[comments-dotorg] Follow-up to comments on Final report of .org Task Force

[Eric Brunner-Williams in Portland Maine <brunner@nic-naa.net>]

There appears to be some underlying points of agreement between Mssrs Meuller
and Neuman, to pick them as arbitrary points of reference in an apparently
bi-polar dispute. Both agree that:

	o the designee SHALL BE a non-profit entity,
	o the designee SHALL NOT operate the registry,
	o not to reference cost data available from the newTLD roll-outs.

As a policy, steering the current registry and its communities of registrants
towards the original purpose of the registry, and the overwhelming majority
of its 1 million registrants, towards non-commercial purpose, is reasonably,
but not of necessity, applied also to the selection criteria of the designee.
As several DNSO constituencies appear to agree on the first point, that issue
is settled.

The second point is probably one which is not controlling. A proposal by a
non-profit is not likely to fail simply because the non-profit is demonstrably
capable of operating the registry, and proposes to do so rather than contract
the service out to another non-profit, or for-profit, either by single-source
or competitive bid. Consequently, that issue is peripheral.

The third point is where the money is. One-time and recurring cost vs pricing.
This is where a discussion would benefit from data. With only two exceptions
known to me, all of the newTLD projects ran 7-figure one-time costs, ranging
from a high of $20 million (.biz) to one-tenth that. Assuming that the terms
of the divestiture are "data only", the application must cost-in an RRP or an
EPP registry, one or more data bases and associated business logic (some of
which isn't necessary until some time after "go-live"), and a constellation
of public data publication facilities hosting dns, whois, and operational
"other" servers.

The community experience and expertise developed in the newTLD roll-out of
late-2001 and early-2002 (.info, .biz, .name, .coop, .museum, and .pro and
.aero) strongly supports costing a "data only" transfer at or under the mid
6-figure mark for one-time cost, with North American/European labor costs
assumed.

The operational experience of SRI during its tenure of the ARPA NIC contract,
with a staffing of 15 EFTS, modified to reflect a CRM model restricted to the
ICANN accredited registrars, a policy-body liaison task, and a infrastructure
maintencne requirement, strongly supports costing either a "data only" or a
"data-plus-infrastructure" transfer at the 7-figure mark for recurring costs,
again with North American/European labor costs assumed.

These figures, a .5M one-time and a 1M recurring cost, are guides for neutral
policy base price, and active policy add-on price estimates. How active the
policy or policies eventually specified in an ICANN RFP, or how creative the
cost management and value multiplier in an Application, these vary from this
base point.

Eric Brunner-Williams
wampumpeag, llc