Re: [soac-newgtldapsup-wg] Financial instrument

[Eric Brunner-Williams <ebw@xxxxxxxxxxxxxxxxxxxx>]

On 10/1/10 3:27 PM, Avri Doria wrote:
> Hi,
>
> thanks.
>
> This would be really good material for one of the FAQ entries.
>
> E.g. Why should the security period be lowered from 3 years

It is not a "security period", it is a "continuity period", and the 
purpose that requires 36 months is not clearly identified. Richard 
made a guess at it -- time required for a registrant to re-brand in 
some other name space. I was unconvinced since there is no possible 
alternative to re-brand from one community's associated name space to 
another community's associated name space.

> I think it might make the recommendation more complicated and make the 
> recommendation would get lost in the discussion.

We all have our thoughts. If you prefer to offer a simple alternative 
to the continuity instrument as-is, fine. What I wrote was to recap 
the mail and vacation-interrupted conversation I had with Karla, and 
to explain that "continuity" is quite different for applicants 
in-isolation than for applicants in-cooperation.
One of our fundamental issues is whether our framework assumes (and 
promotes) needs-qualified applicants acting in isolation from each 
other or assumes (and promotes) needs-qualified applicants acting in 
cooperation with each other.
CORE, Afilias, CoCCA, are all instances of cooperation in the registry 
and registrar markets, as are ARIN, RIPE, APNIC, and LACNIC, and at 
some level of abstraction, ICANN itself, as cooperation among the SOs.
If there is consensus to offer a continuity instrument related 
proposal, then all the rational drops away and one of us, you or me or 
someone else, simply writes the declarative prose proposal.
Eric


> a.
>
> On 1 Oct 2010, at 15:02, Eric Brunner-Williams wrote:
>
> > Tijani, Avri,
> >
> > This (Tijani's variation on Richard's suggestion) lowers the months at some 
> > cost per month from 36 to 6.
> >
> > What this doesn't do is provide a rational basis for the cost per month.
> >
> > There is a fixed costs, the monthly reports to ICANN being one.
> >
> > There are the variable costs, the minimum power, cooling, connectivity, computational 
> > capacity, office space and staffing, for operations "during continuity".
> >
> > My first claim (really a CORE operational observation) is that both the fixed 
> > and variable costs are so small that an operator providing back-end services to 
> > a very few registries would write off the overhead of both costs.
> >
> > So, where a few applicants share facilities (line 253 Avri), the proper amount 
> > the applicants should deposit as a continuity instrument is zero.
> >
> > My second claim is that where two or more registries being "continuity", there is only 
> > nominal cost for each additional registry being "in continuity". The incremental cost of 
> > doing the N+1 monthly report to ICANN is nominal. The incremental cost of operations is nominal.
> >
> > So, as long as some of the applicants who form shared facilities (again, line 253) are not "in 
> > continuity", there is no additional cost from sharing applicants going "into 
> > continuity".
> >
> > Now suppose the Board rejects this too, and the applicants must all set aside 
> > some amount representing N months (possibly 36) at a burn rate the Board finds 
> > credible.
> >
> > What happens when an registry commences "continuity" operations?
> >
> > Does the current operator draw from the fund or does the fund go to some other 
> > operator, who will then draw from the fund, until the fund is exhausted, or 
> > possibly until the operations are profitable, and the operator (original or 
> > subsequent) replenishes the continuity fund?
> >
> > Is this a "contingency" capability, or is it a transfer of resources from an applicant meeting the 
> > "needs" criteria to an operator, perhaps one of VGRS, AF, NS, CORE, ... that does not meet the 
> > "needs" criteria, which is triggered at any time ICANN, or a market dominated by VGRS, AF, NS, 
> > CORE, ..., causes small registries early in their operational history to have revenues that don't meet 
> > expenses for a quarter?
> >
> > Eric