RE: Process forward [RE: [gnso-idng] restarting discussions on IDN gTLD]

["Edmon Chung" <edmon@xxxxxxxxxxxxx>]

In a different light, I think it is important that we also focus on the
registrants and even more importantly the general Internet user.  That is
what ICANN should concern itself about.  Providing a full IDN experience is
what the "market" needs and is what end users want.  This to me is a rather
clear and consistent message, especially from the recent discussions sparked
by ICANN's PR and also during IGF as well.

That being said, market advantage for registries should not be taken
lightly.  Our continued position was that IDN ccTLDs and IDN gTLDs should be
introduced at the same time.  That is also one perspective to look at the
issue.  In balancing the interests, that is sort of why I have suggested the
Track A/Track B approach, where Track B, even if implemented earlier would
not have created new registrant markets that do not already exist, thereby
reducing to a minimum the concern for market advantage.

In any case, the discussion here is probably not the specifics of what
process / tracks are being produced, but whether or not an implementation
team should be formed to look into it (although admittedly they are
inter-related).

In trying to summarize the discussion thus far, I wonder if we can say the
following:
- There is definite interest to explore a possible mechanism to introduce
IDN gTLDs as soon as possible
- There is likely going to be delays in the full new gTLD process (based on
the last known proposed schedule)
- There is continued interest to work towards minimizing the time lag
between the introduction of IDN ccTLDs and IDN gTLDs
- Implementation plans, if any, must address the issue of first-to-market
advantage
- Implementation plans, if any, must address the identified new gTLD issues
including overarching issues


Thoughts?...

Edmon



> -----Original Message-----
> From: owner-gnso-idng@xxxxxxxxx [mailto:owner-gnso-idng@xxxxxxxxx] On
> Behalf Of Stéphane Van Gelder
> Sent: Tuesday, November 24, 2009 3:30 AM
> To: Eric Brunner-Williams
> Cc: Edmon Chung; gnso-idng@xxxxxxxxx
> Subject: Re: Process forward [RE: [gnso-idng] restarting discussions on
IDN
> gTLD]
> 
> I'm not thinking so much of a possible comparison between a .paris and a
.ile-de-
> france, which would probably be able to run within the same track, but
rather
> between say a regional TLD and a company TLD (if single-owner), which the
latest
> CORE comments for example seem to suggest should not run at the same time
> (Werner's comments: single-registrant TLDs ... should not be allowed in
the
> coming round).
> 
> Stéphane
> 
> 
> Le 23 nov. 2009 à 19:07, Eric Brunner-Williams a écrit :
> 
> > Stéphane Van Gelder wrote:
> >> The market I am referring to is the registrant market. What I mean is
the first to
> be in a position to go through the ICANN application process. i.e., if you
are a
> prospective TLD applicant and your application window opens a year after
> someone else's, you have a to-market disadvantage.
> >
> > Thanks again for your patience. I'm still having some difficulty
understanding
> the importance of application submission, when the thing sold is domains,
which
> could take between 6 months and 24 months after the application window
closes,
> depending on the innate characteristics of the applications.
> >
> > "If your application window opens a year after someone else's, you have
a to-
> market disadvantage."
> >
> >
> > I don't see how a .paris in 1Q10 creates a to-market disadvantage to
anyone,
> except perhaps a .il-de-france in 2Q10 or later, should such an
application
> actually exist.
> >
> > I don't see how a .shoe in January creates a to-market disadvantage to
anyone,
> except perhaps a .sock in December, should such an application actually
exist.
> >
> > I'm sure there's a nugget of truth in your assertion, I simply think the
claim
> you're making could be narrowed to be generally true, rather than
generally false.
> >
> > Perhaps, any ".com-like .thingie" creates a to-market disadvantage to
any
> otherwise indistinguishable ".com-like .other-thingie", if the .com-like
.thingie
> precedes the .com-like .other-thingie.
> >
> > That seems a likely to be true statement, possibly even true where one
of the
> thingies, or both, are IDN, whereas the ordering of Paris or Barcelona and
> some .com-like .thingie-or-another, in ASCII or not, seems likely to be
false.
> >
> > I hope you can address the registrar interest in any restriction on the
inventories
> the RAA enables.
> >
> >
> > Eric
> >
> >> Stéphane
> >> Le 23 nov. 2009 à 16:12, Eric Brunner-Williams a écrit :
> >>> Stéphane Van Gelder wrote:
> >>>> Sorry Eric, I don't mean to suggest you shouldn't ask anything else.
Please
> do so!
> >>> OK.
> >>>
> >>> 1. Are you using "first to market advantage" meaning
> >>> (a) first to offer inventory through RAA sales channels, or
> >>> (b) first to offer something else?
> >>>
> >>> If (b), could you explain what the market is, what is bought and sold
in it, that
> kind of thing, and what the advantage is in that market?
> >>>
> >>>
> >>>
> >>>
> >>>> I was simply frustrated at my own apparent lack of clarity and my
inability
> to get my message across.
> >>>> Your latest question seems to highlight this fact, as I am in fact
very much
> in favour of the GNSO's position that gs should be released at about the
same
> time as IDN ccs (I don't recall the RySG's position on the matter, but as
this is not
> my SG anyway, I won't comment on it). I am sorry that my previous message
led
> you to understand exactly the opposite. Please note however that the way
you
> seem to be characterizing the GNSO's position on this seems wrong. The
GNSO's
> position is no limited to IDN gs.
> >>>
> >>> I only have an observer's knowledge of the Council's position, but the
act of
> asserting the desirability of a similar period of availability for ccTLD
IDNs, the
> ccTLD IDN FT excluded, and gTLD IDNs, without explicit reference to also
the
> same period of availability for gTLD ASCII, creates the possibility of
sequential,
> not simultaneous ordering of events.
> >>>
> >>> It is hypothetical, but if the Council had two alternatives, one to
allow gTLD
> IDNs "today" and gTLD ASCIIs "tomorrow", and the other to allow gTLD IDNs
> "tomorrow" and gTLD ASCIIs "tomorrow", it seems possible that they might
settle
> on the first choice, as "tomorrow" might be several years from "today".
> >>>
> >>> If your recollection of the Council's position on this issue is
correct, then
> other than technical nits, such as table consistency, there is no reason
for this list
> to exist, as there could be no policy  with Council support that would
differentiate
> IDNs offered under a registry contract and ASCIIs offered under the same
> contract.
> >>>
> >>>
> >>>> Once again, Eric, please do not hesitate to come back with further
> questions. I will endeavor to answer as quickly and as clearly as I can,
and
> certainly do not want to stifle discussions.
> >>>
> >>> Well, I do have one more, touching on SG positions. As a registrar
(USA
> Webhost and CORE), I'd like to offer more RAA restricted inventory
_today_.
> >>>
> >>> How is it in the interests of the members of the RSG to advocate that
there
> be no more RAA restricted inventory _today_?
> >>>
> >>> I don't mind selling Chuck's or Jeff's or Hal's existing product using
my
> RAAs, I'd just like to know why it is in my advantage as a registrar not
to be able
> to sell more of their, or anyone else's product, using my RAAs.
> >>>
> >>> Eric
> >>>
> >>>> Thanks,
> >>>> Stéphane
> >>>> Le 23 nov. 2009 à 15:18, Eric Brunner-Williams a écrit :
> >>>>> Stephane,
> >>>>>
> >>>>>
> >>>>> Thank you but you've managed not to suggest what the advantage is
that
> you seek to avoid, and what the market is in which this advantage might
exist.
> >>>>>
> >>>>>
> >>>>> Would it be reasonable to conclude from your comment that you are
also
> opposed to the position of record of the GNSO, or at least the RySG, that,
> except for the ccTLD IDN FT, that IDN offering by the ccTLDs and the gTLDs
> occur roughly at the same time, to minimize the "first to market
advantage"
> (obviously, not in the same sense as you use the term), because that
position
> distinguishes IDN gTLDs from ASCII gTLDs, and could, as a "track", result
in
> applications for IDN gTLDs being accepted prior to applications for ASCII
gTLDs?
> >>>>>
> >>>>> I believe that was your initial statement, and I simply want to know
if you're
> opposed to the coupling of the cc and g IDN offerings, and favor either
(a) no
> gTLD IDN until gTLD ASCII or (b) no ccTLD IDN until gTLD IDN and gTLD
ASCII.
> >>>>>
> >>>>> Thank you for your patience. I won't ask anything else, no matter
how
> curious I am.
> >>>>>
> >>>>> Eric
> >>>>>
> >>>>> Stéphane Van Gelder wrote:
> >>>>>> Hi Eric,
> >>>>>> I think what I mean is fairly clear. That no category of applicant
should be
> given an application window before others. What happens once the
application
> window is open for all obviously then depends on the specifics of each
> application and the validation process.
> >>>>>> Thanks,
> >>>>>> Stéphane
> >>>>>> Le 23 nov. 2009 à 14:12, Eric Brunner-Williams a écrit :
> >>>>>>> Stéphane Van Gelder wrote:
> >>>>>>>> Sorry, I should have said "having some applicants be allowed to
> apply before others".
> >>>>>>> Could you suggest a rational, as "first to market advantage" will
exist
> where two or more applications are fully evaluated, contracts entered,
unless all
> the associated DNS data is added to the root as a single unit of change.
> >>>>>>>
> >>>>>>>
> >>>>>>> Or do you mean both "apply before others" _and_ "delegated before
> others"?
> >>>>>>>
> >>>>>>>
> >>>>>>> Or do you mean the "first to market advantage" to be just the act
of
> submitting an application?
> >>>>>>>
> >>>>>>>
> >>>>>>> If the "first to market advantage" arises simply from the act of
> submitting an application, what is that advantage? In what market?
> >>>>>>>
> >>>>>>>
> >>>>>>> Eric
> >>>>>>>
> >>>>>>>
> >>>>>>>> Thanks,
> >>>>>>>> Stéphane
> >>>>>>>> Le 23 nov. 2009 à 13:37, Eric Brunner-Williams a écrit :
> >>>>>>>>> Stephane,
> >>>>>>>>>
> >>>>>>>>> Thank you for the clarification.
> >>>>>>>>>
> >>>>>>>>> Its my experience that the round in 2000-2002, having only 7-10
> applicants, 3 of type sTLD (aero, coop, museum), and 4 of type gTLD (biz,
info,
> name, pro), was not executed to avoid sequential delegation and sequential
> launches.
> >>>>>>>>>
> >>>>>>>>> How can "some applicants [] go before others" be prevented?
> >>>>>>>>>
> >>>>>>>>> Eric
> >>>>>>>>>
> >>>>>>>>> Stéphane Van Gelder wrote:
> >>>>>>>>>> Eric,
> >>>>>>>>>> Seems I really must try and make my emails clearer ;)
> >>>>>>>>>> I am not in favor of a system which would allow some applicants
to
> go before others and hence gain a first-to-market advantage.
> >>>>>>>>>> Hope that is clearer.
> >>>>>>>>>> Stéphane
> >>>>>>>>>> Le 22 nov. 2009 à 15:15, Eric Brunner-Williams a écrit :
> >>>>>>>>>>> Stéphane Van Gelder wrote:
> >>>>>>>>>>>> Edmon,
> >>>>>>>>>>>> Your track A/track B idea is interesting.
> >>>>>>>>>>>> I maintain that track A as you suggest it requires that the
general
> idea of tracks be accepted under the new gTLD program. I cannot see where
the
> only track is an IDN track and other categories have to wait until the DAG
is
> finalised.
> >>>>>>>>>>> Are you arguing for "more tracks" or against "language as a
track"?
> >>>>>>>>>>>
> >>>>>>>>>>>
> >>>>>>>>>>>> Track B. This is a different topic, but it does tie into new
gTLDs
> in general of course as even if they were based on existing gTLDs, the IDN
TLDs
> referenced would constitute a new TLD.  That being so, it will be
difficult IMO to
> make a case for the early release of IDN versions of existing gTLDs, as
that
> would be giving entities who already have the advantage of being on the
market a
> first-to-market advantage for new gTLDs.
> >>>>>>>>>>> Are you arguing for "more TLDs" or for "no more TLDs until an
> unknown set of conditions are satisfied"?
> >>>>>>>>>>>
> >>>>>>>>>>>
> >>>>>>>>>>> Eric
> >>>>>>>>>>>
> >>>>>>>>>>>> Stéphane
> >>>>>>>>>>>> Le 21 nov. 2009 à 22:49, Edmon Chung a écrit :
> >>>>>>>>>>>>> First of all, I sort of agree with Bertrand (GAC) when he
said,
> perhaps it
> >>>>>>>>>>>>> is not so much about "fast" track, but just "track
differentiation".
> The
> >>>>>>>>>>>>> point is, what would be useful may be to have a focused
> discussion on IDN
> >>>>>>>>>>>>> gTLDs in parallel with other discussions with new gTLDs in
> general, i.e. to
> >>>>>>>>>>>>> discuss them in separate tracks such that each track would
not
> hold back the
> >>>>>>>>>>>>> other.
> >>>>>>>>>>>>>
> >>>>>>>>>>>>> Based on that, and the question of getting "beyond the
> criticisms that are
> >>>>>>>>>>>>> being used to stall gTLDs in general", I therefore sort of
> proposed a dual
> >>>>>>>>>>>>> track approach for IDN gTLDs earlier in the thread:
> >>>>>>>>>>>>> 1. IDNG Track A: IDN gTLDs operated as completely new
> gTLDs
> >>>>>>>>>>>>> - The new IDN gTLD will accept registrations completely
> separate from any
> >>>>>>>>>>>>> existing gTLD
> >>>>>>>>>>>>> - This track would need to address the overarching issues
> >>>>>>>>>>>>> - Focused discussion on IDN may (or may not) make resolving
> the issues more
> >>>>>>>>>>>>> effective
> >>>>>>>>>>>>> - Result of the discussion may (or may not) merge the
process
> back with the
> >>>>>>>>>>>>> full new gTLD process (or "track")
> >>>>>>>>>>>>>
> >>>>>>>>>>>>> 2. IDNG Track B: IDN gTLDs operated in accordance with
> existing gTLDs
> >>>>>>>>>>>>> - Only for existing gTLDs, but can be objected to by
> prospective IDN gTLD
> >>>>>>>>>>>>> applicant
> >>>>>>>>>>>>> - Existing gTLD registry must serve the same zonefile under
the
> IDN gTLD
> >>>>>>>>>>>>> (even though the IDN TLD itself would be a separate NS
> delegation at the
> >>>>>>>>>>>>> root) OR offer registration only to the same registrant OR
> implement other
> >>>>>>>>>>>>> registration policies to achieve the same
> >>>>>>>>>>>>> - Where objection is received and not resolved, then the
> application should
> >>>>>>>>>>>>> move to Track A above (that is the part where I proposed the
> "confusingly
> >>>>>>>>>>>>> similar" test)
> >>>>>>>>>>>>>
> >>>>>>>>>>>>>
> >>>>>>>>>>>>> Both IDN Track A and Track B can proceed together and in
fact
> can continue
> >>>>>>>>>>>>> to exist.  Track B can essentially be an ongoing process,
even
> for future
> >>>>>>>>>>>>> new gTLDs.
> >>>>>>>>>>>>>
> >>>>>>>>>>>>> On the topic of limitation on number of script/language,
first of
> all, the
> >>>>>>>>>>>>> IDN ccTLD Fast Track did not set a numerical limit, but
rather, it
> is based
> >>>>>>>>>>>>> on the number of official languages ("official language" in
its
> general
> >>>>>>>>>>>>> sense) in the cc locality.  For gTLDs, the issue would be
tricky
> if a
> >>>>>>>>>>>>> numerical limit is set and could raise technical, fairness
as well
> as
> >>>>>>>>>>>>> political issues.  Take ".Asia" for example, it would not be
> appropriate to
> >>>>>>>>>>>>> pick Chinese over Japanese or Korean, or Hindi or Arabic for
> that matter.
> >>>>>>>>>>>>> In the case of Chinese and Japanese (Kanji) (and Korean
Hanja
> if included),
> >>>>>>>>>>>>> the issue further complicates because of the overlap in
> script/character
> >>>>>>>>>>>>> usage, which could become a fairness issue, as whichever is
> launched first
> >>>>>>>>>>>>> could take away names available for the latter.  This is
also a
> reason I
> >>>>>>>>>>>>> think perhaps we should distinguish this discussion from
"fast"
> track, but
> >>>>>>>>>>>>> rather look at this IDN Track B as an ongoing track for even
> future new
> >>>>>>>>>>>>> gTLDs.
> >>>>>>>>>>>>>
> >>>>>>>>>>>>> Initially, perhaps limiting to non-latin scripts could work.
> Limitation to
> >>>>>>>>>>>>> 1 per language per script may be ok too, which I believe is
> similar to the
> >>>>>>>>>>>>> approach taken by the IDN ccTLD Fast Track.
> >>>>>>>>>>>>>
> >>>>>>>>>>>>> The EOI seems quite separate to this discussion I think.
> >>>>>>>>>>>>>
> >>>>>>>>>>>>> Edmon
> >>>>>>>>>>>>>
> >>>>>>>>>>>>>
> >>>>>>>>>>>>> PS. It seems we are gaining some momentum with this
> discussion.  Perhaps we
> >>>>>>>>>>>>> should schedule a few conference calls to further hash out
> possible ways
> >>>>>>>>>>>>> forward.  Will start a separate thread on this.
> >>>>>>>>>>>>>
> >>>>>>>>>>>>>
> >>>>>>>>>>>>>
> >>>>>>>>>>>>>
> >>>>>>>>>>>>>> -----Original Message-----
> >>>>>>>>>>>>>> From: owner-gnso-idng@xxxxxxxxx <mailto:owner-gnso-
> idng@xxxxxxxxx> [mailto:owner-gnso-idng@xxxxxxxxx] On
> >>>>>>>>>>>>>> Behalf Of Eric Brunner-Williams
> >>>>>>>>>>>>>> Sent: Sunday, November 22, 2009 12:23 AM
> >>>>>>>>>>>>>> To: Avri Doria
> >>>>>>>>>>>>>> Cc: gnso-idng@xxxxxxxxx <mailto:gnso-idng@xxxxxxxxx>
> >>>>>>>>>>>>>> Subject: Re: Process forward [RE: [gnso-idng] restarting
> discussions on
> >>>>>>>>>>>>> IDN
> >>>>>>>>>>>>>> gTLD]
> >>>>>>>>>>>>>>
> >>>>>>>>>>>>>>
> >>>>>>>>>>>>>>
> >>>>>>>>>>>>>>> ... what gets the gTLD fast track beyond the
> >>>>>>>>>>>>>>> criticisms that are being used to stall gTLDs in general.
> >>>>>>>>>>>>>> Good question.
> >>>>>>>>>>>>>>
> >>>>>>>>>>>>>>
> >>>>>>>>>>>>>>> Other questions (perhaps subquestions of the question
> above):
> >>>>>>>>>>>>>> - would it be for non-Latin scripts only?
> >>>>>>>>>>>>>>
> >>>>>>>>>>>>>>> - would it be for IDN gTLDs only
> >>>>>>>>>>>>>>> - would it be only for existing gTLD registries
> >>>>>>>>>>>>>> - would each existing gTLD registry would be allowed to
apply
> for one
> >>>>>>>>>>>>>> "similar" name
> >>>>>>>>>>>>>> - what changes to the existing registry contracts are
necessary
> >>>>>>>>>>>>>>
> >>>>>>>>>>>>>> That, and the item I added above, would make it most
similar
> to the
> >>>>>>>>>>>>>> ccTLD IDN FastTrack.
> >>>>>>>>>>>>>>
> >>>>>>>>>>>>>> I make that one museum in a non-Latin script, one cat in a
> non-Latin
> >>>>>>>>>>>>>> script, one aero in a non-Latin script, one ... , and one
com in
> a
> >>>>>>>>>>>>>> non-Latin script.
> >>>>>>>>>>>>>>
> >>>>>>>>>>>>>> For those counting, that's ~40 change requests to the IANA
> root
> >>>>>>>>>>>>>> arising out of the ccTLD IDN FT process, and ~20 arising
> from a gTLD
> >>>>>>>>>>>>>> IDN FT process, if "this" can be called a "gTLD IDN FT", or
> about
> >>>>>>>>>>>>>> 2/3rds of the budget Thomas Narten suggested was annually
> available.
> >>>>>>>>>>>>>>
> >>>>>>>>>>>>>>> - would applicants have to accept the most stringent of
the
> restrictions
> >>>>>>>>>>>>>>> being proposed by for new gTLDs (full IRT, no Geo related
> name of any
> >>>>>>>>>>>>>>> sort, no word that anyone on earth considers
controversial,
> nothing that
> >>>>>>>>>>>>>>> has the same meaning or etymology as any exsiting TLD ...)
> >>>>>>>>>>>>>> Do we care? Those that get hung up by objections simply
> survive the
> >>>>>>>>>>>>>> objection process or fail. Those to which no objections are
> offered
> >>>>>>>>>>>>>> progress. If a "gTLD IDN FT" is like (see above) the ccTLD
> IDN FT in
> >>>>>>>>>>>>>> the script, number, and equivalence restrictions, most of
> these
> >>>>>>>>>>>>>> restrictions have already been addressed.
> >>>>>>>>>>>>>>
> >>>>>>>>>>>>>>
> >>>>>>>>>>>>>>> How would this fasttrack combine with the EoI proposed
> process?
> >>>>>>>>>>>>>> I'm aware of a Board initiated EOI request for comments,
but
> not a
> >>>>>>>>>>>>>> proposed process.
> >>>>>>>>>>>>>>
> >>>>>>>>>>>>>> I'm aware "there is a group of participants that engage in
> ICANN's
> >>>>>>>>>>>>>> processes to a greater extent than Internet users
generally",
> who have
> >>>>>>>>>>>>>> agreed to attempt to restrict competition through
presenting a
> >>>>>>>>>>>>>> resolution to the Board proposing a "EoI process".
> >>>>>>>>>>>>>>
> >>>>>>>>>>>>>>
> >>>>>>>>>>>>>> I suggest that we ignore the latter, for several reasons,
and
> identify
> >>>>>>>>>>>>>> a gTLD IDN FT profile as a response to the former, as well
as
> a more
> >>>>>>>>>>>>>> general gTLD IDN, and the even more general TLD IDN
profile,
> to inform
> >>>>>>>>>>>>>> the Board.
> >>>>>>>>>>>>>>
> >>>>>>>>>>>>>>
> >>>>>>>>>>>>>>> I am not against this, but I am not sure I see how it
would
> help.
> >>>>>>>>>>>>>> I've a long note which I'll submit as a critical comment on
the
> >>>>>>>>>>>>>> Board's EOI question to the community. In a nutshell, I
don't
> think
> >>>>>>>>>>>>>> the EOI motion helped clarify issues. I'm even more
skeptical
> about
> >>>>>>>>>>>>>> the purpose of a private party "proposed EOI process".
> >>>>>>>>>>>>>>
> >>>>>>>>>>>>>>
> >>>>>>>>>>>>>> I think the next question isn't so much what about the EOI,
its
> what
> >>>>>>>>>>>>>> about the restrictions imposed on the ccTLD IDN FT.
> >>>>>>>>>>>>>> - non-Latin
> >>>>>>>>>>>>>> - one each (or two for CDNC territories)
> >>>>>>>>>>>>>> - limited "creep" from the iso3166 alpha-two value into
alpha-
> three
> >>>>>>>>>>>>>> or other standardized names (for which we have no
equivalent
> >>>>>>>>>>>>>> convenient standards to point to)
> >>>>>>>>>>>>>> - what changes to the existing registry contracts are
necessary
> >>>>>>>>>>>>>>
> >>>>>>>>>>>>>> In addition, there is the problem of expanding the number
of
> entities
> >>>>>>>>>>>>>> holding a distinct IDN delegation in their own right, under
a
> new
> >>>>>>>>>>>>>> contract.
> >>>>>>>>>>>>>>
> >>>>>>>>>>>>>> Are non-Latin scripts to be prioritized? That has been the
> basis for
> >>>>>>>>>>>>>> the ccTLD IDN FT, and for what I suggest above for a gTLD
> IDN FT.
> >>>>>>>>>>>>>>
> >>>>>>>>>>>>>> Are unserved populations to be prioritized? That is the
> foundation
> >>>>>>>>>>>>>> that the non-Latin requirement is based upon.
> >>>>>>>>>>>>>>
> >>>>>>>>>>>>>> I think we can help by suggesting not simply numbers to
> some EOI
> >>>>>>>>>>>>>> effort, but specific groups of applications with specific
> answers to
> >>>>>>>>>>>>>> "the criticisms that are being used to stall gTLDs in
general".
> >>>>>>>>>>>>>>
> >>>>>>>>>>>>>>
> >>>>>>>>>>>>>> Eric
> >>>
> >
> >