[NCSG-Discuss] Closed Generics - history

[Mary.Wong at LAW.UNH.EDU]

Thanks first to all for a very bracing, thoughtful and informative
discussion on both sides, and thanks to Kristina, Ron, Milton and others
for submitting public comments that illustrated just that. 

IIRC, the ROCC was developed partly as a result of the GNSO discussions
over what was called single-registrant TLDs, and concerns over whether
these would exclude registrars (and possibly be problematic for the
non-discrimination policy). The problem with the single-registrant
discussion, IMHO, was that so many people were so focused on the
dot-brand issue - e.g. what if IBM operates .ibm solely for internal
purposes? - that the very strong likelihood that there would be SOME
dot-generic-word registries that would be operated along similar lines
just didn't get the same amount of "air time" or focus. 

That's not to say the possibility was not ever contemplated. It was.
It's inherent in the nature of a gTLD - the operative word is "generic".
 

As I said in my public comment on the issue, we had enough problems
trying to define "public interest" in the whole Rec 6 (Morality & Public
Order) debate, which led to a widening of what became the Limited Public
Interest Objection. At least in that category, one can rely on
principles of public international law and take some comfort in the fact
that a third party neutral who is a jurist in that legal field will make
the decision (whether one agrees with it or not). 

Asking ICANN to say that Registry A can't be approved because it's too
closed a registry and as such against the public interest, while
Registry B can be even though it's closed but NOT so much as to be
against the public interest, would be subjective and poor policy. And
asking ICANN to ban all closed registries (defined by ICANN when neither
trademark law nor public international law has such a definition) while
allowing dot-brands and existing generics to continue would be just as
rationally and legally indefensible. 

Bottom line - the GNSO had a chance to delve more deeply than it did
into the issue of so-called closed generics. Maybe it should have.
However, it would not be true to say it never considered the issue -
probably truer to say that many didn't think it was going to be such a
big issue for some stakeholders, especially as it was distracted during
the single-registrant and standard-versus-community discussion by the
dot-brand issue (and the vociferous trademark arguments.) 

Cheers 
Mary 




Mary W S Wong
Professor of Law
Director, Franklin Pierce Center for IP
Chair, Graduate IP Programs
UNIVERSITY OF NEW HAMPSHIRE SCHOOL OF LAW
Two White Street
Concord, NH 03301
USA
Email: mary.wong at law.unh.edu
Phone: 1-603-513-5143
Webpage: http://www.law.unh.edu/marywong/index.php
Selected writings available on the Social Science Research Network
(SSRN) at: http://ssrn.com/author=437584  


>>> 


From:  
David Cake <dave at DIFFERENCE.COM.AU> 

To: 
<NCSG-DISCUSS at LISTSERV.SYR.EDU> 

Date:  
3/8/2013 12:19 PM 

Subject:  
Re: [NCSG-Discuss] Closed Generics - history 

On 09/03/2013, at 12:19 AM, Avri Doria<avri at ACM.ORG> wrote:

> ( i really wish the NCSG list would accept my email the first time I
send it)
>
> Hi,
>
> I beleive the ROOC is only a concern if they are unwilling to use
Registrars.  If they are willing to become, or set up, a RAA registrar
and offer the same deal they offer themselves to all Registrars, I do
not beleive they need any exception.

    My reading of the ROCC was that you can't remain closed (or at
least, closed and with a usefully large number of registered domains)
while using Registrars. I guess it is theoretically possible to use the
registrar model, yet set the restrictions on registration so tightly
that no one but the registry can meet them. It would seem that such a
model would be designed for no purpose other than to circumvent any
restrictions on closed registries though, and I suspect ICANN might
similarly see look on such an arrangement with a suspicious eye. 

> As for the issue of not being permitted without being in the Public
Interest why should they be subject to a consideration that others are
not subject to?

    In this message I wasn't attempting to argue whether the policy is
the best one - only establish what the policy is. And hopefully put to
rest the argument that everyone considered this resolved in 2008 or
whatever.

>  I think the issue is that someone would have to Object to each on a
Limited Public Interest basis.  If they can argue that that particular
TLD is detrimental to the public Interest it can be disallowed.  I do
not beleive that they can be condemned as a class.

    No, I don't think that is the issue at all. ICANN could probably
decide to interpret it that way if they really wished, but the policy as
written clearly has the onus on the registry to demonstrate that there
is not a public interest problem, rather than the onus on objectors to
claim that there is.  I read it as, if you want to apply and you want to
run it in a closed (registrar-less) way, you need to satisfy the three
requirements in the exemption clause in your application. It isn't
condemning them as a class any more than open registries are condemned
by having to agree to abide by the rest of the ROCC.
    I think that if ICANN decides it wants closed generics, then it
could set the bar very low in order to do that (and equally, if it
decides it doesn't, it can set the bar high), but waiting for an
external objection would not meet the policy as written.

> The notion that there was a case against these as a category of gTLD
is impossible as there were no categories.  Yes, Bertrand wanted them. 
He wanted them real bad.  But he did not get them.

    I followed that discussion too. I am not claiming anything came out
of it. I was just using it as an example of a discussion in which, if
there was consensus about closed generics, it would have been mentioned,
and it was not.

    I am also not claiming that the policy, as written, specifically
creates a category of closed generic, or anything like that. Rather, the
existing policy defines conditions for running a closed registry, that
include the maddeningly vague public interest test, and doesn't
distinguish between branded and generic closed registries as such (and I
guess the core of the substantial, rather than this historical sideline,
argument is whether there is a public interest difference between a
closed .brand and a closed generic).

> The only category* other than Standard was Community.  And yes,
categories have emerged since the new gTLD applications, but that was
the point: until they emerged with the actual application we did not
know they existed.
>
> So, as just another way one might use a generic TLD, private (or so
called closed) were envisioned as just another gTLD that someone could
apply for subject to the same rules as any other gTLD.  No exceptions
were made to the nature or use of a generic TLD other than those imposed
by law or community.

    I think position is only consistent with the existing policy if you
think the exemption clause, and particularly the public interest clause,
in the ROCC is literally meaningless. Because otherwise, clearly there
is a restriction on operating a closed, registrar-less, registry for the
registry operators own use - three restrictions, in fact, one of which
sets an addition public interest requirement.
    Regards
        David
>
> avri
>
> *  and that established by ICANN Staff not the policy process -
though we did differentiate treatment for community applications
>
> On 8 Mar 2013, at 02:45, David Cake wrote:
>
>>     The really baffling part of this discussion to me is the way in
which both sides are completely convinced that the issue was considered,
and resolved, years ago, while disagreeing directly on what the outcome
of that decision was.
>>     It is, clearly, impossible that both sides are right.
>>     So, I've done some digging.
>>     
>>     I looked at, for example, the Categories informal group that
Bertrand put together in Nairobi in 2010 and tje subsequent discussion,
because surely that would have made clear some consensus. It did not. In
that discussion, the idea of single registrant generic terms was
mentioned, but objected to quite strongly by some, while some spoke
loudly in their defence (including Evan Leibovitch, at the time). It was
made clear that single registrant gTLDs were possible, but if there was
any discussion of closed generics it was only to flag that some people
felt it was an issue. So I'm inclined to believe that this issue was not
settled in 2010 or earlier.
>>
>>     The only point at which the issue seemed to be directly
addressed within ICANN policy since then seems to be the VI WG. The VI
WG resulted in the Registry Operator Code of Conduct (henceforth ROCC)
that was incorporated into the Applicant Guidebook. The ROCC
>>
>>     The relevant portion would appear to be:
>>>     • Registry Operator may request an exemption to this Code of
Conduct, and such exemption may be granted by ICANN in ICANN’s
reasonable discretion, if Registry Operator demonstrates to ICANN’s
reasonable satisfaction that (i) all domain name registrations in the
TLD are registered to, and maintained by, Registry Operator for its own
exclusive use, (ii) Registry Operator does not sell, distribute or
transfer control or use of any registrations in the TLD to any third
party that is not an Affiliate of Registry Operator, and (iii)
application of this Code of Conduct to the TLD is not necessary to
protect the public interest.
>>>
>>     So, it seems pretty clear that:
>> - ICANN has considered the issue of closed TLDs.
>> - ICANN has NOT specifically considered the issue of closed
generics, except as a particular case of closed TLDs. Closed TLDs are
explicitly permitted, but are conditional.
>> - by current ICANN policy, a closed generic then is explicitly
permitted, IF and only if it can demonstrate that application of this
Code of Conduct is not necessary to protect the public interest.
>>
>>     Note that Registry operators registering names without going via
an accredited registrar (except a very small set) is not permitted via
the ROCC. If they want to actually a top level domain as a closed domain
(generic or not), the current ICANN policy is that they should have to
ask ICANN for an exemption on the grounds that the ROCC is not necessary
to protect the public interest.
>>     
>>     As far as I can tell, that is the current policy - that closed
generics are possible, but are not, by default, permitted. So, it
appears that neither side is 100% right, and neither 100% wrong. ICANN
still has the right to disallow any closed generic on public interest
grounds as per the Applicant Guidebook.
>>
>>     So, as far as I can tell, the actual current ICANN policy is
that closed generics are possible, but ICANN can choose to allow them
only if they are convinced there is a public interest reason to do so.
>>
>>     Milton is, of course, correct that the 'public interest' is a
problematic term. If you are to go on current policy, then the question
of closed generics depends mostly on how 'protect the public interest'
is defined, and of course both sides of debate could probably find a way
to define it to suit their argument.
>>     But I think it is time to stop saying that 'closed generics were
decided on in 2008' or whatever. Current ICANN policy actually seems far
from clear, and there does not ever appear to have been community
consensus on the issue.
>>
>>     Thoughts? Responses?
>>
>>     Cheers
>>         David
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.ncuc.org/pipermail/ncuc-discuss/attachments/20130308/6e991bd1/attachment-0001.html>