Workshop:  The First Mile:  Additional Solutions for a Successful gTLD Launch
 ICANN Meeting - Cairo
 Wednesday, 5 November 2008
 
 >>ERIC BRUNNER-WILLIAMS:   I've used up my amusing story.
 >>AMADEU ABRIL I ABRIL:   Ten years later, we still wait for a stable system for TLDs.
 >>ERIC BRUNNER-WILLIAMS:   Yes.  It's not self-promotion.  It's just those are the three things I happen to know rather well.
 >>AMADEU ABRIL I ABRIL:   Eric --
 >>ERIC BRUNNER-WILLIAMS:   I have dry mouth and my water is far away, but we're ready, so I can start.
 All right.  The panel before you is Dirk Krischenowski, from the dot Berlin proposal, who lives in Berlin; next is Nacho Amadoz, from punct cat in Barcelona.  The next speaker will be Nick Wood, from (saying name), and Amadeu Abril i Abril, I'm sorry, also from Barcelona.
 And then, finally, Nick Wood and Bart Liben, from Brussels.  We had a cancellation.  Unfortunately, Bertrand De La Chapelle, who was going to be one of our speakers, had to cancel in order to do some GAC stuff.
 I guess I can't be more precise.  Yes.  A sign of respect.
 We have with us, courtesy of Nick Wood, about 50 copies of "The Perfect Sunrise," which I expect many of you got in Paris.  So those of you who didn't get it in Paris, we have some more.  And I'm not going to give them out now, because you might read them instead of listening to our distinguished panel.
 Without any further ado, if you're ready, Dirk.  I'm sorry, Nacho.
 >>NACHO AMADOZ:   Thanks, Eric.
 Well, my name is Nacho Amadoz.  I am the legal advisor for dot cat domain and registry.  And we're talking about the first mile.  But, as we don't use miles, we have renamed it after the first kilometer.
 And we're going to do some briefing about how was the sunrise and the facts related to it.
 First, we have to talk about how we are now, right now.  There's been two and a half years after the sunrise, and we have 32,000 domain names, which is not bad.  High rate of these domain names have already hosts, 75%.
 And it was increased before the summer.  For the summer, it was 40%, I think, and it got to 75%, which is a high increase.
 The renewal rate is also high, 90%.  But it would be even higher if we would let out some promotions directed to individuals in some (inaudible) that had not such a high rate.  So we could be talking about 95% or maybe higher.  That shows a healthy growth.
 And the number of registrars has been increasing constantly since the very beginning.
 We started with some of them, some of them located in our area, which was really helpful.
 But after some registrars from abroad saw that the domain name was successful, they decided also to step inside.
 And talking about the contents and the people who showed these contents.
 We have, as you see, lots of different situations.  We have, of course, people from multinationals that wanted to address their services and to show what they were doing to the Catalan-speaking community.  It's (inaudible) administrations.  But more than that, we had the community, which was the strongest support for dot cat, and who finally gave it its sense.
 These are the (inaudible).  Not much to say.
 And this is important, which is to focus on quality and not quantity, whatever the definition of quality is.
 Because having a lot of domains is good, because you need the income.  But that income wouldn't be -- wouldn't be sustainable and wouldn't be justifiable if it's not because there is a quality behind it.
 There are other metrics --
 >>AMADEU ABRIL I ABRIL:   Names for cats.
 >>NACHO Amadoz:   For example, names for cats, as Amadeu has told me.
 Other metrics can measure this quality.
 You see the domains per market size, 3.2 domain names for thousand speakers.
 You have also the indexed documents by Google, which is more than 9 million.  That is a high number.
 And the quality can also be seen through the number of conflicts and problems that will the sunrise and the past sunrise created, which was really -- a really low number.
 Three of them were UDRPs.  One of them especially was not a (inaudible), because the people who registered the domain name originally were a local dealer.  So perhaps it was lack of communication with the local dealer and the other place that gave origin to that problem.
 And we had a mediation procedure, and one eligibility criteria problem, which finally gave us the reason.
 The renewal rates, as we told you before, are quite high.
 So the conclusion we can draw from that is the dot cat, of course, is top among all other TLDs, which might be replicated through other top-level domains direct to the community.
 The facts of our sunrise.  Numbers show real interest from the community, not only before debate, but during debate and after debate was successful, because it shows a total of 20,000 registrations from applications and registrations.  And those three disputes we talked about.  The media coverage was important, for free, and no influences.  For free seems to be an important issue, regarding the fees and all the stuff we're talking about these days.  But no influences shows also that there was a real, real interest from the community and from the local magazines and newspapers that wanted to reflect that issue.
 Another important thing is that 100% funded from registrations since day one.
 And those zillions of hours answering registrants' questions and manually validating registrations.
 We should enter into the dial of the validation, of the manual validation later.
 Before the D-day, you have to set a plan.  That plan must not only be related to the registration process, but how is it going to work in relation to the eligibility criteria, which is the main thing to decide, because it sets how you can outline the rest of the processes.
 The post sunrise registration process is also something to -- well, to think about.  Because it might give you an idea on how to decide the sunrise registration process in order to be able to adapt it to another situation.
 The local registries, as we said before, are the ones who know better the community, and they can help during the process to get the community even more involved.
 The objective sunrise rules is something really, really important because it's not about judgment but about check-in.  And if you have these predefined registries inside your registries, you must not yet guess whether an application is good or bad.  Because it's not whether it's good or bad.  It's about valid or invalid regarding what are the members of this community.
 You can see here some of these directories.
 Ones are from the public administrations.  Others are coming from the private sector regarding the media associations, the film producers associations, the museums, sports, and so on.
 These people predefine their communities.  So you can use this prior definition as a tool in order to not get more (inaudible).
 If you select these directories, it's easier just to stick to them.
 The registry policy defines its life, during and after.
 Policy shapes domain's desired properties.  That was in relation with quality.
 As I said, whatever quality is, you must define how you want the domain to be, who is it addressed to, and that comes through that shaping of the policies.
 The policy can help as well to reduce the conflicts.  A well-defined policy can reduce to almost zero the number of conflicts that might arise.  Because you might be able to see who is trying to get in and see if it fits into those predefined categories that are the basis of the community.
 And of course for those who are not lawyers, to synthesize the policy, because it does not only help you to know how to apply it and how to use them, but it also helps registrars and registrants to understand them well.
 If it's a very long list of terms and conditions, nobody would really know what it's about.
 Here we have the definition of the three phases we used during the sunrise.
 Phase 1 was the one that had the higher priority.  And that priority was not only in terms of time but also in terms of quality, being, as it were, these fit into this phase, the ones that had the highest priority over the others.
  Even if the others came earlier.  That is because they were entities that were really related to Catalan language and culture, and that definition can be broader or can be more narrow, but it's a definition from those directories that sets which is the main area.
 Then we have second phase, the entities with pre-existent Catalan content online.  And if they were smart, they would apply for the first phase better than the second, but you could find some people who wanted to go to the second phase not knowing they were into the first phase.
 And there's also third phase, addressed to all the individuals that gave explicit support.
 Which names can they get?  That is another thing to define.
 Must they be names that they already have in other domain names?  Must they be domains related to their activity?  Those are things which must be thought of before definition, and it will show some gray areas, but the lesser the better.
 One year after the sunrise, there were symptoms that the community was quite ripe to change from this pre-validation context to a post-validation context.
 Domain content, domain numbers were already there so they gave shape.
 So it was not , how do you say, as necessary as it was in the beginning to check every validation manually, to check every domain name and to check who was applying for every domain name.  That is something that can be checked after if you have that strong base.
 So the post-validation process is a semi-automatic process which starts with an e-mail that the registrant receives when he tasks his application that leads him to a URL that he must validate.  And from that moment, the domain name goes live.
 He must do all these actions in order to first check that the mail address he is giving is valid, which is an important thing.  But he is also accepting the policies and the terms and conditions, which is also something that he should have been doing when he is on the (inaudible) Web site, but double-check makes it better.
 The result is clear rules, easy to understand rules, and that makes easier policy informant.
 There was also an issue about IDNs.  We treat them as a unique domain name.  That was something to help foster the people to get into the IDNs to use them, because they were receiving them -- receiving the domain name they were applying with the IDN and the non-IDN, if they were applying for the IDN, in order to help them to use them.
 From the registry perspective, they are a single object.  That comes an example, the name for the not for profit which is in the register of the dot cat which is fundacio which in cat Catalan means a not for profit.
 And we got them both for the same price.
 And (inaudible) during the presentation, but policy helps to get significant content fast because people know where they are getting in.  Most people already know it because they are perhaps giving support to the bid.  They are interested in being there, but others may come from that definition.
 Good policy avoids conflicts.  That is easy to understand, even for not lawyers.
 There is a focus on customer satisfaction for the main purpose of that quality definition, that must always go together with quality.
 And first quality, and then quantity, because that quality is our main value.
 Quantity may come, but not from nonquality.
 Thank you very much.
 [ Applause ]
 >>ERIC BRUNNER WILLIAMS:   Thank you.  We will do questions at the end.  We'll go on to the next speaker.
 So please hold your questions.
 Dirk, you are on.
 >>DIRK KRISCHENOWSKI:  Berlin is moving north.
 Hello, my name is Dirk Krischenowski.  I am founder and CEO of the dot Berlin initiative, which will apply for the top-level domain Berlin for all Berliners.
 I am proud to make an interesting announcement today because we, as cities, that means Barcelona, sitting here in the room as well, hamburg, New York, Paris and Berlin founded yesterday a stakeholder group which has certainly common interests, and this group will participate in the structures, in the framework in ICANN.  And to bring in our interests there.  And I can encourage other groups as well to, if they are common interests, to found such groups.
 In our group we also discussed best practices in city top-level domains, and one of the best practices I would like to show today in example is the city TLDs namespace mandate.
 So the city is concerned, and we are working with our, as explained, together with PDX, that's for Portland, a new initiative, and Quebec, but Quebec isn't sure if it's a city TLD or region or both.  We will find out in a while.
 So what is a namespace mandate, or a namespace management approach?
 A namespace -- a city TLD namespace mandate a voluntary, that's important, obligation of a city top-level domain manager to build a reasonable framework for the allocation of domain names.
 This can include a lot of things, I will show you later, but the framework should be, and this is important, fair and transparent to the community you are applying for, and should be based on a multistakeholder consultation and approach.
 We did this in Berlin, and we had a very good experience.  We, like ICANN, like to involve all the interest groups that like to show up and speak up.
 The city TLD mandate is designed to foster the use, the growth and added value for the city TLD and for the concerned city community.
 Why do we search namespace mandate?  If a namespace mandate had been in the establishment, for example, in dot US or dot org or other top-level domain, you wouldn't see an empty page on www.whitehouse.us.  It's a name page where nothing is on.  And people are typing in things like whitehouse.US because it's like the parliament of your country and you expect something to be there.
 And this is something people don't like.  They are disappointed when they go to such domain names.
 How can namespace look like within the framework of the following sunrise, as we heard from nacho in dot cat and the land rush potential auction.
 This namespace mandate can be -- we are divided into sub-mandates.  For example, names which are delegated before the sunrise to the sovereign interests of governments, institutions, courts, schools, and other public bodies and institutions.
 Then there might be competitive interests of the city.  Cities are in competition with other cities, so tourism, culture, sports, science is very important.  And a name like visit.Berlin, visit.NYC, those are the striking names city marketing is using.  So they should be using it, and a mandate to get the name, might be very interesting.  Then we have community assigned terms also for reasonable city and cultural TLDs top-level domain names, cities like the yellow page terms like roofer, city map, traffic and others, should be, you should not find a single roofer or this domain name is for sale.  You should find pleas at least a list of roofers where all the roofers in that city are listed.
 Language versions of the terms are important because many cities have big communities.  In Berlin we have the Turkish community with, I think, 20% of our individuals in the city citizens.
 And maybe also terms reserved for WIPO II recommendations, INNS, international organizations, geographical names and so on.
 The namespace mandate we think is a key success factor for at least a city and the regional TLDs but also maybe to other TLDs because the domain names of this mandate, they can immediately start it when they are allocated, immediately start when the TLD is in the root and not after sunrise, after land rush or whatever.
 This can be days after.
 If you prepare with your community these names, the Web sites can immediately go online and create a lot of visibility.
 Certainty is also important.  Whitehouse dot US is another important example.
 Why don't people trust it?
 Direction, people search not only with search engines.
 We have different studies seen in the last years that many people still type in the domain name.  If it's in the Google browser area, if it's in the browser area.  So type in is important, and people will type in whitehouse dot US or hotel dot Berlin or taxi.paris, and it's important that they find something.  And this creates good direction for the users.
 And last term I have here, intuition.  Hard to pronounce word.
 But this basic order of the namespace mandate will establish an intuitive and somehow semantic use of the domain names which are delegated by this namespace mandate.
 Yeah, examples here might be for Berlin city hall.  Dot Berlin kindergarten, or roofer dot Berlin.  People should find exactly what they are looking for under these domains and this is an important task.  The ccTLDs and most of the other gTLDs didn't have the chance, because they were running on first come, first served, but the new TLDs top-level domain names can do it better.
 This is lie last slide.  Namespace mandate might be a very flexible tool.  It is not only for geotop-level domain names but also might be interesting for industry communities, like car, blogging, pharma or whatever.  And also company TLDs where people find at www.jobs@ibm, exactly jobs at IBM.  Not to be in competition to the jobs top-level domain names.  I don't want to confuse with that.
 So that's my view on how namespace mandates for new top-level domain names can be a very good tool which is great for the community and also for the Internet user.
 Thank you.
 [ Applause ]
 >>ERIC BRUNNER WILLIAMS:   Thank you.  During Dirk's presentation, he has one pronunciation problem that was caught in the translation.  He said visit Berlin, not wizard Berlin.
 >>AMADEU ABRIL i ABRIL:   Okay.  Now you will see something absolutely extraordinary in an ICANN meeting.
 I will be the speaker using less time in a presentation.
 Start your watches now.
 So I don't want to explain experiences, but there are some lessons in what we have seen in the last years of launch the TLDs and some proposals on that that can be similar to things.
 We should all of us encourage social responsibility in the TLDs.  But doing the best, and trying not to create problems when they can help avoiding them.
 And that will be also an appeal for cooperation.
 And, you know, that will be also an appeal for cooperation.
 What I mean by "responsibility"?  There's two things on that.  The responsibility should have an effect on reputation.
 Nacho has said that having very good policies will help you avoiding conflicts.
 I beg to disagree.  Having good policies is very important, but not all.  You have policies for speed limits, very clear, very well written, and then you have red lights.  But you need something else.  You need compliance procedures.  You need to tell people that if they pass the limits, they will have a fine that, you know, will -- with this amount that will be paid in this form, et cetera, et cetera.
 But that's still not all.
 You need enforcement.  But you need credible enforcement.
 And this is the last part that he was explaining somehow.  I mean, community TLDs especially, all the TLDs should be thinking about that.  But public community TLDs have a special duty in showing more social responsibility, not creating externalities in creating, really, pollution outside the space they are supposed to handle.  Need to really think about the role of reputation and all this.
 For instance, one of the examples that Nacho gave, dot cat had a classic (inaudible) in the middle domain name per domain name manually checking whether they comply or not.
 And there were some conflicts and some lies and some mistakes.
 Then, you know, after some time of application of taking down, you know, ten domains because they were not complying and especially making lots of noise about that, oh, oh, you know, we are doing this, just once, but we are doing this.
 People learned that some things were not welcomed within.  As a matter of fact, when they moved to -- I just asked him, they have less conflicts, and they have taken down much less domains than having -- so it's not about the policies also.
 If you establish a reputation in your area and you have some strict application of the rules, then you can (inaudible) them.  The reverse is impossible.
 Take dot EU, they engineered a mess in the beginning.  Now they're trying to cope.  But it's hopeless.
 I mean, I don't (inaudible) as always.  But we are all used to see that 80% of the names are just parking sites or for sale, something like that.  And despite all the good efforts they are doing to curb this down, the reputation is there.  So they are still attracting people trying to do very strange things.  Right?
 While other, you know, very strict and very tight initial compliance procedures and strict enforcement helps in reversing that in establishing your reputation.
 Now, the second thing regarding that, therefore, one of the lessons here is that I know that Nick, for instance, wants prevalidation only for any TLD that claims to be community.  But my message is that perhaps not always, and perhaps in other situations.  Perhaps initially.  But given the circumstances, that's not the only thing you can do.
 The next thing is about cooperation and eco thinking.
 Look, one good idea is a good idea.  700 completely different good ideas for new TLDs simultaneously might be a very bad idea.
 And I learned that myself in -- you know, sometimes -- I drafted brilliant policies for new things in registries and registrars, and nobody noticed.  Because, you know, there were 5,000 people doing different things.  And all of them were doing one thing, and I was trying to do the reverse.  So nobody noticed.  So it was a very brilliant and stupid idea, the impact.
 Here, therefore, my claim is that, first, let's try and -- again, the appeal is for the community TLDs.  Because as each one will serve a different community, they have an ability to cooperate beforehand.  Dot Berlin can come here and say, we are doing that for Berlin and cooperating with Barcelona.  Commercial TLDs probably have a different constraint that they cannot optimally discuss their plans and policies until they file the application to ICANN.  So let's just start with the communities, trying to have them cooperating and not creating a complete mess for users within public institutions, I.P. practitioners, general commercial users, whatever, in a maze of brilliant ideas that nobody understands.
 Now, as I said, the goal here is not to have uniformity or homogeneity.  It's to have coherence within diversity.
 Each city TLD or city TLD, language or cultural TLD or a branch TLD have different ideas.  But if they take similar procedures, similar approaches with different criteria, perhaps we will all understand that, we will pollute a little bit less the state of the DNS and the Internet in general.
 Example A, (inaudible) cooperating, defining policies, for example, sunrise policies, why the hell you are always taking just frameworks, and nothing else.  Frameworks are important.  First come, first served is important and very useful.  But as we see, there are other models for doing that.  We always forget geographic indications for something that I can't understand the reason.  But for regional TLDs or city TLDs, this would be no excuse for doing that.
 More than that, why not sharing compliance procedures?  That is, many registries and registrars may be small and have different skills, technical, legal, but perhaps not all of them at the same time.  And therefore some problems with them.
 We'll have thousands of registrars and hundreds of registries.  And then we have, you know, urgent phishing things.  Who goes there?  Who finds what?  Who discovers what's the procedure?
 And then you go to a registry.  And the first time you're confronted with that, what they are doing, waiting for the lawyer they consult in -- somewhere in a remote place, waking up the next morning, and then trying to discover whether the guy calling was really a bank or not, or the police or not, you know, three days have passed.
 So why don't they try to establish this service, which is not a registration service, which is not for money, just for service, one shop for all in which, you know, the third parties can go there, complain, have a very good shared policy way to validate that the claims are legitimate, and then, you know, with registrars and registries taking the name down, if necessary.
 The same probably for I.P. claims.  That's very difficult to do in open, generic, nonrestricted TLDs for communities.  It's much easier to say whether this I.P. claim has anything to do with the registration and the I.P. claim has anything to do with the community or not.  And, therefore, again, they could copyright in defining the policy and studying together with some specialist, share it among all of them that none of them individually will be able to afford, A, and even if they have the interest in spending half the time they need in preparing the application in devoting to this finally marginal, but I think very important things.
 Or, why not having shared mediation or ombudsman services that has to be for coherent communities, city TLDs, language and cultural TLDs, regional TLDs, and you have the -- but you can have the infrastructure behind saying this is the advice you should give; right?  This is how we can solve that and solve the problem.
 And while phishing and I.P. application procedures could be beyond strict, coherent communities.
 One final example.  And one idea, that perhaps we should think about it again, dot TMX, it's not a TLD.  It's a service.  It could be a TLD perhaps one day if this works and this works to the general satisfaction, it could be a TLD.
 Let's just start with a service.  What sort of service?  A service for exclusion of trademarks.
 What the hell, how can I read all the procedures for -- imagine, seven ccTLDs that probably, in the front page, they are strange languages like Catalan or German, people cannot even pronounce "intuition" or "visit."  So deep inside, probably, there's a link for an English version, but where?  Why not have a centralized service -- I'm talking about service, not product -- a service in which I can say, I don't want my trademark being used in those cities.
 Indeed, if there's a conflict with one of those mandate names or institutions that have priority, that's it.  But the trademark will not have anything to say, because that would not be anything like any legitimate claim.  But I don't want that available for the general public, just pretending that.
 So it can be done.  It can be done quite easily to share that.  And you can still (No audio).
 Community checking against that to provide automatic exclusions, to provide I.P. (inaudible).  Or the registrants (inaudible) the service, or even to generate registrations in other TLDs.
 Now, the idea here, I repeat, is (inaudible) what I want is that the community TLDs start thinking about not creating isolated, brilliant, but, you know, useless, because (inaudible) completely invisible ideas, but to have very visible (inaudible) community (inaudible) coordinating these things that are simply trying to prevent bad outcomes of their simple existence in diversity to the third parties (inaudible).
 And this is about services, not products.  If you are starting, oh, I will make lots of money, this is not the goal.  The goal is to prove your social responsibility, that we can create a new space with all type of (inaudible).  Also, domain names that seem marginal to the majority for, you know, remote cities like Berlin or cultures or regions and things like that.
 So that was my (inaudible) try to cooperate.  These are just examples.  Some of those have been used in these and many other TLDs, dot Asia is very active in working in antiphishing, you know, very valid antiphishing practices.  But, once again, we need to develop 700 of them?  300?  Who can (inaudible) that?  Not me.  (inaudible).
 So thanks a lot.  And now Nick will come here and say that all this is wrong and (inaudible) is the only thing that is important on the Internet.
 [ Applause ]
 >>ERIC BRUNNER-WILLIAMS:   It's city TLDs not ccTLDs, and what happens we should say municipal so it doesn't sound like CC.
 >>NICK WOOD:   Unpronounceable.  Trademarks are the only important thing on the Internet.
 [ Laughter ]
 My name's Nick Wood.  I return a -- I'm a registrar.  And we work only with Intellectual property owners.  We only work with pretty big brands.  I'm looking around the room here and I can see dotted around perhaps half a dozen people or so who also work with big brands.
 Would you out there like to raise your hands? .  And I'm going to pick on those who don't.
 >> (inaudible).
 >>NICK WOOD:   A big brand, I would say, is one that --
 >>AMADEU ABRIL I ABRIL:   One that can pay the trip to ICANN meetings?
 >>NICK WOOD:   Exactly, exactly.
 So we have a small number of people here who I would also like to enjoin in the comments that I'm going to make, which I think you who do not work with trademarks will probably be very happy to hear.
 The fact is that most trademark owners believe or are quite, quite sure that good rights protection mechanisms are very, very good for trademark owners, they're very good for communities, and they are very good for registry operators.  So it's win-win-win.
 And the reason it's win-win-win is that good REITs protection mechanisms -- and I'm going to come and talk about what these ones are in a minute -- because good rights protection mechanisms mean that registries are unpolluted, it means they're well run, it means they start on time, it means that they're profitable and stable and they do not fall over, it means that trademark owners do not have to be concerned about them.
 And if there's one thing that trademark owners do very well, it is to follow the rules.
 So, as we've heard earlier, if the rules are clear and the rules are concise, then trademark owners can follow those rules and they're not going to cause you any problems whatsoever.
 So a number of things have been suggested today as mechanisms for helping to protect the rights of others when new registries are launched.
 Everyone here probably is familiar with a sunrise.
 So I think there's no need for me to talk about a sunrise or another mechanism.
 >>AMADEU ABRIL I ABRIL:   With this.
 >>NICK WOOD:   But what I would like to do is to talk about some of the other ingredients that go into sunrises and then to touch on some of the things that were raised earlier.
 Now, I haven't got any slides, because this is a fairly ad hoc session.  And also, dotted amongst the people that -- the people who work with trademark owners here is someone who I can tell you has done a lot of work on a dot TMX, and someone who can also stand up in a minute and tell that you it works because he's used one.  So that's a surprise.
 So if I were in your position and I were to be running, let us say, a dot London, then what would I put into the mix?  What would I want to put into it?
 I would want to make sure that the registrars who I used were perhaps accredited, to use a term, but, actually, to use another term, to make sure that they were credible.  So it's not what they tell you they're going to do; it's how they have behaved in the past.  So can you trust the registrars who you're choosing to be your distribution?  Can you trust them?  Or are they going to take your very good product and are they going to dilute it and spoil it perhaps?  They're not going to tell some of the registrants about your terms and conditions.  They're going to take your special property and they're going to make it less special.
 So be careful, spend time on choosing your registrants is one thing I would do.
 Another thing I would do is I would allow some amendments to applicants.
 So if we take it for granted -- and I think it's very important that the rules are very clear, especially those on eligibility, and then people are applying.  But some people who apply, particularly in community, city gTLDs, they will be relatively unfamiliar with your rules.  They perhaps will not be applying using an attorney to advise them.  They will be doing it themselves.  They'll be sitting at home and they're going to fill in the form.  Some people are going to make a mistake on that form.  But it's a good application.
 So don't box yourself into a corner.  Give yourself some flexibility so that you can correct honest administrative mistakes.
 I would certainly have what Amadeu called an I.P. claim system, some fast challenge process.  And the fact that it was so effectively done under dot cat means that as far as trademark owners are concerned, dot cat is absolutely no problem at all.  It's really rather wonderful.  Do we agree?
 [ Applause ]
 >>NICK WOOD:   I.P. owners?  Any complaints ever about dot cat?  Never.
 The next thing, I think, which is -- which goes back to the clear rules of eligibility, it's to do with charter enforcement.  So what I would do is, I would get people to, when they sign up to my registry, then they're agreeing to a charter.  And I would go back every year.  I would send them something, they have to renew their charter credentials, their eligibility every year.
 Because then again, once again, you keep it as a good, pure space to be in.
 I think that validation of the applications in the first instance -- I'm sure that validation took you hours and hours.
 But the benefit of it is the fact that you have got a very high quality of registrants and they're loyal and it's growing.  And also, what you haven't gotten, if you think of very large examples, you haven't got an example like dot info, if you wind back to the year 2000.  Anyone here remember 13,000 challenges of last resort?  13,000.
 Anyone here remember dot biz?  And the difficulties?  Yeah, the difficulties in their launch.  I'm sorry, I didn't mean that as an insult, because I rather like dot biz.  What I meant was the fact that they had to pay 1.7 million in refunds to applicants in early days.
 So validation to check the applicants or who they say they are is really important.
 I quite like a usage requirement.  I quite like, in some instances, if there's someone who's got a trademark and they say they want to come into your space, is there a trademark, is there a trademark that you're using.  Is there someone that's rushed out and trademarked a very good term and they're not going to use it.  They want to sell it on.  Or is it someone who is using a trademark in the correct way?
 There's then the matter of WHOIS, which wasn't mentioned today but for trademark owners, one thing that really annoys them is when there's no WHOIS detail or the WHOIS details are inaccurate so if you can publish correct and up-to-date WHOIS information and ensure that registrants who do not keep their details up to date at least have their wrists slapped, then that's fantastic.
 So those are the ingredients that I would put into it.  And I think, then, the biggest two things to talk about now briefly that Amadeu raised -- and this is where we'll bring in some other people, and Bart, perhaps could you come up here.  Is it possible for him to have a seat?
 >>AMADEU ABRIL I ABRIL:   Yes.  We can leave.
 >>NICK WOOD:   We can leave.
 He only needs one seat.
 >> I'm not that big.
 >>AMADEU ABRIL I ABRIL:   I'm in a dangerous place now.
 >>NICK WOOD:   Then Amadeu threw out some things.  He threw out the suggestion that there should be coherence.
 >>AMADEU ABRIL I ABRIL:   Within diversity.
 >>NICK WOOD:   Within diversity.
 And I think that's absolutely fantastic.
 With the clients that I have, the trademark owners that we know, the people in this room, the trademark people in this room, would we like coherence with diversity?  Would we like similar types of systems across cities?  Would we if find that helpful?  Yeah, arms up, anyway?
 Okay.
 (No audio).
 >>NICK WOOD:   Oh, okay.
 >>AMADEU ABRIL I ABRIL:   Just in case.  Defensive registry.
 >>NICK WOOD:   So I think that was, in that case, with David's arm, that's everyone, I think.
 I think that the idea of some kind of shared mediation service or shared dispute service is, again, fantastic.
 In the UDRP, there are now three providers, three principal providers of UDRP services.
 The fact that there is there is a UDRP which is post launch and is mandatory upon you, it's harmonized, it works pretty well.  To flip that to prelaunch mechanism so that there is something -- there is someone you can go to with your I.P. claim or someone else who can look at things, that's a wonderful idea.
 Reasonably priced, fast, all those extra bits and pieces that trademark owners would want, but of course come in.
 But one place to go to for consistency across them would be fantastic.
 Any disagreement with that from I.P. people?  No?
 That then comes to the dot TMX.  And I go back to, I don't know, previous gTLD rounds.
 I can remember sitting in Melbourne and speaking to this list of perspective applicants for new gTLDs.  And person after person would knock on the door of this room in Melbourne, and they'd come in and they'd say -- we'd say, "What are you going to do?"  This was intellectual property.  We'd say, "What are you going to do to protect your rights in your new gTLD?"  "We've thought about it, but can you give us a list that we can block?"  Because then everything is very simple.  For a long time, that was a kind of a dream.  Well, this dot TMX isn't that simple, but it seems to me to be not a bad step towards something which would provide clarity for people running registries and that trademark owners, I think, would value greatly.
 So Bart Liben has been responsible for validating applications in the sunrise for dot EU, for dot Asia, for dot mobi, for dot ME.  He's done the numerical sunrise for dot NL.  So here's a man who has done lots and lots of these validations.
 And I know it's something that he's thought about.
 So, Bart, would you like to --
 >> Sure.  Thank you, nick.  Hello, everyone.
 So thanks for this introduction.
 Currently, it's quite clear that in the EVE of the new gTLD that's coming along -- and by the way, we've also assisted as an organization, we've assisted ICANN in drafting the criteria for those -- that rights protection mechanisms is something that will come up, yeah.  And if we're looking at potentially 500 new TLDs that will be out there in the next five years to come, if you make a quick calculation, that would mean about two sunrises a week. So I think certainly for intellectual property owners, that is something of a major concern.
 And, hence, taking into account our involvement in those past processes, we're talking about 400,000 records, approximately, that were flipped over, and said, "Well, it's okay or it's not."
 We, of course, are also looking into ways on how we can make this more effective.
 So what we've actually constructed and that could be useful for those of you who are applying for a new gTLD, being it as a trademark owner, as community-based applicants, or maybe just for open gTLDs, in general, this database is actually a whole set of records.  And we're talking about tens of thousands of I.P. rights that have been matched against domain names.
 Well, this database contains quite a generic rule set.  And for those who are interested, I'm happy to show you more or less how it works.
 This is something that we want to put at the disposal of the community, to the future registries, in order to tap into and in order to make sure that the actual applicant for a particular domain name in a new TLD is actually the right holder or not.  It looks at automating the sunrise process to a very large extent.
 Now, if you look at automation, one of the things that the pro working group in June of last year has established is that automating the sunrise application process against online trademark records was one of the main recommendation there.  Now, there's a risk involved, because if you look at, for instance, dot Asia, the sunrise process for dot Asia, generated about 30,000 applications on the basis of registered trademarks.
 And more than half of the records were wrong.
 So this is what you get when you're going to automate things.  And this is something that you must be very -- it must be very carefully taken into account.  We have the same issue in dot EU, where over 50% of all the applications that were submitted contained erroneous information.
 Now, in dot EU, we could lift the acceptance rate to 70%, so 20% of the wrong applications that went through due to a number of clerical errors that were not considered in determining them.
 In dot Asia, we got a success rate of about, I think, between 97 and 98%.  Although the record set was even worse.
 And it generated zero disputes.  So we combined the application process where we didn't consider certain mistakes in the process, and we accepted on the basis of erroneous information actually claims made by trademark owners or I.P. owners that were genuine.  We could determine that on a number of parameters.
 And the other ones that were contested, if you like, we had developed some kind of an internal dispute resolution I.P. claims service where somebody else could object against somebody who applied for a particular domain name on the basis of a trademark.
 So as I said, we generated zero disputes, so no risk for the registry at all.  We had about 120 of reconsideration processes where application data was corrected at the request of the applicant themselves.
 So, basically, if you look at the risk element involved into sunrise dispute processes, this was one of the most successful, I think, so far.
 So on the basis of all this experience, what we've actually built is this huge database containing trademark records matched against domain names, which can be tapped into, hopefully, in the future when it will be fully live and fully online, by registrars and registries, so you can automate the sunrise process and have something in basically a split second.
 So I see my time is up.
 The ones that do require or do want to have a bit further information on how to use that, how that could be tapped into a policy and so on and so on, I'll be here until Friday.  So just grab me by the neck whenever you see me passing by, and I can give you a small demonstration on how it works.
 Thank you.
 >>ERIC BRUNNER-WILLIAMS:   Thank you.
 [ Applause ]
 >>ERIC BRUNNER-WILLIAMS:   Thank you very much.
 We have "the perfect sunrise."  I'm going to hand a pile of them to Werner to go stand in the back and distribute at the back.  And I'll be up at the front.
 We are out of time.  So I don't think we can take questions, unfortunately.
 But we'll be out in the hallway for quite a while.  So we can take questions there.
 >>AMADEU ABRIL I ABRIL:   I have one thing here.  We asked for only one hour and a half, we had only 45 minutes.  The goal was, my personal goal and Eric's and many others, is that I said to encourage cooperation.
 I am interested in I.P. protection, but even more interested in having policies and mechanisms that help making TLDs a success for their intended use, especially for the community ones.  Because the community ones, I said they are more likely to cooperate for (inaudible) of the communities.
 So if you -- I think we should take up ourselves, nick, that, for instance, we'll try to coordinate all these ideas and come in Mexico well before the final thing for the guidebook or whatever name it has at that time to come with ideas for people to use.  Because I repeat, those that had experience before now have the duty to help the people that have to right applications to have easy access to a choice of good policies and good decisions instead of trying to reinvent the wheel or even worse, choosing the wrong policies.
 And that was the intent.  Okay.  Thanks.
 [ Applause ]
 >>ERIC BRUNNER-WILLIAMS:   Thank you very much, everyone.
 [ Applause ]