Re: [gnso-vi-feb10] VI - An RSP Question..

[Eric Brunner-Williams <ebw@xxxxxxxxxxxxxxxxxxxx>]

Statton,

There are three distinct sets of interests.

1. CORE the IAHC-MOU formed entity, with the interest in the increase
of the number of TLDs which are informed by RFC 1591, section 3,
subsections 2 and 3 -- the "duty" language. May and June 1997, and
continuous.

2. CORE the ICANN accredited registrar, with the interest of selling
COM, NET and ORG, prior to 2001, extended by most of the subsequent
new gTLD startups, with the exceptions of .aero, .cat, and .museum.
April 21, 1999, and continuous.

3. CORE the registry services provider, initially for .aero, later for
.museum and .cat. December, 2000, and continuous.

For simplicity, these are CORE-RO, CORE-RR, and CORE-RSP, and as my
first note "Agenda item offer" pointed out, the November 1998 IETF
draft by Kent Crispin, Dave Crocker, Roberto Gaetano and Sylvan
Langenbach for a Shared Registry System Protocol (SRSP) envisioned
that even within a single name space, multiple RSPs could exist,
distinct from the RO relationship with the IANA, modernly a contract
with ICANN. The idea of a "competitive registrar" is actually predated
by the idea of a "competitive registry services provider", for any
given registry operator.

You ask: As a back-end registry provider, why would CORE advocate for
greater restriction on CORE's ability to contract with another party,
such as a registry operator or registrar?

A. CORE's RR interests are adequately addressed by the existing gTLD
registries, minus .museum and .cat, and some ccTLD registries. We
expect that CORE will not be the only source of applications in the
new gTLD round, there may be one or two others. The "registrar"
interest does not need access to every new inventory to make
reasonable ROI for CORE's members.

Here we've 60 or so competitors.

B. CORE's RSP interests are in the growth of its partner or client
RO's registry, which does not exist for the benefit of CORE as a
registrar. If CORE placed its registrar interest ahead of its partner
or client's own interests as an RO, the RO would reasonably seek an
alternate RSP vendor. This wasn't the case with SITA's change of RSP
vendor, but it general, the rule is make the client or partner happy,
don't try and own the client or partner.

Here we've 3 competitors: the RSP that operates com/net/name, the RSP
that operates org/info and the RSP that operates biz.

C. CORE's RO intrests are probably quite similar to every other
would-be RO. Lots of happy registrars is a lot more important than
CORE as a registrar, see also A above.

We think the restriction results in a more predictable set of
relations and expectations in a market that is necessarily
multi-actor, international, and in transition from monopoly to
competition.

Thank you for an interesting question.

Eric