RE: [gnso-vi-feb10] Proposed Addendum to Proposals

[Jeff Eckhaus <eckhaus@xxxxxxxxxxxxxxx>]

Thanks to everyone for the input and response to my proposed addendum. As I 
mentioned at the close of my email this is meant to be a framework and open to 
new ideas and constructive criticism that will hopefully start a path towards a 
resolution. It is not a closed proposal that is take it or leave it and of 
course will need input from the group.
I would like to address some of the specific questions and items brought up, in 
no particular order:


*         The proposal is an addendum to the JN2, which would allow a 
registry-registrar bundle to sell its own TLD after 18 months and an 
application. There is no mention of never being able to sell names in its own 
TLD

*         Yes, audit is the same thing in Beijing, Brussels and New York. Thank 
you for asking that question and clearing that issue up.

*         I personally am OK if VRSN decided to become a Registrar for .shoe or 
another new gTLD, but that is something the group can decide

*         The reason I set the framework with Registry/Registrar/VI group 
because this group has stated ICANN is not up to the task and existing 
Registries like PIR and Afilias are the ones that have brought up issues of 
Registry data so they would be in the best position to set up the rules to 
guard against it.

*         This proposal did not take into account exceptions and I leave it up 
to the group to decide if there needs to be any (ex: community, brand)

*         The costs of these audits may indeed outweigh the benefits but what 
are the options with moving forward on cross-ownership when some in this group 
are convinced there will be enhanced harms from any type of cross-ownership?

*         I did not propose a complete separation of registry-registrar. That 
is a huge cost for any new entrant and effectively serves a barrier to entry 
which may be the goal of some that support structural separation.

I have stated from the beginning that my personal belief is that the harms that 
some are claiming will occur from the type of cross-ownership are non-existent 
and that it is more an approach to keep out a class of applicants than anything 
else. Harms will happen with the 15% proposal as all the harms that have been 
brought up occurred with the 15% ownership limit in place. If people were most 
concerned about harms they would be pushing for the DAGv4 version which is 2%.
But.........we need to come to a consensus and I think this a path forward.  
Having a spirited discussion on whether or not harms will occur while 
intellectually stimulating will not get us to a conclusion and the protections 
that some require to move ahead with cross-ownership.


Regards,


Jeff Eckhaus



From: Antony Van Couvering [mailto:avc@xxxxxxxxxxxxxxxxxxxx]
Sent: Tuesday, June 15, 2010 8:48 AM
To: Jeff Eckhaus
Cc: 'Gnso-vi-feb10@xxxxxxxxx'
Subject: Re: [gnso-vi-feb10] Proposed Addendum to Proposals

Jeff,

Thanks for the constructive attempt.  I think it's a useful way into the 
problem I see with many proposals.  Unfortunately, the imposition of 
significant costs and regulatory burden would be injurious if not fatal to 
smaller registries.  I've tried to add more color to that assertion below.   If 
this issue were addressed properly, I would be much readier to sign on to a 
compromise proposal.

I note that one of your thoughts is that a registry-registrar bundle would 
never be able to sell names in its own TLD.  I do not believe that "exceptions" 
for orphaned TLDs make any sense, especially if they are combined with a cap 
that would essentially discourage the registry from succeeding and would entail 
a very expensive migration/transition process, not to mention customer 
confusion) when the cap was reached.  The people from .coop graphically 
described the pain they went through in this case, in one of our earlier VI 
meetings, I forget which one.   Furthermore, there is the horrible case where 
the registry is not "orphaned" but might as well be -- only one or two subpar 
registries are carrying the TLD, not marketing it, not answering their phone, 
etc. -- and the registry impotent to do anything about it.

Suppose however that the "same TLD" restriction were not in place, and that a 
small registry could act as a registrar for its own TLD under your proposed 
rules. For a small registry, the restrictions would seems so bizarre, punitive, 
and anti-competitive that the "Eckhaus Rule" would shortly be as well received 
as the imposition of shariah in Cancun.

Consider what I believe will be a common case: a small registry with about 6 - 
10 employees, registry/registrar tech functions outsourced.    One person in 
charge, two sales and marketing people, an accounting/financial person, 
possibly a technical person to manage the registry tech functions, and someone 
to deal with ICANN, who may also be a lawyer.  Even if someone were to argue 
that a registry needs twice as many people -- evidence among the ccTLDs 
notwithstanding -- the case is the same.

This small registry may be a cultural/linguistic "community" applicant, or it 
may simply be a small registry that for some reason doesn't reach the high 
"community" bar, for example .indigi, which will fail as a "community" because 
it is not a community, but a community of communities (don't blame me, I didn't 
write the rules).   This registry may also be a small entrepreneurial venture, 
such as (among those announced) .cal, or .board.  This registry may also be a 
government-supported geographical TLD, such as the registry for a small city.   
What these registries have in common is that they need to or want to be able to 
market and sell directly to the public, through their wholly-owned 
ICANN-accredited registrar.

The proposed audits will cost between $50,000 and $100,000.  There are certain 
auditing costs that don't depend on the size of the enterprise.   At a gross 
profit of (say) $10 per name, that's an extra 5,000 to 10,000 names that need 
to be sold to be pay for those audits.  To put this in perspective, this range 
of sales is approximately the entire annual sales of .cat.

Now consider the issue of separating the registry and registrar -- this is an 
even bigger cost.   Basically you would have to double your staff, and possibly 
-- in order to maintain real separation -- pay a separate rent as well.  I 
won't attempt to estimate the cost, but it's large and very possibly would make 
operation impossible.

All of this in order to prevent hypothetical harms to existing 
registrars/registries.  Your suggestion may solve one problem, dear to members 
of this Working Group, but it would create another much larger one.

Antony





On Jun 14, 2010, at 4:54 PM, Jeff Eckhaus wrote:


All,
After today's discussions with this group and reading the emails on the list, I 
have noticed a consistent concern coming from many group members, that they are 
worried about ICANN's ability to enforce any rules that are put in place. It is 
one of the main concerns opponents of the JN2 proposal have expressed with the 
issue of co-ownership. Specifically being able to police the following issue: 
"Registry Operator or its Affiliate may serve as an ICANN-Accredited Registrar 
in any top-level domain other than the TLD for which Registry Operator or its 
Affiliate serves as the Registry Operator" .
Those opposed to JN2 and other proposals seem to agree that cross-ownership is 
appropriate, but that ICANN will not be able to police any restrictions on data 
sharing between a registry and registrar. They believe that we cannot simply 
rely on Registrars to adhere to a signed agreement. Thus, because compliance 
will be too difficult to enforce, we must limit cross-ownership.
While I disagree with this viewpoint, my opinion does not matter at this point. 
What does matter is assuring the people who are concerned with the above, 
attempting to bridge the gap and reach consensus. To that end, I am proposing 
an unsolicited addendum to the JN2 proposal (maybe JN3 now??) and to any other 
proposals that allow a Registry to own up to 100% of a Registrar (vice-versa) 
but not distribute the owned TLD.
This will only apply to co-owned entities that have an ownership level above 
the 2% threshold as discussed in the DAGv4:
*         The Registry/Registrar must agree to an annual audit for the first 2 
years. Every 18 months for next 3 years, and every 2 years thereafter. (timing 
can be negotiated)
*         The audit will focus on ensuring that Registry data is not shared 
with the co-owned Registrar, co-owned Resellers, and any related Affiliates
o   Details of what data would need to be audited would be supplied by a 
working group/committee led by current Registries
*         Stiff penalties would be levied if there is an audit failure (amounts 
TBD)
*         All costs of the audit would be borne by the co-owned entity
o   The co-owned entity would pay fees into a pool, not directly to auditors. 
This avoids any thoughts of impropriety
*         This audit would be in addition to the audit and compliance 
requirements already agreed to in ICANN Registry and Registrar agreements
*         Auditors would be independent of ICANN but would work with ICANN 
Compliance on fees and remedies
*         Auditors would be rotated among assignments to avoid capture
This is the framework of my proposal that I believe would cover Registrant 
rights and concerns and bring comfort to those who believe there will be 
enhanced harms if there is any type of co-ownership. Most important it would 
cover the policing/enforcement issues that seems to be the roadblock to 
consensus. I welcome feedback on this idea and look forward to hearing more and 
seeing everyone in Brussels.

Regards,

Jeff Eckhaus