[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [wg-c] Eureka?
I think Kevin's post is extremely helpful. There's been a lot of ink
spilled attacking the idea that, on Day One of the new regime, we should
immediately add 100 new gTLDs to the root. It seems to me, though, that
that's a strawman: I'm not aware of anybody in the WG who's advocating
that. So far as I know, everyone here who believes in the addition of lots
of new gTLDs is in favor of some sort of phased rollout. Such a rollout
might, for example, look like the one contemplated in
draft-postel-iana-itld-admin-00.txt: ten new registries and 30 new gTLDs
per year, for a period of five years.[*]
The debate within the WG, near as I can tell, is between three positions:
[1] We should agree now on a phased rollout under which a lot of new TLDs
end up being authorized over the next few years.
[2] Having a lot of new TLDs sounds good in theory, but our action should
be to add a few new gTLDs now, and later on we can revisit the possibility
of adding additional ones (or not).
[3] We should add a few (or no) new gTLDs and stop, because having lots of
new gTLDs isn't a good idea.
I think [2] won't work -- if we try it, odds are we'll end up adding a few
new gTLDs and no more. If we add three or five or seven new TLDs, and then
replay the entire argument from scratch, the weight of inertia will be
heavily on the side of those who argue against expanding the name space.
Current TM interests will still oppose future TLDs; the new TLD operators
added by ICANN won't be thrilled about further expansion either. So if we
are *in fact* to add a lot of TLDs in the long term, I think option [1] is
the only way to do it.
Kent has argued that it's important to structure the name space; I think
that's a separate issue. FWIW, though, the interest in structuring seems
to me to cut in *favor* of an extensive phased rollout. If we wanted in
the long term to have a lot of gTLDs in a structured name space, it would
make sense to plan out that space *now*, and then get those gTLDs out there
in a phased rollout. It would make less sense to plop in a few new gTLDs
now, and only work out the structure later on, after deciding whether there
are in fact going to be any more gTLDs.
I've explained before why I think adding a lot of new gTLDs is a good
idea. Some folks have argued that having many new gTLDs would be bad b/c
the proliferation of gTLDs will greatly increase trademark policing costs.
It seems to me essentially inevitable, though, that ICANN will adopt
policies ensuring that all new registries have uniform search protocols.
If ICANN does so, then the impact of new gTLDs on trademark policing costs
shouldn't be great. (Nothing I'm saying in this post, BTW, goes to whether
we should have a large number or a small number of *registries*
administering the TLDs. That's also a separate issue.)
If we reject [2] and [3], we're back to [1]. How quickly might the phased
rollout go? That depends, I guess, on the risks that it poses. I've seen
two arguments that a too-rapid rollout would be risky. One is that
establishing new registries too quickly will make it difficult for ICANN to
ensure that all of the new registries have appropriate dispute resolution
policies and protocols for finding contact information. The second is that
rollout should be slow in order to allow gradual testing of the new dispute
resolution process.
I don't see the first argument as substantial. Unless ICANN flops
entirely (in which case our entire discussion is moot), I expect it to have
the power to impose on the registries, via contract, any dispute-resolution
and contact-information policies it finds desirable, and I don't expect to
see the registries to flout the contractual terms they've just agreed to.
I do think the second argument raises a legitimate concern. But I don't
think our desire to avoid overwhelming the dispute resolution process means
we need to dribble out domains at the rate of only a few a year, say; I'm
not sure why something close to a Postel-style rollout won't do it.
Thoughts?
Jon
---------------------------------------------
[*] Postel raised the possibility of adding a larger number during the
first year.
Jon Weinberg
weinberg@msen.com
At 08:43 AM 8/4/99 -0400, Kevin J. Connolly wrote:
>I accuse myself of the sin of failing to parse the issues
>and thereby engaging in fuzzy thinking. Gomen nasai :-)
>
>There is a difference between the end stage steady
>state of the DNS and the trip through the desert to
>get there.
>
>I don't think there are any responsible people who deny
>that in the end stage steady state, there should be LOTS
>of gTLDs. As many gTLDs as there are viable registries
>for them. If technology still imposes an upper limit on the
>number of TLDs in the root, then we'll have to devise a
>mechanism for allocating the domains. It will also be
>important to develop a method for policing the root to cut
>out the dead domains, as well as some method for forcible
>transfer of custody of the registry function for a TLD if (for
>example) the registry operator goes bankrupt and
>abandons LARGENUM of paying customers. (This last
>criterion will improve user confidence in the stability of the
>DNS.) In that steady state, there will be no dictation by
>ICANN of the business model and policies of the registry
>beyond those dictated by the trademark protection and
>dispute resolution policies.
>
>The reasons for this conclusion are legion and don't need
>to be rehashed here. That can be done by a drafting subcommittee.
>
>The trick is how to get from where we are now to where
>we wish to be, without getting derailed in the process. This means
>that the process has to start under a yellow flag. We will also
>need to estalish the mechanism for red- and/or green-flagging
>the addition of additional TLDs.
>
>(A) The process has to be open, transparent, and responsive
>to the wishes of the entire Internet Community.
>This is OUR responsibility.
>
>(B) We will need to decide whether this WG will propose a
>list of gTLDs, a mechanism for introducing gTLDs, or a
>combination of both. I think we should do both: propose a
>list and a schedule sufficient to cover the first calendar
>year (+/-) of operation and a method (other than this blessed
>workgroup fussfest method) for introducing new domains.
>
>(C) The introduction of new gTLDs will need to be coordinated
>with other policy decisions being taken by ICANN, particularly
>with respect to the protection of famous trademarks and the
>development of fair and economical techniques of dispute resolution.
>
>(D) The transition process needs to moderate the impact of creating
>new monopolies so as to prevent those entities which are favored
>during the rollout phase with a period during which monopoly profits
>will accrue to registry operators.
>
>(E) The transition policy needs to recognize the special status of
>Network Solutions, in order (among other things) to prevent NSI
>and its affiliates from engaging in predatory practices.
>
>Okay, I think that's enough to convey my thinking and enable
>progress by picking my ideas into sawdust.
>
>KJC
>
>:include <yada yada>
>
>**********************************************************************
>The information contained in this electronic message is confidential
>and is or may be protected by the attorney-client privilege, the work
>product doctrine, joint defense privileges, trade secret protections,
>and/or other applicable protections from disclosure. If the reader of
>this message is not the intended recipient, you are hereby notified
>that any use, dissemination, distribution or reproduction of this com-
>munication is strictly prohibited. If you have received this communi-
>cation in error, please immediately notify us by calling our Help Desk
>at 212-541-2000 ext.3314, or by e-mail to helpdesk@rspab.com
>**********************************************************************
>
>
>