Re: [soac-newgtldapsup-wg] registry back up/ financial instrument

[Eric Brunner-Williams <ebw@xxxxxxxxxxxxxxxxxxxx>]

I'm sorry I missed this part of the call, as it is a set of issues 
that I've been interested in for some time, dating back to the initial 
registrar failover problem, and the subsequent registry failover problem.
Whatever ICANN's rational for having a continuity interest, I think 
our interest is in examining the cost, necessity and utility to the 
applicants qualified for support.
For Community-based applications, rebranding seems unlikely as a 
registrant response to registry failure. The same observation applies 
to registrants of an NGO's failed registry, and registrants in an 
emerging market/developing country having only one registry, failed in 
our test for utility, and for registrants of a limited presence 
language registry. It isn't even clear what the registrants of a 
failed entrepreneur-originated registry should do, given their initial 
registration preference.
For these reasons I can't accept registrant rebranding as the rational 
for a continuity interest.
Additionally, the "protection of registrants" and related statements 
of rational in 
http://www.icann.org/en/topics/new-gtlds/registry-transition-processes-28may10-en.pdf 
appear to miss the possibility of uniqueness of the choice of registry 
by registrants, through a focus on termination of one standard type 
registry and transition to a second standard type registry, rather 
than upon the sustainability issues a Community-based registry may 
face intermittently, or chronically.
Now let me offer another <brave_hat="on"> statement.

In both success and failure our interest in promoting improvement to 
the representation goal contained in the Green Paper and maintained 
since is guiding. We don't want to end up with Verisign as the general 
beneficiary of new gTLD registry failure, being the default failover 
operator. This non-goal applies to NeuStar, Afilias, and CORE as well.
If an applicant can identify one or more existing operators, such as a 
ccTLD registry operator, and/or a non-operator with sufficient 
reserves and competencies, such as a database operator, needing only 
the EPP and business logic wrapper around the database operations, 
which are willing to provide continuity services, then the fundamental 
continuity interest is met.
In so far as registry operations necessitated by failures of new gTLD 
registry can be kept out of Reston, Toronto and Dortmund, some aspects 
of the representative interest is preserved, even in registry failure.
<brave_hat="off">

Now existing registry operators, even registrars, but not general 
purpose database operators and applicants, should have a feel for the 
incremental cost of importing a failed registry's data and any 
business logic necessary for continuity purposes. At a minimum, the 
cost is bounded below by the mere staffing cost to generate monthly 
reports about unchanged data to ICANN. Going back to the Cairo meeting 
with Doug Brent, outgoing COO, ICANN's generic staffing cost for 
registry oversight, independent of success or failure, was set at 
$25,000/yr. The labor cost of reading reports in Los Angeles is likely 
to be significantly greater than the labor cost of generating reports 
in emerging markets/developing countries.
Using the PuntCat published data, we have numbers to inform our 
estimate of the cost of continuity operations for some of the 
applicants who qualify for support. 40K registrants with 1 year leases 
yeilds 3.3k lease renewal events per (20 day) month, 166.6 per (8 
hour) day, or 21 full database dips per hour.
There really is minimal operating cost, at least for some of the 
applicants who qualify for support.
So what _should_ satisfy the continuity instrument for qualified 
applicants?
I suggest the following ordered by preference list:

o A credible offer from a cooperative formed by two or more 
independent applicants,
o A credible offer from a ccTLD operator,

o A credible offer from a gTLD operator,

o A credible offer from a entity capable of becoming a registry upon 
demand,
o An irrevocable standby letter of credit, and

o An irrevocable cash escrow account.

Under no circumstances should the standby letters of credit or cash 
escrow accounts of applicants from emerging market/developing country 
be given to continuity operators in North America or Europe.
Eric