RE: [registrars] OUTCOMES REPORT OF THE GNSO AD HOC GROUP ON DOMAINNAME TASTING

["Paul Stahura" <stahura@xxxxxxxx>]

John, although you may not know BulkRegister still has a membership fee
pricing structure; everyone on this list knows many registrars have a
variety of volume discount structures, bundling structures, business
models etc.  It's the essence of what ICANN has successfully brought to
this space: competition, variety, choice.

Any registrar can sell 10,000 names for $0.07 each (a total of $700) and
be assured of payment of the $700 (but not only that, actually get paid
the $700), and activate all 10,000 names.  Then it can delete 9,900 of
them (before or after the AGP).  I do not see how this is in any way
violation of 3.7.4

Ross, I know you are not advocating ICANN mandating what prices
registrars or registries should charge for domain names.


-----Original Message-----
From: John Berryhill [mailto:john@xxxxxxxxxxxxxxxxx] 
Sent: Wednesday, October 03, 2007 2:24 PM
To: ross@xxxxxxxxxx; 'Registrars Constituency'
Subject: RE: [registrars] OUTCOMES REPORT OF THE GNSO AD HOC GROUP ON
DOMAINNAME TASTING



>The second a registration gets transmitted to and accepted 
>by the registry as a valid transaction, a fee is payable to the 
>registrar. The clause in question clearly states that the registrar 
>must be satisfied that the RNH will pay for the registration and that 
>such payment is final and non-revocable. In other words, a registrar is
> not permitted to issue a refund to a registrant for cancellations made
>during the AGP according to the terms of this contract.

Again, this is the hazard of "other words" rather than the words written
in
the clause in black and white.

3.7.4 Registrar shall not activate any Registered Name unless and until
it
is satisfied that it has received a reasonable assurance of payment of
its
registration fee. For this purpose, a charge to a credit card, general
commercial terms extended to creditworthy customers, or other mechanism
providing a similar level of assurance of payment shall be sufficient,
provided that the obligation to pay becomes final and non-revocable by
the
Registered Name Holder upon activation of the registration.

First, "Registrar shall not activate any Registered Name unless and
until it
is satisfied that it has received a reasonable assurance of payment of
its
registration fee."  What this states, in the negative, is that the
Registrar
is to have a reasonable assurance of payment, defined as "general
commercial
terms extended to a creditworthy customer", before activating a name.

Second, the obligation to pay becomes "non-revocable by the Registered
Name
Holder" upon activation.  You are reading the words "by the Registered
Name
Holder" out of the clause, and replacing it with "non-revocable by
either
the Registered Name Holder or the Registrar."

Third, and backing up, it's not clear how you are reading "its
registration
fee".  The word "it" appears three times in that sentence, and each time
"it" refers to the Registrar:

"Registrar shall not activate any Registered Name unless and until it is
satisfied that it has received a reasonable assurance of payment of its
registration fee."

You are reading the first two "it"s as "the Registrar", and then
switching
the last "it" to "the domain name".  You implicitly assume "it" means
one
thing the first two times, and then a different thing the third time.
Why?
(and "because I want it to mean what I want, is not a sufficient reason
for
changing the meaning of a pronoun in mid-sentence).

It raises the question of "what is the Registrar's fee" for the domain
name
in question, but it certainly doesn't refer to the Registry fee, or any
other defined fee than what the Registrar is charging.

There are Registrars who provide free domain names as an incident to the
purchase of other services entirely.  I believe there are at least two
incorporation services that provide them incident to incorporation
services,
and it is common to provide them incident to hosting services.

So, Ross, if I am running a "buy two get one free" domain registration
promotion, and if the customer has paid for two, then under your reading
of
the RAA, have I received reasonable assurance of payment for the third
domain name, or not?  No, I haven't received one red cent for that third
domain name.  The point is that I have received assurance of payment of
MY
fee for that third domain name - which was conditioned on payment for
the
first two.

BulkRegister used to have a scaled price structure under which you paid
(a)
a fixed membership fee, and then (b) a variable fee dependent on how
many
names you registered.  The point of the clause is that - whatever terms
the
Registrar is offering - the Registrar must be satisfied that the
Registered
Name Holder is and will be bound by those terms on a basis that is not
revocable by the Registered Name Holder at the time the domain name is
activated.  Further, that satisfaction may arise from the
creditworthiness
of the Registrant as determined by the Registrar.

I'm not trying to be difficult here, but someone was paid an awful lot
of
money for doing a bad job of choosing pronouns if that sentence really
means
what you think it means.  This is not the worst-drafted section of the
RAA.
If my workload lets up, I could write a dissertation on the positive
impenetrability of other sections of the RAA.

>Whatever. Verisign can and should implement the restocking fee, 

The silence from Verisign on this issue is fairly loud.

The registry is the single control point for
limiting/stopping/controlling/banning domain tasting.

>From the "bucket of bad analogies"... there is a reason why, for
example,
birth control methods have historically been oriented more toward one
gender
than the other, and it is a simple question of numbers.  Controlling
what
happens to one cell is a numerically more efficient strategy than trying
to
control millions of others.

There are a number of registrars who all may read the RAA in various
legitimate ways.  There is, for each TLD, a single registry.  Piling
more
potentially ambiguous language into the RAA with yet-undiscovered
unintended
consequences simply seems less efficient than restricting registry
operation
in this context.