*** Disclosure: The following is the output resulting from
transcribing an audio file into a word/text document. Although the
transcription is largely accurate, in some cases may be incomplete or
inaccurate due to inaudible passages and grammatical corrections. It
is posted as an aid to the original audio file, but should not be
treated as an authoritative record.***


ICANN - Singapore
Trademark Clearinghouse Implementation Plan
22 June 2011

  >>MARGIE MILAM:  Okay.  Let's get started.  Okay.  Everyone please
have a seat.  We're going to get started.  Everyone... okay.  Good,
good.  I'd like to introduce Kurt Pritz, who will provide some
opening remarks, and then he'll hand it over to Karen, who will give
us an overview, and then the format for this session is really a
roundtable with our panelists to talk about implementation details.

 So let me hand it over to Kurt.

 >>KURT PRITZ:  Thank you, Margie.  Thanks for coming.

 This is intended to be a working session, so what we wanted to do
was have a big roundtable discussion about trademark clearinghouse
and it's implementation, but the rooms don't always lend themselves
well to the topic or the format we want to pursue, so bear with us.

 The idea is this:  That ICANN is in the middle of its new gTLD
program implementation that it's been working on for some time.  One
important aspect of that is secure the services of a trademark
clearinghouse provider, and the trademark clearinghouse has a few
duties.  It must take in data from trademark holders and maintain the
database of the trademarks that are -- and the other intellectual
property that's registered in the clearinghouse, and obviously that
has to be done with a great deal of accuracy and care.  So
maintenance of that database.

 And then it has to help administer at least two processes:  The IP
claims process and the sunrise processes for when new registries
actually launch.

 So to administer those processes.

 And so we are in the process of securing the services of a
clearinghouse provider.

 Rules for how the clearinghouse operates are in the applicant
guidebook.  The intended purpose of the clearinghouse, as I describe
with more specificity, and some rules for how it's going to operate.

 But really, there's a lot of detail to be worked out in order to
ensure two things.

 One is that the clearinghouse operates efficiently and well; that
in charging for its services, it's not charging too much; and in its
operation, it's not creating a burden for registries, registrars, and
others that would use the clearinghouse but, rather, sort of fits in
with its operation.

 So in our procurement of clearinghouse services and our
implementation of the clearinghouse, we want to consult with the
experts in creating that very specific set of rules for how the
clearinghouse is going to operate.

 Some of that work's already taken place.  Jeff Neuman, from
NeuStar, led an effort of different experts in fashioning some
operational rules around the IP claims process and how that might
work, and this session is really to build on that.

 So what we have here are some of my -- some of the friends I've
developed over the last six years in working on this to discuss
operational aspects of the clearinghouse, and we've invited registry
operators, registrars, to talk about the operational aspects and how
to make that economical for the registries and registrars to operate.

 And we've also invited members of the IP community because we don't
want to lose sight of the goal.  Why did we establish this
clearinghouse and what are the benefits we're trying to confer and
what are the protections for trademark holders?  We don't want to
lose that as we start to delve into the operational details.

 So we've asked these people up here -- you know them all, but Mary
Wong, J. Scott Evans, Nick wood, Jon Nevett, Russell Pangborn,
Jonathan Robinson, Debra Hughes, Joe Waldron, Jeff Neuman, Matt
Serlin, Bart Lieben, and Kristina, did you just -- ah, Kristina
Rosette just came up and so we're going to ask Karen to kind of lead
them through a series of three sets of questions, one about the
database, one about IP claims, one about sunrise, and discuss some of
the operational details here or what the thoughts are.

 Following this, what we want to do is carry on with our work and
continue an iterative discussion with the community and experts, so
as we, you know, do a formal request for proposals and write these
set of rules, we're writing them in a way that meets the goals of the
program.

 So thanks very much for coming.  I hope you find it really useful
and what we want coming out of this is, you know, any advice you have
for ICANN in implementing that.

 So with that, I'm going to turn it over to Karen and I want to
thank you guys in advance for all the thought you're going to put
into this.

 >>KAREN LENTZ:  Okay.  The agenda is up on the screen how we think
the session will run.  There are essentially three processes, three
major processes arising out of the trademark clearinghouse and what
it's intended to do, and we'll spend some time with each of those.

 As Kurt mentioned, the staff is here in listening mode.  These are
considerations that we're looking for that we need to be aware of in
going out to the RFP and finding a trademark clearinghouse provider.

 There's a rough time line up here.  

 Where we are now, June 2011 is developing process flows with the
community.  That's part of what we're doing now.

 The orange arrows along the bottom have to do with developing an
RFP and going out and making that process to secure a provider.

 And then further along, there's an integration of those two
processes to turn it into something operational.

 So the workshop goal, as has been said, is to identify the
implementation details we need to be aware of, collect feedback from
the affected parties here, the stakeholders and everyone here in the
audience, and then determine how we're going to proceed with those
discussions as the implementation continues.

 So the first aspect that we're going to talk about, the trademark
clearinghouse is a -- meant to be a repository of data concerning
trademarks and trademark rights.  It's intended as a central location
that will create efficiencies for trademark holders and others who
are part of the process by having a central location to go to for
this information.

 So the first aspect of this will be:  How does that data get
entered, and if I'm a trademark holder, how do I maintain the
information that I've put into the clearinghouse?

 And going through the thought process on this, the main questions
that have come up are here on the screen here.  How do we populate
the clearinghouse?  What kind of authentication process will the
trademark clearinghouse perform on that data?  How will they ensure
that the data is accurate?  How will users be able to access and
update that data when they need to?

 Essentially, there's a pretty simple three-step process here.

 When you go enter data into the clearinghouse, you're going to
provide information and supporting documents.  So the -- according to
the documentation that's been developed to date, the eligible
trademarks are those that are registered, validated by a court, or
protected by a statute or treaty.

 So the first step is obviously to provide the data of what sort of
trademark you have and how that can be proven or documented.

 There is also a set of terms and conditions.  You're attesting that
the information is accurate.  You're providing some consent to the
clearinghouse provider for that data to be used for certain mandatory
required services that the registry operators will be providing.  And
there's also an option whether you'd like the data to be available
just for the mandatory basic services or whether you'd also permit
the data to be used for certain ancillary services that may be
offered.

 And then there's an expectation that there will be some fee
associated with entering that data, and that's another detail to be
worked out.

 So going through the documentation that's been developed by the
community so far, the list I have up here is the data that's been
listed as what would be collected at the time a trademark holder is
entering their information in the clearinghouse.  That will differ
according to what type of trademark it is.  It may be registered, in
which case you're providing certain information.  It may be court-
validated or protected by a treaty.

 The line up there about the class of goods or services is something
that was listed in the STI's sample trademark claims notice, which is
something that we'll talk about later, but the assumption is that if
that is to be provided in the notice, that's something that needs to
be collected up front, although that hasn't been specified in the
other documents that I've seen.

 So that's something to discuss.

 And then in addition to collecting that information, as I said,
there's a declaration that the information is correct, will be kept
current, et cetera.

 There's a consent to use of that information for the purposes of
the clearinghouse, and then there's the option for how you're
permitting that data to be used.

 So the first question I'll pose here is -- you know, it seems
fairly straightforward, this is what we want to collect.  Is there
anything missing?  Is there some capacity that is -- will be expected
or needed by rights holders when they are entering this data?

 So I'll open that up to anybody who would like to comment.

 >>MARGIE MILAM:  Karen, it's Margie.  I think Amy wanted to provide
some explanation on the types of intellectual property that could be
included in the clearinghouse.

 >>AMY STATHOS:  Yeah.  So I just wanted to supplement the first
slide that talked about the registered marks or marks that are
protected by statute or treaty or validated by court are the ones
that must be recognized in one or the other of the protection
mechanisms.  Either claims and/or sunrise obviously it now has been
opened up to other marks that also constitute intellectual property,
but those can be put into the database, in the clearinghouse, but
aren't necessarily required, and the manner in which those will be
used will be something that will be up to the registry who is
providing its own other trademark protection mechanisms.

 >>MARGIE MILAM:  Kristina, go ahead.

 >>KRISTINA ROSETTE:  I apologize if I'm getting too far ahead of
ourselves, but I would say in terms of a copy of a registration, I
think it would be fantastic if it would be possible for the trademark
owner to -- if a physical copy is required, to have it be a simple
copy.  Getting certified copies, notarized copies, et cetera, can be
very time-consuming and when you're talking about what is likely to
be multiples, that expense can add up very quickly.

 The one thing that I would actually also encourage you to consider
is whether it would be feasible if the submission in lieu of a
physical copy of the registration could provide a link to the record
for that registration to the online publicly accessible trademark
database from the country that issued it, so that, you know -- again,
I think you'd have to -- and I think there is somebody from the PTO
who could probably speak to whether or not you can rely on links like
that.  

 But I think doing that -- I mean, let's take advantage of the fact
that so many of these trademark offices now have the records online. 
Certainly there are those that don't, but to the extent that there
are and you can, let's at least explore that.

 >>MARGIE MILAM:  Okay.  Jeff has a comment.

 >>JEFF NEUMAN:  Yeah.  I was just going to agree.  In part I guess
with Kristina.  I think -- I'm not sure why a physical copy of a
registration is actually needed, if you can have some automated
mechanism of checking the online Patent and Trademark Office.

 I'm not sure you could do a direct link, because at least with the
USPTO they don't allow deep links.  They usually -- at least up to a
few months ago.  I haven't done it in a while.  But it usually just
links into the homepage.

 But -- so -- or to the search page.

 But I think, again, an automated mechanism of checking with the
Patent and Trademark Office should be enough.  I think the physical
copy is just going to -- is just kind of -- I mean, there are some
offices that aren't online, and in those cases you would get the
copy, but otherwise I would just -- it should be automated.

 >>MARGIE MILAM:  Okay.  I think Nick has a comment.

 >>NICK WOOD:  Yeah.  It was a simpler point, really, about the data
to be collected that I think you probably need the data first
registration and the date of expiration as well.

 >>MARGIE MILAM:  J. Scott?

 >>J. SCOTT EVANS:  I think there needs to be some data point for
assignment documents, because many times the original certificate of
registration could be listed in the name of an entity that no longer
owns that, because they don't -- at least in the United States, and I
believe in most jurisdictions, they don't reissue certificates.

 So it's only through assignment records.  So you need to do some
sort of documentation to show whether there's been an assignment, so
that the person putting it in can show that they actually have title.

 Another comment to Amy.

 Amy, I think that the guidebook's use of "other marks" confusing,
and that needs to be changed to "other indicia," because we're
already talking about marks in trademarks, and if you just make that
one change, because I think the idea is there are other indicia other
than trademarks that might be protected by the local registry, and we
just need to make that clear.  Because I think there's some
confusion, well, if trademarks are here, what are other marks.

 >>AMY STATHOS:  Right.  And with respect to the assignment issue,
sometimes the database that has the assignment documents may be
different than the actual database that has the actual registration,
so looks like it's something that --

 >>J. SCOTT EVANS:  So I think as the data is collected, you need to
put in a bullet point that says, you know, if -- like you do for use,
"if applicable, assignment documentation showing title" or, you know,
something to that, but as a data point that will need be provided.

 >>MARGIE MILAM:  Thank you.  I think we have a comment over here.

 >>OLIVER SÜME: Thank you.  My name is Oliver Süme and on behalf of
dot Hamburg, and I have a question regarding the kind of trademarks.

 According to most of the trademark laws, at least according to the
national legal frameworks in Europe, there are very different kind of
trademarks.  You can enter a string as a trademark, which probably is
the most well-known one, but you also can enter -- or register a
picture or a logo as a trademark.  And in addition to that, you can,
at least according to the law in Germany, register tones that might
be a symbol for some commercial spots or things like that as a
trademark.

 And will all these different kind of trademarks will be entered in
the clearinghouse or is it just the string that is protected?

 >>J. SCOTT EVANS:  As I understand the guidebook, it's going to
adhere to the other sunrises and claims that have existed in the
past, and it is only for text marks and only protects the text.  And
I believe there is some allowance for the deletion of textual
representations that are not available in the string.  So if it has
an apostrophe, it's a plural possessive or if it has an ampersand or
a dash, those can drop out, but it would still be protected if it's
textually represented in the registration.

 >>MARGIE MILAM:  Thank you.  I have Bart.

 >>BART LIEBEN:  Yeah, maybe commenting on that, I think that you
should -- well, ICANN should look into a way on how to define the
word "trademark" because there's no universal definition of that.

 So, for instance, if you -- in quite a few PTOs, if you reject, for
instance, a trademark in Chinese script, then they say that it's a
device mark.  Certainly when Chinese is not the official language of
that country.

 So there are things where it can be dubious.  So, first of all, you
have PTOs that do not distinguish between word marks and not word
marks or device marks or figurative signs, combined marks, so this is
something where I think the guidebook or the guidelines should shed a
bit of light on what that could be.

 >>MARGIE MILAM:  Okay.  Thank you.  And Kristina, right?

 >>KRISTINA ROSETTE:  Two points.  And I goes this is a
clarification -- a question and then perhaps a suggestion.

 If it's the -- is it the intention that each trademark owner would
only identify one point of contact, so that -- not to pick on Russ,
but that it would be -- you know, Russell Pangborn, Microsoft
Corporation, ta-da, ta-da, ta-da, ta-da, ta-da, or can a trademark
owner put in multiple contacts?  Because I think what you might want
to suggest to trademark owners to do is that there would be one
individual identified, but that in order to facilitate effective
implementation of this, that there be a -- that they create some kind
of e-mail alias.  So again, not to pick on Russ, but that it would go
-- instead of his e-mail address, it go to
trademarkclearinghousequestions -- or whatever --  @microsoft.com, so
that people have a way to do that.  

 The other thing is that although I believe the IRT recommendation
was that registered trademarks be limited to the principal or primary
register of the country of origin, that delineation didn't make it
into the guidebook.

 And in the U.S., marks that are registered on the supplemental
register are deemed inherently descriptive and for purposes of
trademark claims I would think that a potential domain name
registrant would not know that.  So I think you are going to have to
indicate that at least for U.S. registrations, whether or not if it's
a registration on the supplemental register.

 >>MARGIE MILAM:  Good comment.  I think Debbie was next and then Joe.

 >>DEBRA HUGHES:  Just a quick comment, going back to documentation
for proof.

 I think for treaties and for national legislation, I agree with the
comment that was currently made about going to official databases
instead of sending in paper.

 I think there's ways to do that, Amy and Kurt, that make sense. 
There's official databases for many legislations throughout the --
throughout the country.  For example, for the Geneva convention, I
could point you to the exact location where you could find the exact
relevant sections rather than sending in paperwork.  So just kind of
piggybacking on the documentation issue for those types of marks.

 >>MARGIE MILAM:  Joe.

 >>JOE WALDRON:  Yeah.  I'd like to just expand a little bit on the
comment you made about a Chinese trademark that would be registered. 
I think we need to address all of the -- kind of the standard IDN
issues when that occurs in terms of what's being actually recorded in
the registry.  And then what's being transmitted to the actual domain
name registry, the TLD registry, to ensure that we don't have a
mismatch of the representation and the way that is converted -- you
know, that we use Punycode to be able to do that, so you could have a
complete mismatch.  Even though they may appear the same, they're
going to look as very different entries in the registry.

 >>MARGIE MILAM:  Okay.  Matt?

 >>MATT SERLIN:  Yeah.  So then Joe, would you -- would you enter in
the Punycode as part of the -- in the clearinghouse database?  I mean
do you think?

 >>JOE WALDRON:  To be honest, I don't think I have a solution at
this point.  I think it's an area that needs to be addressed to
ensure that there is consistency with the -- with the trademark the
way it's registered and then the representation of the -- of the
domain name that goes into the registry.

 >>MARGIE MILAM:  Go ahead.  Bart.

 >>BART LIEBEN:  Yeah.  So just to -- to follow up on that -- well,
in the interest of declaration, I am the architect of a clearinghouse
that's currently being used in a couple of TLDs, and that could be
part of the proposal for becoming the ICANN clearinghouse.

 The way how we do it is we have a representation of the mark and
then we have basically the representation of the domain name without
the extension, then, that could then be forming part of a sunrise or
a claim, so we have both fields, actually.

 So we have the Punycode and we have the actual representation of
the mark in addition to that.

 >>JOE WALDRON: Right.  But that also then implies that there's some
consistency and standardization across all of the registries that
you'll be dealing with, so that they've all done the same
implementation of different language tables, so there is a very tight
coordination that needs to be done between all of the gTLDs and the
trademark clearinghouse.

 >>BART LIEBEN:  Yeah, and even -- so the way we currently do it,
even if there's a different implementation, we basically lock the
implementation of the IDN tables to the extension, so we -- we can
translate all of them, if you want, so --

 >>MARGIE MILAM:  Thank you.  We have one more comment on this and
then we'll go to Karen who is going to go to the next slide.

 >> Yes.  Very quick comment and thanks, guys, for bringing this up.
I mean, the IDN part.  And you've talked about some details about
linking the converted Punycode to the -- to the IDN mark.  Just to
make sure that you add those -- for example, the details like
language tag and how you can do batch processing when you deal with a
large amount of trademarks -- I mean, registration in the
clearinghouse, just to add that.  Thank you.

 >>BART LIEBEN: And maybe very, very quickly with respect to the
name and contact information for the trademark holder of record, our
experience learns that in quite a few cases, this data that is kept
by the PTO is not up-to-date.  There are various reasons for that. 
For instance, if the company merges or if it changes its name, it
generally waits until the time where the trademark expires in order
not to pay additional taxes to the PTO for just changing the name, so
do take into account that about 15 to 20% in some of the PTOs, you
have inaccurate contact data.

 So basically what we do is we get the contact data from the brand
owner itself and if we see that there's a mismatch between what we
get and what the -- what we see in the PTO, we ask the brand owner
for an explanation.

 >>KAREN LENTZ:  Thank you, Bart.  That's actually a great segue to
the next slide -- oh, Nick, did you want to add something?

 >>NICK WOOD:  Yeah.  Just before we moved off this topic of the
data that needs to be collected, I wondered if I can ask ICANN about
the ancillary services that you imagine that the provider of the
clearinghouse could also offer.

 Because that goes to the heart of the type of data that's collected.

 Would you allow an operator to collect other data?

 >>AMY STATHOS:  So obviously we haven't imagined all the ancillary
services that a clearinghouse might want to provide, but there are
other services.  For example, a watch service is an ancillary service
that will proceed beyond the initial 60 days of the IP claims notice,
so it would be something along those lines, and whatever other
ancillary services there are, I'm sure there will be some -- some
innovation going on.

 >>BART LIEBEN:  So maybe --

 >>MARGIE MILAM:  We need to go on to the next topic, if that's all
right.  Sorry, I'm trying to manage the time.

 >>BART LIEBEN:  Well, I'll just very briefly mention that the data
that we generally get from the brand owner in 50% of the cases, it's
generally wrong so don't ask for too many data.

 >>KAREN LENTZ:  Okay.  So the next section of what we're going to
talk about here has to do with accuracy and authentication.  So the
requirements that have been developed have an expectation that the
clearinghouse will come up with a list of data authentication
sources, Web sites of patent and trademark offices in various
countries around the world, and then any third -- third-party
providers that might have access to that data and would be
authoritative.

 There are a few tasks here.  One is to verify the accuracy and
matching of the contact information that's given.  As has already
been mentioned, there are some issues with that.  And then the
authentication of the trademark data itself, so you are looking at
how do I make sure that you are who you say you are when you are
providing me with this data; and, second, how do I ensure that the
trademark documentation that you're giving me is real.  

 So any comments on how those steps should be -- should operate?

 >>MARGIE MILAM:  Kristina?

 >>KRISTINA ROSETTE:  I just have a threshold question, and I
apologize if I missed this.  Is it the anticipation that trademark
owners would be submitting the information directly?  Or would it go
indirectly through the registrars?  I would actually suggest the
former, but I'm just wondering what the intention it.

 >>KAREN LENTZ:  I don't think that's been established.  I think
we're open to suggestions on that.

 >>MARGIE MILAM:  I think that's right.  I think the initial concept
is it would be submitted directly.  If people have other ideas, we
are open to them.

 >>J. SCOTT EVANS:  I would say based on past experience, directly
is probably what registrars, if they have any sense, would want
because --[ Laughter ] -- in prior rounds there were more mistakes
made by registrars who were dealing with volume issues than there
were by -- and it caused a lot of service problems for them.  I'm
casting no aspersions, but it was a lot to take on.

 >>MARGIE MILAM:  Matt?

 >>MATT SERLIN:  No aspersions taken.  

 I think it is probably something that needs more discussion; but
from my perspective, the option to do it both ways frankly.  If the
trademark owner wants to go direct, if they want to go through a
registrar, I don't see why it would -- I don't see why we wouldn't
have that option.

 >>MARGIE MILAM: Nick and then Russ.

 >> Jonathan actually.

 >>NICK WOOD:  I would support what Matt said.  Not withstanding the
fact that the marks owners may be the best original source of the
information, many of them use agents whether they are registrars or
not.  I would say the ideal perhaps would be to have the option of
both and leave that in the hands of the market to decide.

 >>JONATHAN ROBINSON:  I was going to say the same thing.  What goes
into it is going to be twofold.  One is there are going to be costs
where people will want to do it directly to save money, and then
there is the other resource aspect.  If you don't have the horsepower
internally to do it, you may want outside agents to be the one to do
it for you.  So you should keep the options open.

 >>MARGIE MILAM:  Jeff.

 >>JEFF NEUMAN:  I was going to say same thing about leaving the
options open.  We are only talking about registrars and direct.  But
let's face it, there could be law firms.  There could be companies
out there that already do trademark protection that have nothing to
do with domain names.  So just leave the option open to submit
directly to the clearinghouse.

 >>MARGIE MILAM:  Anyone else?  Go ahead.

 >>NICK WOOD:  Just a very small supplementary point.  There is a
question of some sort of asset accreditation to engage with the
clearinghouse.  So it is a question that will need to be thought
about as well, under what circumstances and what tests do you have to
pass to engage with the clearinghouse.

 >>MARGIE MILAM:  Jeff?

 >>JEFF NEUMAN:  Yeah.  I'm not sure about that.  I think if you
have just a direct Web portal, I think you're not going to accredit
anyone to do that.  I think that just adds needless steps.

 >>NICK WOOD:  "Accreditation" is probably the wrong word.  But you
need some form of registration as we all do for any consumer Web site
or something somewhere for identifying yourself as a submission
agent.  On whose behalf are you doing this?

 >> JOE WALDRON:  And I also think it is not just a Web page.  It
could be an EPP connection from a registrar to the clearinghouse.  So
I do think that there is some level of validation or authentication
or something that needs to take place outside of just obviously going
to the Web site and filling in the data.

 >>BART LIEBEN:  I do agree.  I think it is absolutely necessary to
have different types of interfaces, so a Web-based interface which
one single brand owner with one trademark can use on its own.  We
have XML that we're using for bulk submissions.

 >>MARGIE MILAM:  Any other comments on this from the floor?  Okay.

 >>WERNER STAUB:  Werner Staub from CORE.  I have a question related
to that, which ends up almost the same question.  If you say should
the registrar be able to go directly or not, to send the data to the
clearinghouse, are we really sure there has to be one single,
monolithic clearinghouse or could it be organized in a way that we
have the UDRP, where we have more than one forum actually available?  

 If you look at the underlying data, contrary to domain name, we
don't have a situation where any string would be exclusive to any one
who possibly want to have a claim or a trademark.  It could be more
than one who have it.

 If you look at the problem that we have, very different legal
frameworks and the fact that even the data that has collected may
have to evolve naturally.  And this could be a question of supply and
demand that people are adapting in the context of their environment.  

 We might think of having a multiple component system.  But it could
be European, Asian or several Asians and North Americans, South
Americans component of the clearinghouse.  That would be separate and
exchange data with very simple synchronization methods.

 >>MARGIE MILAM:  Questions?  Jeff?

 >>JEFF NEUMAN:  I thought it was pretty clear in the guidebook now
that there is only one clearinghouse.  There could be multiple
validators in theory.  I thought that was pretty clear in the
guidebook that was approved.

 >>MARGIE MILAM:  We're getting confirmation from Kurt that that's
correct.  

 Kristina?

 >>KRISTINA ROSETTE:  I apologize for not raising this on the
previous slide, maybe it is this one.  But y'all are going to have to
figure out what you are going to do in terms of evidence of use.  I
don't even know where to tell you to start on that, to be perfectly
candid.

 >>MARGIE MILAM:  Do you want to go on to next slide?

 >>KAREN LENTZ:  Okay.  So the proof of use was, indeed, on the
previous slide.  That is a function that the clearinghouse would be
expected to provide.  So certainly we're open to suggestions on how
that can be effectuated also.

 Anything else on this slide?  Okay.  And then this is kind of to
wrap up this session, there are a few subprocesses that would go
along with maintaining data in the clearinghouse.  First of all, who
has access to it?  And what's the process for validating that access?
How do I update if I've changed my address or I've stopped using a
trademark or some other type of change?  How do I go in and make that
change?

 And then what is the process for removing a trademark from the
clearinghouse?  There is various ways that could occur.  If a
registration for example, was found to be invalid, there would be a
procedure for removal.

 So any comments on that aspect?  Jonathan?

 >>JONATHAN ROBINSON:  We think we were talking about this with some
colleagues last night.  Certainly one of the issues is thinking about
this in terms of a multiyear process because let's just -- I mean,
we've heard many, many numbers about how many applications might come
in for new gTLDs.  But take any number you'd like and envisage a
certain percentage, go through relatively easily and are out there
and will go through then a cycle of implementation.  

 But then as one goes down into the long tail of the applications,
these are likely to take longer and longer to come out.  And it may
well be that some go beyond, for example, an initial period envisaged
for the clearinghouse.

 So notwithstanding the fact that there may be other rounds of TLDs
even if one simply envisages a single round, those may take a
substantial period of time.  So it is a requirement, I would imagine,
to not only have an update process but a requirement to -- for the
clearinghouse to operate on an ongoing basis at least insofar as the
first round has a long tail but, as I say, extending potentially into
other rounds.

 >>MARGIE MILAM:  J. Scott.

 >>J. SCOTT EVANS:  I think it is very clear that there is going to
have to be something similar to a registry lock on all the
information once it's entered and been put there.  And then there
would have to be some sort of validation that's done now.  I mean,
they do that with my domains, the most valuable domains.  There is a
process I go through with my registrar.  And they're going to have to
have something very similar to that to unlock and let this
information be updated because you run the risk of some sort of fraud
occurring in a situation where it has just been entered in through a
portal.  

 So you are going to have to have something where there is some sort
of lock in the information that has to go through a process to get it
opened back up, even if you -- you know, a simple passcode that you
put into the system, I think, is -- we've seen is not necessarily the
best firewall for this type of thing.  But there needs some thought
given to that.

 >>MARGIE MILAM:  So making it more difficult to update the
information so that you avoid fraud or some inadvertent change.

 >>J. SCOTT EVANS:  Just as innovative as people are who do the nice
things, they are going to be just as innovative in doing the bad
things.  So we need to on the front end know for sure that given the
fraud we've seen in startup mechanisms in the past, dot EU being a
prime example, when it was much more complicated to be fraudulent
than it could be in this situation, we need to build in protections
now.  

 So I would say some sort of firewall or double firewall protection,
even if you give a trademark owner something more complicated than
entering a simple password to get into that information.

 >>MARGIE MILAM:  John, you had a comment?

 >>JONATHAN ROBINSON:  Real briefly just for follow up on Jonathan's
point, the clearinghouse is going to be used for URS in the future as
well.  So it is not only using it for new TLDs and their initial
periods but the URS will go on obviously forever.

 >>MARGIE MILAM:  Bart?

 >>BART LIEBEN:  I would suggest to keep this as lightweight as
possible for brand owners.  Due to all the complexities -- I see on
the slide here "penalties for failing to keep information current,"
as I mentioned, quite a lot of information that is officially
available is not current at this point.  

 So you would actually be imposing a requirement upon a brand owner
that currently does not exist even in trademark law.  So do keep it
as lightweight as possible if a brand owner says, Okay, I have taken
over this company or we have merged with this organization X, Y, Z,
that there is some kind of an update process, that there is indeed a
lock as J. Scott suggested because in EU, it was impossible to change
the contact data just in order to avoid this type of gaming that
potentially could have taken place which hasn't at that point.

 And certainly if registration is found to be invalid or if there is
fraud on behalf of the self-proclaimed brand owner, at that point in
time, of course, you need to have a removal process in place.  But
that was, of course -- what we in the clearinghouse have is a process
that basically encompasses that.

 >>MARGIE MILAM:  Okay, thank you.  Back to Karen.

 >>KAREN LENTZ:  Okay, thank you.

 I think there has been a lot of very helpful insights already just
in this section.

 I'm going to move on to the next section of the agenda which covers
the trademark claims process.  This is a process associated with the
clearinghouse where clearinghouse data is used.  Within the first 60
days of general registration, a new TLD or registry operator would be
required to provide this trademark claims service which involves
presentation of a claim notice to a potential registrant if they are
attempting to register a name that matches an entry in the
clearinghouse.

 And it also involves facilitation of notice to a trademark holder
that has entered data in the clearinghouse when a name is registered
that matches their trademark.

 This process involves several parties.  It involves the person
trying to register the domain name, it involves their registrar, it
involves their registry operator, and it involves the clearinghouse. 
So there are a few -- well, I guess a number of possible permutations
of this process and how the information could flow, how the
communications occur between all of these parties.  So that's where
we're going to look at next.

 There's been a group of registries and some others, I think, that
did some initial thinking on this and constructed a model.  

 And I'm going to ask, Jeff, if you would like to go through that
briefly.

 >>JEFF NEUMAN:  Okay.  I won't go through each of the steps.  I'll
just say the premise behind this model was, at the last minute, I
guess it was decided that both a trademark claims process and a
sunrise process had to be offered, which is fine.  The harder part
was the prospect of having to do it the first 60 days of, quote,
general registration.  And that was all the guidance that was given.

 So what does "general registration" mean?  Well, for different
registries, that could mean a whole bunch of different things.  That
could mean a first come, first served.  That could be a founders
program.  That could mean an auction.  That could mean probably 30
different things that I can't even think of right now.

 And so to come up with a model that was that flexible to handle all
those situations but also the premise of the fact that there may be
hundreds of these all going live around the same time, all with the
first 60 days, some were overlapping, the fact that even from our
experience of launching dot CO, being the back-end operator, in the
first 15 minutes, we got 100,000 registrations.  But that wasn't even
all the hard part.  We got 19 million create commands from registrars
and 5 million checks all within 15 minutes.

 So the discussion we had with the registrars -- with some
registrars and some registries was that we didn't think that relying
completely in realtime on a clearinghouse to do everything was
necessarily the best way to go for all situations and that we adopted
kind of a model that at least NeuStar has when we operate the North
American Numbering Plan which basically helps route all the phone
calls in North America.  So we maintain all the databases and all the
updates.

 But rather than actually having every call come through our
database, which would be a lot, we push out updates every 15 minutes
to all the switches of the carriers.  And so it is the carriers that
actually maintain those databases so they actually route the calls.  

 Kind of a similar model here where the clearinghouse would be
collecting all the information but would be disseminating that
information to the registries so that the registries could then host
the claim notices to display out to the registrants but also be able
to -- instead of the registry having to query the clearinghouse with
those -- you know, just one registry, 19 million creates in the first
15 minutes, which obviously would be a hardship if every registry is
using the clearinghouse at the same time, rather than doing that, if
the registry could query itself for the information and then spit
back the notices -- It is not just taking notices, it is also getting
input on whether they want to proceed with that registration or
alternatively cancel that registration.  

 So this model again without going through it just kind of walks
through how all of that would work in our minds.  Obviously it is
just a first cut.  But there is some -- a lot of thinking that went
around this, and so just -- I will just leave it at that.

 >>MARGIE MILAM:  Does anyone want to comment on the Jeff proposal?  

 Okay, Kristina and then Mary.

 >>KRISTINA ROSETTE:  Jeff and I have had this conversation, but I
think a lot of trademark owners are going to be very nervous about
having a registry have access to all of their trademark data.  And it
seems to me that assuming that we go with this model, the only way
you're going to be able to eliminate those fears is to put very
significant restrictions on the scope of what the registry can do
with it.

 And in order to do that, you're going to have to have language in
the registry agreement, which, as I understand it, is done.  So I'm
not quite sure how we move forward with this.

 >>MARGIE MILAM:  I think I had Mary next, right?

 >>MARY WONG:  So I had a specific question for Jeff, and then I had
a comment about the trademark claims form.  Should I bring both up
right now?  Okay.

 So I will first ask my question to Jeff.  On behalf of the group
that prepared the straw man proposal, it is a really specific
question.  And I think I'm just asking for some clarity on what you
have called Step 2 in the process where the clearinghouse is supposed
to provide, I think, a list of eligible registerable strings.

 And I'm just wondering if we could get some clarity as to what that
really means.

 >>JEFF NEUMAN:  Okay.  I will answer Mary's question first.  The
list of eligible strings are the strings minus the -- I will call
them, quote, illegal characters.  So minus the ampersands, minus some
of the other marks that aren't recognizable in a domain name string. 
So the clearinghouse would collect the trademark.  But then at some
point, it needs to convert it to the identical string.  So that's
what that meant.  It wasn't any kind of review.  It was really that
simple.

 And then on Kristina's question, she's right.  We did have those
conversations.  Those conversations happened interestingly enough in
1996 with the North American Numbering Plan when we were selected as
the provider of that as well.  And so, you know, I hear what you're
saying.  Those arguments were made back then.  And I think with
restrictions, as you said, I think it could be enforced.

 I will add, though, that the way to fix it is not by the registry
agreement but it is by making the registry sign an agreement with the
clearinghouse provider.  That agreement hasn't yet been set.  

 Without reopening up the guidebook or reopening the registry
agreement, we could perhaps work it that way.

 >>MARGIE MILAM:  I think Mary had another question, is that right? 
And then we will go back to you and you.  I'm losing track of the
queue.  We'll keep going.

 >>MARY WONG:  I can hold my comment.  It is not about the proposal.
It is about the form.

 >>MARGIE MILAM:  Okay.  Now we back to Jeff unless you finished
your comments.  And if not, Joe.

 >>JOE WALDRON WALDRON:  One, I just wanted to say the diagram that
Jeff and whoever else has worked on -- I'm not sure.  As I reviewed
this, I thought it was a very good description of a way that this
works.  But I think you can't read too much into it and expect that
it is the only way that it would work.  

 As Jeff mentioned, doing a large volume of transactions during a
land rush, you know, we see this pretty regularly as well.

 Many times we have registrars that don't do checks before they do
an add.  They will do the add first.  That changes the dynamic of the
process.  

 I think what's most important in this is really identifying the
responsibilities and roles of what has to happen, who has to do that
and where it fits into the work flow.  But there isn't going to be
one standard way that you can expect that all registrars will work
with registries.  Or, again, to echo Jeff's comment, you know, it is
not going to be a standard way of the way the domain name registries
work with the trademark clearinghouse.  

 In some cases, it will make sense to do a realtime transaction. 
And in other cases, it makes more sense to do a bulk.  And that
should be left up to the registry to decide.

 >>MARGIE MILAM:  I think I had Matt, and then I am over there with
Russ.

 >>MATT SERLIN:  I wanted to pick up on what Kristina said.  I don't
see why this would be any different than previous TLD launches where
the trademark information was passed directly to the registry.  So I
don't think that's a huge change.

 >>MARGIE MILAM:  Russ, did you want to comment?


 >>RUSS PANGBORN:  The response to that, it is different because it
is going to be all of your marks being put in simultaneously and it
is a significant amount of data.

 And the question I would have back to Jeff is:  If, in fact, the
registry did an agreement with the clearinghouse, would ICANN still
have compliance oversight over the clearinghouse in that scenario? 
That's the debate -- on the registry, I'm sorry.  So in that
scenario, then, if the clearinghouse is going to do something
violative with the data, how would you go about getting recourse if
there wasn't that compliance oversight?

 >>MARGIE MILAM:  Kristina, did you have a follow-up?

 >>KRISTINA ROSETTE:  No.  I mean, Russ answered -- gave the
response I would have given to Matt.  

 I just think it is really important for people to understand that
from a trademark owner perspective, if this is all going to reside at
the registry level and there's no contractual nexus between ICANN and
the registry leading to registry use of this data, I'm going to tell
my clients not to do it.

 >>JEFF NEUMAN:  So, yeah, good question.  I think that's handled at
least in our situation with the North American Numbering Plan again. 
Just as an example, it is handled with third-party beneficiary rules.
So you could grant ICANN third-party beneficiary rights.  And I'm
just abbreviating it.  It is actually pages of our agreement with --
actually with industry in that instance.  So there is a lot of
parallels.  I would be happy to walk through it. 

 I think your issues are very legitimate, but I think they're
solvable.

 >>MARGIE MILAM:  Jonathan?

 >>JONATHAN ROBINSON:  I want to come back to the previous point
about realtime.  So if we -- I mean, I'm not sure if we are done on
the dates or if anyone wants to respond to that.  I wanted to pick up
on this issue, as I understand it, what is proposed and envisaged is
that we are taking a pretty significant step forward here from what's
been done in the past in terms of the realtime availability query
followed by, you know, what would normally be a domain create
command.  We're proposing to insert something in a well-established
and universal process.

 And so -- and I just wasn't 100% sure -- I think, Joe, did you
mention that in all cases, it may not be realtime?  I just want to
make sure we're all clear that that is what is envisaged as I've just
described it.  We are putting a unique and new step, if you'd like,
into the process.  And that has particular demands.  It may not be,
as Joe said, Jeff's model is the only -- for the sake of shorthand,
Jeff's model is the only way of dealing that.  There may be others. 
As long as we are all clear on what we are trying to deal with.

 >>MARGIE MILAM:  I think Joe wanted to respond.

 >>JOE WALDRON:  We have a pretty high-performing system that does a
lot of transactions within the dot com registry.  But we still have
registrars that do local checks just further on performance.  If you
are trying to do a high volume of activity and your customers are
waiting for your responses, in some cases, it makes more sense that
you allow the registry to do a batch process, that I have that file
locally.

 Or you provide, you know, one of your services that is in my data
center because the location of where the registry is and the location
where the trademark clearinghouse provider, I could be talking about
tens or hundreds of milliseconds of latency and that's not going to
be acceptable if I'm trying to maintain the performance and meet my
SLAs.

 >>  Just to be crystal clear here, I think we are talking about
doing perhaps availability check against a copy of his own data, for
example.  Yeah.  So there are some subtle technical points here that
need to be fleshed out.

 >>MARGIE MILAM:  Excellent.  I think I have Chuck now who wanted to
make a comment and I forgot.  We will go to Jeff after Chuck.

 >>CHUCK GOMES:  Is this on?

 >>MARGIE MILAM:  Can we have this on, please?

 >>CHUCK GOMES:  Testing.  There we go.  Thank you.  

 Kristina made the statement that the applicant guidebook -- or the
registry agreement is done.  And so my question is really directed at
staff.  I think it is mostly done.  I think it is mostly done.  But I
believe -- And this is why I'm directing it at staff.  I mean, none
of us want a lot of changes from here on forward.  

 But if that's the snag in terms of making this thing work right, my
question to staff is there's -- I believe there's probably a
mechanism for making that happen.  Is that correct?  Or am I speaking
out of hand there?

 >>MARGIE MILAM:  I think Kurt will answer that.

 >>KURT PRITZ:  Chuck, was your question really about mechanism for
making a change in the guidebook or was it specific to this technical
question?

 >>CHUCK GOMES:  It was specific to the agreement that's in the
guidebook because the assumption in a lot of this discussion is you
can't change the agreement from this point forward.  And I'm sure we
don't want to do much there.

 >>KURT PRITZ:  Right.

 >>CHUCK GOMES:  But if that was a real critical issue, is there a
mechanism for doing that?

 >>KURT PRITZ:  The really short answer is yes.  After this, we'll
publish a version of the guidebook that will be under rev control, so
it will be Rev 1.0.  

 Simple changes, there will be a mechanism where staff might be able
to make simple changes.  More complex changes would have to go
through public comment and board review.

 Also, with respect to the clearinghouse, the guidebook establishes
a broad set of rules and then a provider has detailed sets of rules,
just like the dispute resolution providers.  You know, they provide
detailed rules that fit within the broader guidelines.  Anyway, I
hope that was helpful.

 >>MARGIE MILAM:  We will go over here because you have been in
line.  Please state your name.

 >> CHRIS WRIGHT:  Chris from AusRegistry.  I have a number of
comments I could make about Jeff's proposal.  But just putting those
aside for the moment, I think it is important to understand that
there are two distinct things going on here.  One early involves a
subset of the data that's maintained within the clearinghouse.  So
one is a registrar is trying to establish whether a name that is
attempting to be registered is going to hit a record in the
clearinghouse.  Now, all of the information contained in the
clearinghouse is not required for that check to be done.  All that's
required is the list of the names that are in the clearinghouse, and
that check can be completed.

 The second part of the process is then if you hit one of those
names that are in the clearinghouse, you then need to display a
notice to that potential registrant that does contain all of the
information that is within in the clearinghouse.

 Now, if you separate those two processes, I don't think many people
would have a problem with the first one being done with local copies
of data kept by registrars or kept by registries or so forth.  That's
just a big list of names.  So there shouldn't be too many issues with
that.  We don't even know who owns the names at that point.

 In fact, one entry on that list could actually correspond to
hundreds of registrations in the actual clearinghouse.  So nobody
should have too much of a problem with that.

 Now, the reason that in Jeff's proposal that he is trying to get
the rest of the data into his registry is because he doesn't want to
be relying on a third-party.  He doesn't want to rely on the
clearinghouse being up and available and working to have people
register names in his TLD.

 Now, if the trademark owners are prepared to accept the concession
on the notice that's being displayed, then we can eliminate that as
well and then we don't need to get -- registries don't need to get an
entire copy of all of the trademark base.

 Essentially, the concession would just need to be if the trademark
clearinghouse is down, registrations can go ahead with just a generic
notice that says, Hey, Mr. Registrant, the name you are trying to
register is in the clearinghouse.  But at the moment, we are unable
to get the details of the exact information in the clearinghouse and
here's a URL that you can go to later to get that information,
Registrant.  Do you still want to proceed?  

 This is only in the unlikely case that the clearinghouse is down. 
So registries get what they want.  They can still go ahead and do the
registrations while the clearinghouse is down.  But you guys get what
you want in terms of all that data that's contained within the
clearinghouse isn't out there in the wild with all the registries and
the registrars.

 So if mark holders would concede that would be an acceptable
approach and we put SLAs in the clearinghouse contract, I don't know,
99.9% of the time we are effectively only talking four hours a month
where it's possible that this modified notice would be displayed.  It
probably doesn't seem like a big deal.

 >>MARGIE MILAM:   Okay.  Thank you.  I think we'll respond to your
comment.  We'll have Jeff do that and then we will come back to Zahid
who has been in the queue.

 >>JEFF NEUMAN:   I think that's exactly right, and you all helped
with that too, so there were a bunch of people who were on developing
that model.  That's exactly right.  And I don't think it's three 9s. 
I think it's at least -- we have to do five 9s, right, at the very
least, which is only eight minutes or something like that.

 So I think that's the excellent question.  I think you posed it
really well to the trademark owners that that's what they are going
to have to put up with because we as registries can't just stop
registrations.  That would not be acceptable.

 So that's a good question.

 And to address the other, on a final point, this is just one model.
And I don't necessarily say this is the right model that we proposed.
But I do want to just emphasize that while having flexibility is good
for different models, I would stress that the registrars we talk to
did not want to have 500 different models for 500 different
registries.

 So you have got to kind of balance that.

 >>MARGIE MILAM:   Jonathan.

 >>JONATHAN ROBINSON:   Thanks, Margie.  Just a response.  I think
that's a really interesting and innovative suggestion, and it
corresponds quite closely with a systemwide build at net names a few
years ago where we had the online registration system and people
could query a -- for match and there was a kind of traffic light
system at that point that said red, proceed at your peril, you have
got an exact match in a trademark database; amber....

 And so I think a variant at or around that suggestion makes a lot
of sense where there is essentially a traffic light system related to
correspondence in trademark databases.  And as you say, as the
previous speaker said it, doesn't necessarily have to then throw back
all the data, depending on the status.

 >>MARGIE MILAM:   I have J. Scott and Bart, were you in the queue? 
And then we need to go back to Zahid.

 >>J. SCOTT EVANS:   One clarification.  What's in the notice is not
a trademark owners' issue except to tell them that their there is
something in the clearinghouse and do you want to proceed.  That's
all trademark owners care about, and then that we get a notice back
should they proceed to let us know that there's something.

 So all this other information were concessions that were required
by other people than trademark owners; okay?

 So that's not our issue.

 Next, I just want to say with all this, the back-end stuff,
trademark owners are not going to be caring about if there is
different models as long as the front end works the way it does.  You
put it in, they register, you get a notice.  That's all they really
care about.

 But the one thing I did hear earlier, we talked about innovation
and things, is different entry points for information that would come
in and might be regional, and that info -- and some of the data
points might be slightly different because of regional differences in
the systems.  Fine.

 The biggest thing, though, is whatever you do within some region
has to be standardized.  There can't be four different types within
one region and four different types within another region.

 Everybody needs to decide on what's going to go in the Latin
American region, and those are consistent across whoever takes it in.
Because where it gets complicated and the efficiencies fall out is
when there are five different data points, there are two here, three
here, all five over here, and that gets confusing for the people
trying to make sure they're getting the data into the system
correctly.

 >>MARGIE MILAM:   Got it.  Okay.  Bart, Mary, Zahid.

 That's the queue.

 >>BART LIEBEN:   We actually are looking at various options, so I
think that the model that was introduced by Jeff is a sensible one.

 One of the things that we're looking into, to have the registrar
doing an EPP domain check amount on the clearinghouse as an option.

 So at that point in time, you bypass one party.

 Could be a more efficient way in looking at it.

 >>MARGIE MILAM:   Mary.

 >>MARY WONG:   I am actually making a similar point to J. Scott,
but to follow up on it because he is absolutely right.  There are a
number of concerns over the notice, and before you run away with this
idea of the generic notice, which conceptually might be a great idea,
and certainly for expediency and getting things going, that's good,
but we do need to think about -- whoever it is, whether in the RFP or
the operator -- what's in that notice, because you don't want
something to just say you have bumped into a trademark.  Would you
like to proceed?  From the potential registrant perspective, there
are at least two concerns.  One would be the possibility of being
intimidated but the other would be the possibility of the dreaded "G"
word or some other kind of incentivization to proceed without due
inquiry.

 And I would caution against that.

 >>MARGIE MILAM: Zahid.

 >>ZAHID JAMIL:   Zahid Jamil.  Just a short question.  I like the
slide and I just wanted to ask at the risk of asking maybe something
obvious and starting with Bart and then other people on the panel,
what is the problem that you foresee having this process become
perpetual post launch as well, perpetually?

 >>JEFF NEUMAN:   Could I --

 >>MARGIE MILAM:   Jeff, go ahead.

 >>JEFF NEUMAN:   This is really an operational issue.  I don't
think we should talk about the policy.  That's done.  That's decided.
Let's save that for other arena.

 >>ZAHID JAMIL:   My question was first directed to Bart.

 >>JEFF NEUMAN: Okay.  Opine on policy.

 >>ZAHID JAMIL:   What would be the problem of having this happen
post launch perpetually?  Could this be done continuously even after
launch?  Would that be a problem?  That's my question, and maybe Bart
can answer.

 >>JON NEVETT:   We were under strict instructions we wouldn't get
involved in the policy discussion.  Thank you.

 >>MARGIE MILAM:   We are trying to do implementation, not policy at
this point.  I apologize for that.

 One more comment.

 >>WERNER STAUB:   The question is whether the list of domain names
has to be just something that the registries can find or is actually
given the fact that the domain or string is in the clearinghouse of
public record where the public should be able to look it up before
registration is made?

 So I understand that there is a question about confidentiality of a
trademark being there, but quite frankly, if you want to have a right
and then keep that confidential, this is kind of a contradiction.

 So why wouldn't the public be able to query this trademark
clearinghouse?  And we have got a very good protocol, it's called the
DNS protocol.  It's extremely efficient and I think most much us here
are precisely because of that protocol that could be used for it.

 >>MARGIE MILAM:   I think Jonathan was first.

 >>KRISTINA ROSETTE:  I was just going to say I think this point has
come up before and obviously I can't speak for all trademark owners,
but the general view is individual queries by the public of the
clearinghouse for a particular name is one thing, and is likely to
not be a concern.

 But I think what folks are really concerned about would be the
ability to do a search for all of the deposits made by a particular
brand owner.

 That is where I think people are going to have some real issues,
just because you are essentially talking about in many cases, here is
their most valuable portfolio of assets.  Here is all the information
you need.

 >>MARGIE MILAM:   Jonathan, did you want to comment?

 >>JONATHAN ROBINSON:   I think Kristina has answered it in the way
that I -- I guess I am trying to think about whether there is a way
of doing some form of exact match check without revealing.  

 So it comes down to a kind of WHOIS stroke availability check
within the trademark database, and the opportunity there is to do an
exact match check without any further data coming back.  But, I mean,
mining the database is clearly going to be an issue, so that needs
thinking about.

 >>JEFF NEUMAN:   Yeah, this chart does not address the issue of
whether the clearinghouse is publicly accessible or queried.  I just
find Kristina's argument about -- it sounds like the WHOIS argument,
I think.  But that's not meant to be addressed by this.  This is
merely the interaction between the clearinghouse, the registry, the
registrar, the trademark owner, and the registrant.  I think I got
all five.

 To answer Mary's question about the notice, the notice is actually
in the guidebook.  It was drafted by the STI group.  And I don't know
if it was modified much, but that's been in the guidebook.  So we're
not saying that the notice should be changed from that.  In fact, it
has to be in multiple languages.  I think in all of the ICANN -- or
whatever languages ICANN supports, like five or six or whatever the
number is.

 So that's another complication in there.

 >>MARGIE MILAM:   I'm going to go back to Karen because I know we
have to go on to additional topics.

 >>KAREN LENTZ:   I think Mary wanted to respond to the last one. 
Go ahead.

 >>MARY WONG:   Actually, it was about the form, but not on that
last point because I think we are talking about two forms, the new
idea of the generic form and then the form in the guidebook.  If I
can make a quick comment or question about the form in the guidebook.
And it does cover a lot of ground and it was the result of community
negotiation and compromise and consensus.

 I think my specific question now they are in operations mode is
with respect to the notification of the class of goods and services. 
And when the final form comes up, I hope it isn't just you bumped
into a trademark in class 2, or, you know, class 11 or whatever it
is.  Because there are things in there that registrants will need to
know what exactly is a hand tool, for example.

 So if some kind of link could be provided to the list of the very
least for information, instead of just the class or even the class
plus fabrics, that would be very helpful to applicants, I think.

 >>MARGIE MILAM:   All right.  Quickly.

 >>JEFF NEUMAN:   Just to respond, that's in our proposal.  All the
information -- All the information that's put into the clearinghouse
would be displayed back to the prospective registrant.  So it's not
just a general it's class 42 and then no description.  Everything
provided for the clearinghouse would be spit back in multiple
languages and all that good stuff.

 >>MARGIE MILAM:   Okay.  We're back to Karen now.

 >>KAREN LENTZ:   Okay.  Thank you.

 The next slide that I wanted to show was another depiction of the
process.  And this is something that staff had come up initially, and
this is our first attempt at how to scope out this process this,
trademark claims process and integrate it into the existing
registration process and how all the parties would work.

 And I put it up mostly to show that although there are some
differences with the graphic that Jeff described, it's actually
similar in a number of respects.

 One of the principles we settled on was that the parties should
interact with the -- with one party, and the party that is providing
communication should be someone they are familiar with and
understand, have some sort of working relationship with.

 So in that respect, the registrant is getting and receiving their --
the notices and communications from the registrar, not from the
trademark clearinghouse, which is an unknown entity to them.

 Similarly, the trademark holder is getting the notice at the end
from the -- from the trademark clearinghouse rather than some
registry or registry operator.  So that, I thought, was a positive
development that we had kind of come up with similar processes in
that regard.

 So the -- there's been a lot of helpful discussion on the
clearinghouse.  I have taken some -- a lot of things that need to be
clarified, a lot of work on the back end in terms of coordinating
between the registries and registrars.

 Any other comments anyone wants to make on the claims process
before we move on?

 >>MARGIE MILAM:   State your name, please.

 >>CHRIS WRIGHT:   Chris from AusRegistry.  Similar thing that's on
both this picture and Jeff's picture.  I would raise some questions,
is it necessary for a registrar to use the registry, effectively
being the registry acting as a proxy to the clearinghouse?  There was
mention before about registrars worrying about having different
implementations from all the different registries, et cetera.  Surely
it would be better for a registrar to implement once to the
clearinghouse and be done with it, regardless.

 And then there's also questions about given that I'm just
realistically doing an exact match string to a list of strings in a
database, perhaps if I'm a registrar, what I might do is implement,
for example, VeriSign's service and use that one to check all the
domain names in all the other TLDs just by changing the TLD on the
end and send all my claims queries through to them and so forth.

 So there are lots of potential issues with that, and I'm not really
sure it's required.

 Also, I'm not sure it's really required for the registry to proxy
back the trademark information from the clearinghouse back to the
registrar.  They are effectively just taking something from the
registrar, it sending it off to the clearinghouse, getting a
response.  They are just being a proxy.  It doesn't make much sense. 
Surely the registrars would rather just go directly to the
clearinghouse and get the information, rather than have to deal with
all the different implementations that the registries come up with.

 And then finally, just before when we were talking about the notice
and so forth, all the information that you guys want on the notice
isn't necessarily able to be displayed, and the current vision in the
guidebook sort of shows dot dot dot to show the information is too
big and can't be displayed and allowed to be truncated.  It mentions
it's allowed to be truncated.

 So you are actually expecting the potential registrant to go
somewhere else to get the information anyway, and presumably that's
the point of the public access to the clearinghouse so that the
registrant knows, and there is supposed to be a URL in the notice for
the registrant to click on and get more information anyway.

 So if the notice just contains that URL, and the registrant then
clicks on that URL and that trademark clearinghouse site happens to
be down, well, then, the registrant is going to have to go back there
later and deal with it.  But that's sort of back to the issue about
registries proxying all the information.  So that's it.

 >>MARGIE MILAM:   Jeff, you want to respond?

 >>JEFF NEUMAN:   Yes, in that one, we had discussed that, actually,
in a group, and I think the registries didn't want to necessarily
rely on just an assertion by a registrar that it actually did
everything; that because it's the registry's TLD, the registry wanted
to make sure that it knew, and it would actually be in the middle of
the process.  Because under your process, it sounds like the
registrar would do all the checking, the registrar would send out the
notice but there would be no way for the registry to actually know
that that was done.

 And so we didn't know, using that kind of model, nor did the
registries really want to necessarily rely on that, of the registrars.

 >>CHRIS WRIGHT:   You could deal with that contractually in your
registrar agreement, forcing the registrar to have done that stuff,
and then they can send a notification in the create command that
says, yes, I have done everything I am supposed to do so then you are
legally covered.  They are telling you that they have done everything.

 >>JEFF NEUMAN:   Right.  The registries in that group weren't
comfortable with just relying on an assertion.  That's why.  But I
understand your point.

 >>MARGIE MILAM:   Chuck.

 >>CHUCK GOMES:   Sure, it's the same thing, kind of, that Jeff is
saying, but I want to take it a step further.

 The way things are structured right now, the registry has the
contractual obligation, and so I'm not sure our legal staff is going
to be comfortable with that kind of process.  No that we don't trust
our customers, the registrars, but there is an important obligation
we have with regard to that.

 And so that needs to be kept in mind when we look at that process.

 >>MARGIE MILAM:   And I am going back to Karen now.

 >>KAREN LENTZ:   Okay.  Thank you.

 Just to quickly finish up briefly.

 We did want to touch on the sunrise registration process and how
the clearinghouse would be involved in that.

 The kind of basic questions are up here on the slide.  How will the
clearinghouse be involved in the sunrise?  How would they interact
with the registries/registrars, and how would they be involved in the
dispute resolution aspect of the sunrise process?

 So there is a requirement for a sunrise period when a new TLD
starts up.  That needs to be at least 30 days.

 There are some eligibility requirements.  A trademark should be
nationally or regionally registered, validated with proof of use or
court validated or specifically protected by a statute or treaty.

 And the registry has the discretion to create its own sunrise
requirements.

 And then another new aspect to this is that where there is a
registration in a sunrise period that matches an entry in the
clearinghouse that a notice would be provided to the trademark holder.

 This is a -- kind of a rough attempt to sketch this out.  It's way
oversimplified to be able to fit on a slide.  And the -- One point to
make clear about this is it depicts a scenario where the person -- or
the entity registering the domain name, the sunrise registrant, which
is the top row there, is not the same as the trademark holder that is
getting a particular notice.  So there may be some overlap there.  In
many cases, they would be the same.

 But you have got the sunrise registration taking place,
determination of whether it meets the eligibility requirements.  And
then when the registration proceeds, there is that notice that is
generated to the trademark holder that has registered in the
clearinghouse.

 So any comments or feedback on this?

 >>MARGIE MILAM:   Jonathan.

 >>JONATHAN ROBINSON:   Mine is just a very simple comment on the
flowchart here.

 I'm certain or almost certain that your diamond down at the bottom
and eligibility requirements refers only to trademark eligibility
requirements, but bear in mind many TLDs might be introduced with a
range of other sponsorship or other eligibility requirements.  It's
worth just making that crystal clear, that we are only referring to
those pertaining to the trademark clearinghouse.

 >>MARGIE MILAM:   Thank you.  Any other comments?

 Bart.

 >>BART LIEBEN:   I think one of the things you should look into is
if you do the check, so from the registry to the trademark
clearinghouse, so basically you are checking the sunrise registrant
information with what is in the clearinghouse.  So there must be an
exact much between all those data points that you are asking for.

 I think there may be a risk in doing that. 

 >> Yeah, that's a good point because oftentimes the domain name
registrant does not match exactly the trademark holder.

 >>BART LIEBEN:   Just maybe briefly explaining, and it's, of
course, difficult without having a real chart in front of us.

 The way how we currently implement is that a brand owner comes to
the clearinghouse, and all the various checks that need to be done
are done.  So trademark validity and so on are checked.  Also
eligibility requirements are checked.

 And when for each of the TLDs that are supported by the
clearinghouse that we've developed, when those requirements are all
met, the registrar that will sponsor the registration or will submit
the registration to the registry will get an AUTH code.  And that
AUTH code is then being submitted in the create command to the
registry.  The registry checks whether this particular domain name
submitted by this particular registrar has this particular AUTH code.
That in order to avoid gaming.

 So that's the way how we do it.  It's slightly less complicated
than this and prevents gaming, in my opinion.

 >>MARGIE MILAM:   Jonathan.

 >>JONATHAN ROBINSON:   I guess just in line with what Bart is
saying, really one has to envisage it in two separate packages.  The
one is the authenticity or accuracy of the trademark data and the
second is the correspondence between the domain name application or
the sunrise application and that trademark data.

 And they are two separate processes, in many ways.

 >>MARGIE MILAM:   Anyone else want to comment?

 Anyone from the floor?

 Kristina.

 >>KRISTINA ROSETTE:   I have a question.  I was just wondering how
in this process you built in the sunrise eligibility challenge
process.

 >>KAREN LENTZ:   That's actually the next slide.

 >>MARGIE MILAM:   Do we want to move on to that now or does anyone
else have questions on this?

 I think we will move on.

 >>KAREN LENTZ:   Based on the clearinghouse documentation, there
are these four grounds on which a sunrise registration may be
challenged.  One, the registrant does not hold an eligible trademark
registration.  Two, the domain name is not identical to the trademark
on which they base their sunrise registration.

 Three, the registration is not meeting the requirements; i.e., of
national, regional, court validated or protected by a treaty or
statute.  Or four, it doesn't meet the date requirements.  It either
was not issued before the registry agreement was executed or not
applied for beyond -- before the date on which ICANN posted
applications.

 So this is a new -- This is another aspect to the clearinghouse
which we'll open it up for comments on.

 >>MARGIE MILAM:   Werner.

 >>WERNER STAUB:  I just have a question.  I don't understand what
you mean by registrant here.  Registrant of what?  Of the domain name
or of the trademark clearinghouse --

 >>KAREN LENTZ:   On number 1.

 >>WERNER STAUB:   Yes.

 >>KAREN LENTZ:   The entity making the sunrise registration.  So
the registrant.

 >>WERNER STAUB:   The applicant for a domain name in a given TLD.

 >>KAREN LENTZ:   Right.

 >>WERNER STAUB:   And on three, it's the TM registration.  So these
are two different cases where something may be wrong?  They are not
going the same direction, are they -- do they?

 In one, we have somebody who says, look, you cannot have this
domain name because you are not eligible.  In the other one, the
potential domain applicant says, look, this trademark application is
not acceptable.  They are opposite.  Did I misunderstand?

 >>KAREN LENTZ:   If I understand your question, I think there's
some overlap or -- yeah, between one and three.

 One I envision the -- the registrant doesn't have any documentation
or it turned out that they didn't have an actual authenticated
trademark.  Or three, they did.  They submitted some documentation,
but it was not found to be what was required.

 >>WERNER STAUB:   Okay.  But this here does not contemplate the
case where actually whatever is in the sunrise -- sorry, in the
trademark clearinghouse turns out to be not applicable to the
specific case of a TLD.

 >>KAREN LENTZ:   Well, this is about an actual sunrise
registration.  So challenging a registration that has been made.

 I think you're asking about whether an entry in the clearinghouse
itself -- right? -- with the challenged.

 >>WERNER STAUB:   Can be challenged at this point, because this is
where things will be visible.  People will say, well, what do you
mean you have a right here?  No.  This is impossible.

 So then it should be possible to challenge.

 >>KAREN LENTZ:   Okay.  Thank you.

 >>MARGIE MILAM:   We'll go to Bart and then go to your question.

 >>BART LIEBEN:   Maybe a question to ICANN staff.

 To which extent will you allow registries to determine itself what
is an eligible trademark registration?  Giving an example, if dot
Berlin comes across and they say, well, in order to play in sunrise,
you need to have a trademark that's valid in Germany and in U.S.
because we have various Berlins in U.S., to have those nexus
requirements, is that something that the registry itself can impose? 
Or will it be required for every registry to accept every single
trademark that is in the clearinghouse in order to form a basis of
the sunrise?

 >>MARGIE MILAM: Amy, do you want to respond?  Or Dan?

 >>DAN HALLORAN:   Thanks, Bart.  This is Dan Halloran from ICANN
staff.

 I think the rules say that any trademark has to be eligible to
participate.  So they couldn't discriminate based on the trademark. 
But they could still impose restrictions on who could register in the
TLDs, which would kind of overlap that.

 Go ahead.

 >>JEFF NEUMAN:   So it was our understanding, basically, that it
was just any -- you would only have to accept sunrise registrations
for those registrations that would otherwise be eligible in that TLD.

 So if you are only accepting -- if your TLD is limited to German
registrants, then there would be no reason why your sunrise would
have to go beyond that.

 >>MARGIE MILAM:   We'll go here and then I think we are wrapping it
up because we are running out of time.

 >>BART LIEBEN:   So concluding, that would mean that the registrant
is not necessarily, then, the trademark owner.

 Yeah?

 For instance, if you are a U.S. company and you have a U.S.
trademark and you want to register a dot Berlin, but dot Berlin says
you need to be a resident in Berlin, then you need to give the right
to somebody else.

 Well, you license a trademark for -- 

 >>JOE WALDRON:  I don't think you necessarily need to give the
right to someone else.  Whoever applies for a domain name has to meet
all of the qualifications within that registry.  So if it's an
unrestricted registry, restricted, whatever the conditions are within
the charter of that registry defines the registration requirements. 
And the trademark check is one part of that.

 >>MARGIE MILAM:   Over here, please.

 >>J. SCOTT EVANS:   I mean we never had a problem, to my knowledge,
in a restricted small pool TLD that you had to meet eligibility
requirements to get in.  We had people faking who they were to get in
to say, "I'm a museum" when you're not a museum, but I own a
trademark for MOMA.  So I get it in because I licensed it from the
Museum of Modern Art in New York.  To my knowledge, I mean, the
registrant has to qualify to get into the registry based on being
eligible.  Then, if you past that threshold test and they are running
a sunrise, which they have to do, and you have a trademark
registration for the string you want, you get that unless they can
challenge you based on one of these.  And I think most small,
restricted TLDs -- or restricted TLDs also have a charter dispute
resolution policy that allows you to challenge them by saying wait,
wait, wait.  They are not a museum.

 >>MARGIE MILAM:   Tom.

 >>TOM BARRETT:  Tom from EnCirca.  

 I wonder if there is a clause missing here to challenge a sunrise,
which is challenging its use, whether it has actually been in use or
not.  The word "use" doesn't appear here.

 >>MARGIE MILAM:   Werner.

 >>WERNER STAUB:   I'm still unsure about this, the question of
whether the presence in the trademark clearinghouse would be
sufficient to reject any applicant.

 Suppose there is a trademark for something like Champs Elysees, a
wonderful name, and there isn't actually an application based on
other rights, not trademark rights, for a domain that would be Champs
Elysees dot Paris, specifically in Paris where the place is.  

 Now, on what grounds could any trademark anywhere in the world say,
look, if you are stronger than whoever is the legitimate
representative of something that is a local resource and has been
appointed by the city, or whichever way this has happened.

 So I don't think the trademark can be considered to be, in all
cases, a superior right, especially as -- if you look at city names --
it is even by law not allowed to make them trademarks.

 So there is many trademarks, actually, that are morally less strong
than a collective right to that name by many, many people.

 >>MARGIE MILAM:   Would anyone like to reply to that?

 Isn't that the eligibility requirements, though?  Separate from the
trademark?

 In other words, in the situation that you are talking about, you
have got the trademark rights, but you also have the additional
eligibility requirements that the registry might have.

 >>WERNER STAUB:   Well, taking this example, there could be a U.S.
trademark, Champs Elysees, and you have to question whether the city
of Paris can put its monuments and place names into domain names. 
And by what right would some trademark somewhere else in the world
stop this from happening?

 >>MARGIE MILAM:   Right.  And I think that's where the registry
would decide in their -- am I correct -- in their eligibility
requirements whether they have some policy to deal with that.

 >>WERNER STAUB:   So the registry could say trademarks that are not
of a certain realm would not count.

 So the registry or a TLD could say that only trademarks of a
certain realm would count in terms of the clearinghouse.

 >>DAN HALLORAN:   Dan Halloran from ICANN staff again.

 I think what we are trying to say is only registrants from a
certain realm can apply.  They can't discriminate based on the
trademark but they can discriminate based on the registrant.

 >>WERNER STAUB:   But in the specific case, the opposite problem. 
We've got a trademark that has absolutely nothing to do with the
place.  They chose (indiscernible) outside, and they were concerned
possibly about this being used somewhere in a TLD about where they
are.  Here, we are in the place that -- actually where there are
collective rights with that name and where is the original place, and
why would somebody from the outside be able to say this domain name
cannot exist.

 >>DAN HALLORAN:   So to be eligible for a sunrise registration in a
restricted TLD you would have to both meet the sunrise trademark
requirements and the registration requirements for that TLD.

 >>WERNER STAUB:   Yes.  But the problem is we don't have a
mechanism to at least challenge, in that specific context, the
clearinghouse registration.

 That is missing.  We cannot have a situation where any registration
of a trademark anywhere in the world is going to stop things
happening somewhere else have nothing to do with it.

 >>DAN HALLORAN:   I think we are misunderstanding something.  We
can talk off-line.

 >>J. SCOTT EVANS:   What I think he is coming down to is to say if
there are terms that have -- are words in the dictionary that have
meaning outside the trademark and somebody owns it, so -- and that
somebody can say, "We have rights other than trademark rights in this
term, too," so they can't get a sunrise registration.  Okay?  I can't
get it because I'm Champs Elysees, and what I am is I am the shop
owners of the Champs Elysees, and I can't get a sunrise registration
because I don't have a registration, but there is a U.S. trademark
owner that is trying to get it, and he is saying there is no way for
the Champs Elysees shop owners without a registration to say, "We
have rights, too, and you shouldn't give them that premium.

 >>DAN HALLORAN:   So I think another thing that is missing here is
registries could have other registration policies, like sunrises for
things other than just trademarks, this kind of trademark.  They
could have other sunrises or other --

 >>J. SCOTT EVANS:   And I think that's been done before.  That's
not a concept that's new.  I think Asia did something similar to that.

 So you just have to look at other models that have been done before
to deal with these situations, but they are out there.

 >>MARGIE MILAM:   I am going to now hand it --

 >>WERNER STAUB:   -- handled outside of the sunrise mechanism,
possibly, before.

 >>MARY WONG:   I think all he is saying is if you look at specific
sunrise processes run by different operators, they are able to put in
other requirements, processes, different subphases, and there has
been experience in the past, say as in dot Asia, where that has
happened.  So that kind of thing might be a useful model for whoever
the operator is in this case.

 >>MARGIE MILAM:   Thank you.

 I am going to hand it now to Kurt.

 >>J. SCOTT EVANS:   It just needs to be remembered that is minimum
floor.  This is only a minimum floor that the registry has to do. 
They can do more if they so choose to do more and if it fits a
business model.  So there is innovation that can be built on top of
this.  This is not all you can do.  It's the least you have to do.

 >>KURT PRITZ:   So first, thank you for the fantastic amount of
information that was provided just during this really short period of
time by the panel and by those who participated in the audience.

 I'd offer you a consulting fee, but evidently we have this anti-
revolving door thing now, so....

 [ Laughter ]

 >>KURT PRITZ:   But I can offer you the continued opportunity to
work on this problem with us.  Just this session indicated, I think,
there's X more amount of hours to work to flesh out details and to
get an understanding so that when this implementation is done, it
meets our original goals and does so in a way that works good, stacks
neatly, and is available for a reasonable price.

 So thank you very much, everybody.  Thanks for attending this
session.  I thought it was great.

 Thank you, Karen and Margie, for all your work on it.

 [ Applause ]