Re: [alac] new gTLDs

[Wendy Seltzer <wendy@xxxxxxxxxxx>]

I was similarly impressed by Ross's proposal.  It seems as neat a
separation of "technical" and "policy" layers as we're likely to see
in this institution. Further, if we formally separate the layers, and
create a contractual interface between them, we diminish the risk of
failure -- if the (technical) registry fails, its functions can be
transferred elsewhere and the gTLD keep operating. 

It's my feeling that a gTLD string should be approved so long as it
meets minimal no-harm criteria on technical and policy grounds.  Both
questions should be merely threshold ones: is the proposed addition
likely to harm the Internet?  But the evaluations of "harm" on each
axis differ: failure to resolve names in a timely fashion or store
data accurately on the one hand, or creation of confusion among
Internet users on the other.  Separating the evaluations, and giving
minimal basic criteria for each would help streamline the process of
new gTLD addition.

(Perhaps it also forces ICANN to confront more squarely the subjective
nature of policy choices among gTLD policies, and leads them in the
direction of approving more strings rather than trying to refine their
"criteria", but I'm not holding my breath.)

--Wendy

On Wed, Apr 02, 2003 at 03:45:01PM +0200, Thomas Roessler wrote:
> On 2003-03-31 22:26:27 +0200, Vittorio Bertola wrote:
>
> > - Applicants should also be allowed to present more than one
> > technical option to establish the registry - for example, a plan
> > to develop internal operations *and* tentative agreements with
> > existing registry operators as a fallback in case the (more
> > complex) internal plan fails. Otherwise everyone will go for the
> > existing registries just to be sure not to lose points on that
> > evaluation item. The odds of practical implementation are not
> > easily foreseeable, so the evaluation should be focused on the
> > technical and managerial know-how of the applicant, rather than
> > on its current plan, which may be outdated anyway by the time the
> > TLD is awarded. You want to be sure that these people can do it,
> > not to force them to state in advance up to the last detail how
> > they will do it.
>
> I think that Ross Rader's approach (as presented in Rio by Bret
> Fausett) is really the best thing to do about this: Introduce an
> accreditation regime for existing and prospective (!) registries,
> which checks minimal technical thresholds. Once a TLD proposal comes
> with an accredited registry operator (be it internal or external),
> that proposal should be considered fine on the technical side.
>
> With this approach, ICANN could finally create a competitive market
> on which registry operators can act -- they'd have to compete in
> terms of technical quality, price, services etc. in order to attract
> gTLD delegees.
>
> --
> Thomas Roessler                 <roessler-mobile@xxxxxxxxxxxxxxxxxx>

--
Wendy Seltzer -- wendy@xxxxxxxxxxx || wendy@xxxxxxx
phone: 415.436.9333 x125 / cell: 914.374.0613 / fax 415.436.9993
Staff Attorney, Electronic Frontier Foundation
Fellow, Berkman Center for Internet & Society at Harvard Law School
http://cyber.law.harvard.edu/seltzer.html