Comments on "New gTLD Applicant Guidebook Version 3"

[Patrick Mevzek <contact@xxxxxxxxxxxx>]

Following the ICANN announcement at
http://www.icann.org/en/topics/new-gtlds/comments-3-en.htm
please find below my comments.


I've specifically commented on module 5 and my comments are archived
at 
http://forum.icann.org/lists/3gtld-transition/msg00005.html

I will here only comment on the generic procedure for new gTLDs.

The previous round of new gTLDs was presented as a testbed. This
would have implied there are some tests to be made and some
conclusions to be reached. As none of this was planned before
starting new TLDs at the end of the day the new TLDs do not test
anything, they are just new services for the public, with success or
not depending on who you ask and on what metrics do you count.

I'm not overly enthousiast on all new gTLDs as currently announced
in various media, but as the same time I believe that competition
should be possible, and that people should be able to innovate and
test new services. I do hope that some new innovative ideas are
currently lurking in the dark and that they will later on blossom to
fruitful new gTLDs as soon as the process is ready.
That would make me think that lower barriers of entry are better, as
well as reduced timeframes, but without hurry either because after a
10 years delay, 6 months more is not a big deal, and the Internet
will continue to function and people to register domain names in
current TLDs even if the new gTLDs are not launched tomorrow.

I doubt that all these new gTLDs will be able to compete against the
current ones that will still exist, but all new gTLDs will compete
against one another, that is for sure.

Among all cases, there are discussions on 'brandTLD', 'geoTLD' and
'communityTLD', where a 'standardTLD' should probably capture all
other cases. All these TLDs should be to the benefit of the whole
Internet community and be able to provide new features and values to
the day to day life of Internet users.

For that, it is often given as example that the new gTLDs will
provide easier access to content for users, like someone searching
for a pizza in new york will (should) immediately (?) find something
worthwile by going to the website at pizza.nyc
Now, I fail to see how things would be easier than today when,
besides .nyc, some other parties also decide to create .pizza,
.food , .restaurant or even .manhattan : now where should the content
be searched ? At pizza.nyc ? or nyc.pizza ? or nyc-pizza.food ? 
or pizza-nyc.manhattan ? etc...
Of course it is not for me to judge the merit of all of these, as I
said previously the competition should be open, and all Internet
users will decide at the end of the day what they prefer, hoping that
during that course true innovations may happen.

As for geographical needs of TLD, I would note that there was at
least one prior instance of that, namely .US (before its redelegation
to Neustar) with its 3rd and 4th level of domain names, to take into
accounts the US state and locality. The whole scheme has been scraped in
favor of a .US running now like any other TLDs, maybe because non-2nd
level domain names proved to be difficult to grasp to the public,
that may find a geography encoded directly at the top level to be
more attractive.
Again, the competition should be open, and the players free to test
what they want in that aspect.

And for the "communityTLD" case since it seems very hard to be able
to define what a "community" is in this specific context, and how it
should participate in choosing a registry and voicing its support or
its disagreement, so I fear that more cases like .XXX will happen in
the future, where there are many voices claming to be the community
served by it, but without all voices saying the same thing.
Maybe the whole notion of "communityTLD" should be scraped.

I would however even more be worried about new TLDs started where
basically no registration can take place (very small/closed community
and brandTLDs), or no use is made at all. This would be a kind of
"defensive registration" at the DNS apex, which would then become a
tree with dead leafs.
I would think that this case of TLD do not provide any benefit to the
Internet at whole, and hence measures should be taken so that this
case is limited as much as possible. This does not mean that future
registry operators should be required to have some minimum number of
domain names, just that the TLD is used in practice because the
creation of the TLD has been done so that registrations are expected
to happen in it, and not just to "reserve" the TLD and lock other
players out of it. And I'm not talking about TLDs where the volume of
domain names fall during the years or such other cases, but only on
TLDs created for the specific purpose of not being used at all.
Otherwise a lot of time and energy would have been spent by various
actors, besides obviously the registry operator itself, but also many
other Internet actors, without any clear benefit to anyone.

On a more procedural level, I do note that if the plan is, in the
future, to have gTLDs creation on a rolling basis while the first
round would be only during a limited timeframe, there are some points
in the current guidebook that will be problematic such as string
contentions issues and ways to resolve it, as using time and
artificially delay or hurry things would change the fate on some
issues.


More guarantees should also be given upfront about the timelines and
level of public data that will be available regarding applications received and 
their
reviews at various stages.

Of course, post round, a full assessment would need to be made on the
efficiency of the whole procedure as well as the costs involved and the
adequacies of fees asked by ICANN to applicants.
If these fees change in the future (for future rounds), specifically
if they are lowered or if some new cases are described to provide
reductions for some specific cases, some first round applicants may
feel to have paid more than necessary.

Finally, adding many new gTLDs (on the order of hundreds per year)
may have various consequences, some already being studied such as
higher strain (or not) on the DNS root server and such, but it will
have also operational consequences on dealing with registrars
accreditation/deaccreditation and listing on ICANN website,
registries requests on new services (RSTEP), IANA work and rate of
changes such as top level nameservers, amount of correspondence and
legal issues to take into account, etc.

-- 
Patrick Mevzek
Dot and Co <http://www.dotandco.com/> <http://www.dotandco.net/>
<http://www.dotandco.net/ressources/icann_registrars/prices>
<http://icann-registrars-life.dotandco.net/>