competing registries, and all other parties?

This is a long note, and I appreciate the courtesy of your reading it,
but are you all utterly adverse to proposing to the May 31 Board
meeting that continuity with the general limitations on registries is
a better choice for the current round of applications than either
"zero cross ownership" or significantly less restrictions than those
registries currently operate under?

Eric