World Wide Alliance of Top Level Domain-names

ccTLD-ICANN Meeting in Geneva, Switzerland, 19 Feb 2001

ccTLD-ICANN Meeting in Geneva, Switzerland, 19 Feb 2001

CICG, Centre International de Conférences à Genève, http://www.cicg.ch.

ICANN
1) Michael Roberts, ICANN CEO/President
2) Andrew McLaughlin, ICANN Staff
3) Herbert Vitzthum, ICANN Staff
4) Amadeu Abril i Abril, ICANN Board Director
5) Jonathan Cohen, ICANN Board Director

ccTLDs
6) Vojo Spahiu, Albania
7) Stephen Deerhake, American Samoa (US Ter)
8) Marianne Wolfsgruber, Austria
9) Michael Haberler, Austria
10) Juan Carlos Namis, Belize
11) Christopher Corje, Belize
12) Nick Faries, Bermuda
13) Clifford Paravicini, Bolivia
14) Demi Getschko, Brazil
15) Filiga Michel Sawadogo, Burkina Faso
16) Victor Ciza, Burundi
17) Patricio Poblete, Chile
18) Bart Mackay, Cocos (Keeling) Islands
19) Brian Cartmell, Cocos (Keeling) Islands
20) Jim Trevino, Cocos (Keeling) Islands
21) Ricardo Pedraza, Colombia
22) Beatriz Alonso Becerra, Cuba
23) Carlos Ferro Suárez, Cuba
24) Agathoclis Stylianou, Cyprus
25) Per Koelle, Denmark
26) Marko Lahtinen, Finland
27) Elisabeth Porteneuve, France
28) Annie Renard, France
29) Jean-Christophe Vignes, France
30) Stephan Welzel, Germany
31) Sabine Dolderer, Germany
32) Carsten Schiefner, Germany
33) Anastasia Ladopulos, Greece
34) Luis Furlán, Guatemala
35) Sudhir Mudor, India
36) Gina Hosein, Ireland
37) Michael Fagan, Ireland
38) Doron Shikmoni, Israel
39) Stefano Trumpy, Italy
40) Daniele Vannozzi, Italy
41) Franco Denoth, Italy
42) Maurizio Martinelli, Italy
43) Hiro Hotta, Japan
44) Makiko Matsumaru, Japan
45) Tsugizo Kubo, Japan
46) Rasma Kaskina, Latvia
47) Daiva Tamulioniene, Lithuania
48) Haja Ramboasalama, Madagascar
49) Shariya Haniz Zulkifli, Malaysia
50) Oscar Robles, Mexico
51) Pavel Chirev, Moldova
52) Abhisak Chulya, ccTLD Secretariat
53) Fay Howard, CENTR
54) Eberhard W Lisse, Namibia
55) Bart Boswinkel, Netherlands
56) Peter Dengate Thrush, New Zaeland
57) J. William Semich, Niue
58) Alf Hansen, Norway
59) Annebeth B. Lange, Norway
60) Ghassan Qadah, Palestine
61) Sidia de Sánchez, Panamá
62) Erick Iriarte, Peru
63) Joel Disini, Philippines
64) Maciej Kozlowski, Poland
65) Artur Piechocki, Poland
66) Pedro Veiga , Portugal
67) Eugenie Staicut, Romania
68) Kira Litvina, Russia
69) Elena Gertseva, Russia
70) Abdulaziz H. Al-Zoman, Saudi Arabia
71) Maimouna Diop, Senegal
72) Muditha Gunatilake, Seychelles
73) Patrick O'Brien, Singapore
74) Ka Yew Leong, Singapore
75) Ivan Lescak, Slovakia
76) Marko Bonac, Slovenia
77) Calvin Browne, South Africa
78) Neil Doundas, South Africa
79) Pilar Luque, Spain
80) Anders Janson, Sweden
81) Sven Düring, Sweden
82) Marcel Schneider, Switzerland
83) Dr. Constantin Tönz, Switzerland
84) Ching-Yi Lui, Taiwan
85) Kanchana Kanchanasut, Thailand
86) Anthony Bishop, Tuvalu
87) Blythe Holden, Tuvalu
88) James Ross, Tuvalu
89) Naser Sulaiman, U.A.E
90) Richard Francis, United Kingdom
91) Christopher Roosa, United States
92) Peter de Blanc, Virgin Islands
93) Jacob Munodawafa, Zimbabwe
94) Matthew Chigwende, Zimbabwe

Minutes Written by Abhisak Chulya

Chaired by Elisabeth Porteneuve and Mike Roberts

Starting time: 9:15 AM

Elisabeth started off by asking all the participants to introduce themselves. She then proceeded to her presentation on - ccTLD the history in the making. Her power point viewgraph can be founded at this URL: 20010219.ccTLD-history-in-the-making-EP.html

After Elisabeth presentation, Mike Roberts and Andrew McLaughlin of ICANN presented the same slides that they did in Hawaii. Please see Minutes of ICANN-ccTLD Manager Meeting in Hawaii at this URL: 20010201.ccTLD-ICANN-minutes.html

During Andrew presentation, the audience asked him to clarify the issue of property right in his last slide.

Andrew McLaughlin: Do you know the difference between property rights and intellectual property rights? The property rights mean you buy something such as computer then you get the property right on that computer. You can sell it. Nobody can take it away from you without due process of law. The intellectual property rights is a set of statues or laws, not fundamental legal principles but defines statues that say there are rights to data and database, there are rights to copy rights at work, trademark names and so forth. This intellectual property protection may well apply to ccTLD 's operation database that you run. The basic principle that domain name is not property should be preserved to this agreement.

Question: Do we need to respect the local laws?

Andrew McLaughlin: Well, sure we obviously have to respect the local laws which often look to the structure this agreement to determine whether the domain name is the property or not. If the domain name is the property, you can not just cancel it when the person refuses to pay. Instead, it is a set of right created by contract. The contract says you lose the domain name if you do not pay, for example. You have to provide valid name server information on ccTLDs. Those kind of requirements could not apply if the domain name is a piece of the property that once purchase is permanent and forever. So that is the distinction I am trying to get at. It seems to us that this agreement is to embody that notion.

Coffee break:

Eberhard Lisse of .na: I used to be a non-profit, noncommercial local registry only. Now ICANN is actually forcing me through funding to become the commercial registry I personally think that we should amend the situation a little bit, for example, the registry who want to pay domain name tax ,for example $1 per domain, can do so. This would allow small commercial registry to grow into a position to be able to pay this $10,000. US$10,000 translates into 800 domains. We have about 400 domains inside at the moment and 200 domains from the outside. In spite of being the commercial registry, it's simply not possible and I suggest that we find the way of being fair on both sides. I am quite willing to add this tax to the customers. So it is simply say that we have to pay US$1 per one domain name to ICANN so please put this you're your account. When that reaches the level, I am quite willing to pay more than US$1.

Mike Roberts: I just like to point out that Class 3 structure which currently goes to $500 number happens at approximately 1000 names at the moment. The question of where the registries are sufficiently small size that the annual contribution ought to be a very nominal amount is the one we look to your people to give advice on. We had some inputs that $1000 was excessive for the floor. That is why we set at $500. But certainly there is expectation that small registries would not be required to make a contribution in their term was exorbitant.

Andrew McLaughlin: The set of slides that you saw from Mike on financing is our latest effort to synthesize the different positions and to combine them into some kind of workable model. That either means that our model has the virtue of pulling together the best part of different models or has suffered from the sin of having all the worst features of different models. It got the fixed component and variable component. And it tries to do the varies of thing that we have been told we need to do such as don't be a burden to smaller registries, don't force non-profits to become for-profits, respect the diversity of different ccTLDs around the world. As Mike said earlier if anybody can actually come up with better alternative, we are all ears. The goals is to come up with something that meet our needs for stable funding, well at the same time meeting the appearance test of being fare to this vast diversity of different kind of ccTLDs.

Agathoclis Stylianou of .cy: Why we need to pay this amount? My community would like to have some clear answer. What other services are we getting?

Mike Roberts: I would like to defer this question to this afternoon. We certainly will get into this.

Stefano Trumpy of .it: I learnt with pressure that now the issue of election with country code is one of the priorities of the board because this is something that was highly recommended. And we know that you had a lot of pressure for the at-large membership in Yokohama and then for the new gTLDs and so on. We also know that you have pressure for establishing the contract with new gTLDs registry. My fear is that if this attention is enough and I hope it will be enough and in Melbourne we will reach substantial improvements. The key point is a number of chicken and eggs problems for ICANN. For example, this relation with Department of Commerce (DoC), the possibility that we achieve the goals of ICANN in due time before September or October. It's the end of extension of the agreement with DoC. It would be hard to have perhaps 245 contracts in place by that time. And this implies maybe the goals of ICANN will not be completed and then will be another extension of agreement. The chicken and egg problem is here because the National governments outside of US want to be the political play if the tie with US government is not finished. The process in establishing this contract will need more time. Then today we learn that for the first time a clear position not yet well defined of the ICANN position concerning the contract, because we have all these Best Practices documents from CENTR and ccTLDs. We have GAC principles but up to now we didn't have the real Best Practices from ICANN (bilateral or trilateral). In many countries, you got complaint from the local internet community and the government, they expect ICANN saying something clear which is the recommended procedure to approve a contract or to start the redelegation. We have been discussing at length the possibility of having disputes resolution for the redelegation. All this matters imply that ICANN perhaps should delicate more staff and resource to deal with country code. This is another chicken and egg problem. If you don't have contract, then ICANN fails to get money from ccTLDs in the measure that you expect and then you don't have enough resources to speed the resolution of the problems. This is something that I'd like to see the progress in the short time and I wonder if present timing is too long. Until this kind of problem is not clear up with firm position, we are still navigating with no precise direction. This is very general on the discussion for today.

Mike Roberts: There has been an expectation for some months that we would do an updated version of draft contract we first posted last July based on the work you folks have done over the summer. The decision was we should have these three meetings (Honolulu, Geneva and Melbourne) as efforts to find common ground on this before we start passing legal paper around. As you can see from slides, they are not very far away from legal form. So perhaps we can use some time this afternoon to see how far apart or how close we are on the principle.

Stefano Trumpy: In notation to the next step, the famous letter from ICANN to the government. If there is no answer after two months, then this government agrees on the present situation. If this is prerequisite, why wait to establish the contact with governments if the decision is to go ahead with the GAC principle. Just a question of timing and having a firm commitment from ICANN. I think this is a contentious issue and many registries don't like this at all ICANN send a letter to governments. ICANN needs here a fair position how to proceed about that.

Mike Roberts: The fact of the matter is we now have 6-8 countries that are actively talking to us about the trilateral agreement. I felt that working our way through the trilateral agreements will help us and help those countries and the registries involved and others understand the extent to which we will work off common terms versus tailor terms. We certainly have no desire to have 250 different versions of a trilateral agreement. That just doesn't make any sense at all. On the other hand, we can't go all the way the other way to say there is only one piece of paper and has six terms on it and has no deviation from the words about those six terms. What we are searching for hear is satisfactory middle ground that allow us to proceed together. I think that on the specific issue of the choice of agreement form, our only requirement is that if the registry wishes to execute the 2 way legacy agreement, it is required that the local Internet community of that registry including the Government be aware that intention. We are not specifying what awareness means but I suggested in Honolulu that the appropriate way for that awareness to be documented prior to signing of a legacy agreement would be that you would post the agreement on your web site. This is the typical ICANN process the agreement are posted for comment and it seems it's reasonable thing to do.

Sabine Dolderer of .de: For Germany, it's a bit different what we want to get from ICANN. I look on ccTLD Manager-ICANN responsibility that we have to agree on. Just like commitment to operate ccTLDs for full interest of local internet community. That's fine with me. Acknowledgement that no property right are acquired for ccTLDs is also fine for me. Compliance with ICANN developed policy concerned domain names resolution, I don't think this is appropriate for ccTLDs. We have long time experience with domain names disputes and we have domain names resolution solved perfectly for our local situation. I do not think that I need global domain name dispute resolution dictated by ICANN. Of course, we can position our domain names registrants. Apparently they have a very strong position because they are residing in Germany and governed by Germany Laws going to local court. That is a very strong position for domain name holder. I don't think domain name holder wants to be weakened by domain name dispute resolution policy which I have to implement because of ICANN domain name dispute resolution policy. I don 't think ICANN should regulate things in Germany. That is up to the two parties, denic and German government. The other thing is participate and development such as protocol standard. I think protocol standard should come from bottom up. There is another body that does the other way around and that is ITU. They are sitting together every four years developing new standards and tell you to follow their standard. That is not the approach we want in the internet. So standardization should also come from bottom up. Let people or users decide what will be the standard, not because some organization wants it to become standard.

Andrew McLaughlin: Let me answer the first question, we are not proposing ICANN dispute resolution policy globally. Let me say it here for the twentieth time. ccTLDs are responsible to set up their own dispute resolution policy consistent with local laws, local practices, local customs and local needs. I don't know how I can say that more clearly. The fact of having a dispute resolution mechanism could be a basic element of the contract. You get to decide what it is but maybe what we have to say that all ccTLDs have to have some kind of mechanism for dispute resolution even if that mechanism is as simple as you have to go to court and we then agree to abide by court ruling at a registry level. Don't confuse by that bullet point. It's not written very clearly because the parenthesis at the end to the extent applicable is intended to capture basic notion that we don't tell you how to do dispute resolution policy. As you know, the three gTLDs that we are more directly responsible for .com, .net and .org, we have a UDRP that we take responsibility for ourselves because those gTLDs are not responsible for any particular local community. But other than that, the administrator is responsible for dispute policy.

Your second point, on mechanism for conversion on trilateral situation, the point is the agreement that we sign has to allow you and government to change it if both of you reach agreement some where down the road if you want to do that. It's very simple. The responsibility to decide whether you want to do that rest with you. It has to be upgradeable by all the parties concerned. We are not proposing to have a unilateral upgrade where we get to decide whether or not to deal government in. That is not what we are proposing to do. It's two-way contract both parties to the contract have to agree. We are talking about those that contract can't be ironclad because that creates all kind of suspicion by people that are not ICANN or ccTLD manager. The community as a whole has to understand that if local community develop some kind of better structure within the country to handle the public protection function, that this contract can be converted into that kind of agreement. If you have an agreement which is basically indefinite, there has to be some kind of mechanism where it can be converted. This is not intended to say that it's going to be converted without your consent.

Your third point, regarding standard. ICANN is not a standard body. It can not set standard. IETF sets standard. World wide web consortium sets standards. The bullet point you are referring to has to do with policy, not standards. Another words, whether or not Whois data is provided, whether ccTLD registries escrow their data, that's sort of thing. Those of things we don't have any policy for right now and I know some of ccTLDs have work on Best Practices document like CENTR that covers some of the topics. I think one of questions for ccTLDs as a whole is what things all ccTLDs expected to do. Maybe the answer is not much but maybe there are things that ccTLDs want to encourage. I don't know where you are getting standard out of that but it's not intended to be a standard point. Standard is developed and then they will be implemented or not.

Sabine Dolderer: I want to discuss firstly, the technical solution and secondly, technical mass to offer it to the community. There are 2 steps. First of all is to use a common registry-registrar protocol in your registry as a communication body to your registrar. That is the standard possibility to use it as mass like you have to participate in a universal Whois however it' s implemented. That's the policy decision. I think there is a border between what is part of ICANN and part of IETF. The development of a protocol, it is a clue or it is hierarchy, that is up to IETF. It's up to the ICANN to decide. That's the standard that the registry has to offer this, this, this. That's the policy decision. I think a line where ICANN wants to define things. Well, we think it is up to us if we want to offer some services even if there is a technical standard for that service.

Bill Semich: I just want to add a comment or observation and maybe get some response to it. Since we start this process in Barcelona with IFWP, and then in Monterey and finally in LA with final creation of DNSO. It's really been an issue of definition that has not been clarified in terms of relation between what's going to be ICANN and what is going to be ccTLD community in which didn't really exist prior to the creation of ICANN as a community. The definition is who's calling the shot, who is making the decision? Is there a body of ICANN which is a centralizing regulatory authority, which ccTLDs are negotiating with and who has license to the ccTLD top level domain management authority. Always there is a group of ccTLDs who would consider proposing a contract with ICANN providing the services to them for their going businesses, their going concerns which have been in operation in some cases over a decade. I think it's figuring out what the basis of our relationship is. Is it causing any other problems in term of agreement, negotiation, payment or whatever? Maybe others may not agree. That's how I perceive.

Mike Roberts: We ought to have some more discussion about that in the afternoon but basically Andrew & I and other directors are largely messengers of the privatization decision of the US Government. Our bylaws and articles in corporation are all constructed on the framework of the white paper. Now it obviously room for people having quarrel with the white paper but that's not quarrel we are a part of. We are following fairly literally the privatization actions of the government. Let's think outside the box, it's fine. But you know as an officer of the corporation, I do not have as the director the ability to do something different. That is really not on the agenda for discussion.

Andrew McLaughlin: Let me just tackle that from some different angle. On one hand, you have a centralized DNS that administered by somebody they do allow a little piece of it to people to run on behalf of the overall DNS. That is Model 1. Model 2, you have independent ccCTLD registries that magically sprout up somehow from the ground and then they collectively designate central coordinator authority. Let me suggest both of those models are wrong. Neither one of them makes any sense any more. The model that I prefer is one where DNS is a public service, public function, something that were all doing with all the benefit of the internet community. The right way to approach is not try to say who is in charge, is not try to say who is the king of the system. The right thing to do is to say let's build a system & network that serve public interest. That means it accommodates the various kinds of registry who sit in this room for profit and non-profit, government, university, in coordinate them to provide services to global community. That is the right way to think about it. I tend to think of this because I work for ICANN as a non-profit. Some of you think about it as a bottom line thing because you run businesses. It's perfectly reasonable. So my sense is that this is neither a kind of feudal kingdom where little barons are recognized by the sovereign nor is it some kind of a radical Marxist state where registry springs up from nowhere and recognizes the central coordinator. It is something quite different. It is the public service function that we all try to perform in a different way on behalf of the global community.

Pilar Luque: I have 2 questions, which are related. One is how does ICANN propose terms and each registry class within the fixed component and how does ICANN propose to obtain accurate registration figures, by which calculates variable component. I do not know if this should defer to this afternoon.

Andrew McLaughlin: We don't have answer to these questions. Our model is synthesis to a lot of points that we have. We are not proposing to classify everybody. What we try to do is to describe what system might look like. I don't have any particularly powerful persuasive way to classify. And by the same token at the moment we do not have access to everybody zone file. We don't have any magic answer to that. We are hoping to talk about it today as whether at a broad level this makes sense. One of the questions that you have to ask - Is it practical? Can it be implemented? I am not sure that we know the answer to that.

Mike Roberts: I got a question in Honolulu about how would we know that the people are cheating or not. It seems to me that among other things the largest number of the registrations where that would make a significant economic difference are relatively small number of registry perhaps fifty out of two hundred fifty. In general if you are committed to a formula based on the registry sides that your own best judges of whether anybody among yours is cheating.

Bart Boswinkel of Netherlands: I got 2 questions. One is what is the relationship between agreement we suppose to sign up in the future and RFC1591. Is it going to be successor or is it just a clarification of RFC1591? Second question is I have some problems with the concept of trust. I know a concept of trust is a common law notion. I wonder if you use this notion in your presentation, is that the concept you are referring to or is it just something someway in the blind? That 's all.

Andrew McLaughlin: Let's we take your second question first. Trust in the common language sense of the word, not common law sense of the word, when I talk about trust, it is a function that you perform for the benefit of the community, not any legal sense. As your first question about RFC1591, we are proposing to replace RFC1591 with formal agreements. 1591, as was later clarify as what we call ICP1, simply describes what the IANA does with regard to delegation/redelegation and sets expectation on ccTLD managers. If you read it, nothing more than the description of policy. So what we are trying to do is make more transparent, more formal, more concrete, more reliable, more stable. That 's set of the relationship. As I said, in the legacy contract that we contemplate, we intend to capture the basic element of 1591. We are not proposing to change the structure or the substance of the relationship but rather to spell it out. Right now, it is not very clear to the people in this room how the IANA handles the redelegation. One of the things we propose to do through this agreement is to make that process significantly more predictable and transparent. I know the way to think about it is - what steps should we take if we get the redelegation request? It would be nice from our prospective to have a checklist where we go down, we know exactly what we are going to do and when and so forth. Right now we follow 1591 as we have clarified.

Andrew McLaughlin: There is a distinction that you have to draw between consensus policy development and the drafting of organizational agreement. The agreements (that we are trying to sign with the gTLD registries, with the ccTLD registries, with the RIR, with the root server operators) are mandated in the white paper, call for it as part of MOU with US government. That's not the same as consensus policy development. The agreements themselves are supposed to be open to future consensus policy development, one way or another. The NSI contract, the RIR agreements, they are subject to the public comment, they have to be posted but they do not go through a full DNSO process. These are the basic institutional foundations of ICANN and then the policy process has to be built on top of that.

Peter de Blanc's Presentation: Details can be found in his briefing paper at 20010219.ccTLD-Briefing-document-PdB.html

Close for lunch at 12:45 PM.

Afternoon session started at 14:00 PM

This session will be mainly Q&A.

Paul Kane: Thanks for ICANN's board members and staff coming to speak with ccTLDs. We heard this morning about thin and thick model relationship with ICANN. Would ICANN staff or board members like to elaborate? What would ICANN board or staff like? Would you like a thick or thin model as a relation with ccTLDs?

Andrew McLaughlin: Peter (de Blanc) and I just conceptualize about this problem differently. I think there is a lot of merit in the thing that Peter puts forward. I think about it somewhat differently which just to say our objective ought to be made this problem as easiest as possible and to get done the things that we know we have to get done first and so to my thinking the first thing that has to get done is the formal agreement. The broad outline of that agreement seems to command a lot of consensus that it should be lightweight. It should locate primary policy authority in the ccTLD and the local community. It should include some general stability provisions for the overall Internet and it should include funding component for ICANN, that is basic element. Whether you call that thick or thin, I am not really sure we certainly get into details of it, but I do believe it ought to be very lightweight. Now as for may other things that Peter references, it seems to me that they could arrive from that relationship. Clearly ccTLDs need to have a secure in ICANN budget process, budget revenue and budget setting. I like to see a kind of advisory organization of some kind rather to help coordinate in peer-to-peer relationships among ccTLDs and ICANN. That is not the exactly a policy setting body but rather a kind of communication body so that we have a place we can all talk to each other. Again, I don't think that is particularly thick but I am not sure whether that falls within Peter's spectrum. I believe in basically this sort of basic elements of a relationship and an agreement which might then revolve into other things in the future if ccTLDs think it's worthwhile to them to do that. But what's on the table from ICANN staff perspective anyway is a lightweight step by step approach to get through the things that we have to get through first.

Jonathan Cohen: Personally I have a lot more experience about thick and thin. Speaking for myself, it's quite clear that relationships with ccTLDs are essential to continue success and smooth operation of ICANN and Internet. And I think achieving a working relationship that is satisfactory to both is what really our objective whether you call it thick and thin I don't know. I think when I look at Peter's summary, I tend to favor a more robust relationship but not necessary exactly as he put.

Peter de Blanc: For the first part of my presentation, I simply say that there is a dividing line between thick and thin and it occurs as soon as the issue of the policy consensus formation, politics and so forth enters into it. As long as you have something like the standard IETF, Protocol and these engineering situation. As soon as the political side, the consensus formation, the meeting, the back and forth draft, the publishing on web site and so on. I call that "thick". There is obviously degree of thickness. That's dividing line that I am using. And our model right now in terms of the way I look at it, we are already in the thick area because we have 4 meetings a year in different parts of the world. We have a big cost. We have a bureaucracy. So that's we are at.

Amadeu Abril i Abril: The reality has put ICANN in thick model, ccTLD interests are probably on the thick model despite what you made. Quite different question is the statistical implication of this division perhaps really more helpful to see how we move for world. We try to find each one's places from that starting point than concentrating on conceptual model. In reality interests are here and reality more or less the same interests on perhaps different from ccTLD to ccTLD and from ICANN to ICANN. I prefer to divide the way to reach an agreement here. My own problem is that after so many years, I still don't have a way in my mind to think about geopolitical ccTLD. By the way, the real question is that we should not try to treat your political TLD as one group and all try to divide single mechanism or if we need to that, then we need to go to what is relevant from this TLD perspective regarding a relationship with ICANN. The less I would not be interest at the moment.

Eberhard Lisse of .na: I like to find out what you consider broken that needs to be fixed in the delegation. As far as I am concerned, RFC1591 works perfectly well, has worked for 10 years, you are using it at the moment. I am not saying we should not go to ICP1. But why do we need the government involvement in the first place?

Mike Roberts: The first answer to that question is the US government look at the management and the system that involve its research agencies and decided to scaling up the internet and commercial activity in particular required a new approach and then spent a year inquiring very widely on more than just US National basis what that approach ought to be and the answer was the white paper and the creation of ICANN. As my slides point it out with regard the legacy agreement, the point of departure for that is 1591 and the only additions to 1591 in the legacy agreement are the ones that are necessary to have written legal document and a few other things are consistent with white paper principles. From the standpoint of the fundamental relationship which exists in Jon Postel era, the bilateral version of that, there is no particular change on that. The interest of government in this and the reason for the trilateral options is that in many countries the governments either have or are in process of involving themselves into this area. So we have to respond to that interest.

Andrew McLaughlin: Let me add one more thought to that. I can not be neutral as to the government involvement. We should neither seek it out, nor on the other hand should we fight it where governments and ccTLD managers choose to come up with some relationship that accommodates for both of them. Many governments are going to get interest they see as a national resource that merits their involvements to protect the public interest. From our perspective, we should be neutral. We should have an optional agreement which calls the legacy agreement which facilitates no government involvement. And we should also facilitate government to involvement where the governments and ccTLD manager come up with some kind of a mutually agreeable relationship. That is just reality.

Stefano Trumpy: Can you tell something more from your perspective since ICANN is there? There has been an increase in the ccTLD of government involvement. There is a tendency and also more specific question this morning you mentioned that you are already in negotiation with small number of registries in trilateral negotiation. Can you tell something more about that? I suppose you are talking of at least Australia and Canada.

Mike Roberts: A number of countries are recently in the process of restructure that need redelegation: Australia, Japan and Canada all are in that group and there are others coming along.

Stefano Trumpy: So this group of registries that you are already in trilateral negotiation.

Mike Roberts: We are developing one. That is more or less in line with the content of the slide we put up this morning. If you read the principle document, it has language which obliges the government to act responsively. ICANN does not intend to execute a trilateral arrangement that does not include a written statement by the government that it will observe the responsible behavior language contemplated by the government. Now people said to me how can you trust the government to do that. I think trust is earned from all these things but the general intend or general process thus far in our particular discussions I have with people at the national level and I now listed about half of a dozen countries on this, and Andrew just finished the tour listed with three or four others, is that the government people we are talking with are proceeding in a constructive and progressive manner.

Paul Kane: I would agree with you, Mike, 100 percent that take a while to build up trust. But most of ccTLD managers have already established that trust with the local internet community and possibly with the government relations. My original question was: was it to be lightweight, or thick or thin model and I am very concerned that we try to muddy water. If it is thin, another words, the local internet community, brackets the government, are happy with the status quo. Then, the 1591 is in place and is running monthly, the agreement that just required it could be relatively lightweight. If one is going for a three prong approach, does one not run the risk of bringing governments in to really get involved in policy issues. Or more importantly, bring ICANN in to policy issues where possibly ICANN through transform common consensus position could alienate some government strikes ccTLD managers if their policy would disagree with ICANN position. I just try to understand way of going.

Mike Roberts: The decision as I said and we made clear. We don't believe that we currently need to get into this business about the letter to government that was discussed last summer because we find it is more constructive for everyone involved to sit down with the parties and discuss the legacy agreement which is acceptable to the local internet community and the government or to discuss the trilateral agreement when the government does not wish to be a legacy agreement concluded and thus far from my personal experience the governments are not interested in legacy agreements. I am not saying there are not going to be any because there are not going to be governments in that situation. I am just saying the ones that I am personally acquainted with the countries I just named off they are not.

Andrew McLaughlin: In one key respect the trilateral agreement from our perspective is even more lightweight than the legacy agreement because in trilateral situation we are recognizing the redelegation structure inside the country and it does not have to be set forth in our agreement. So from our perspective the trilateral agreements are even less involvement from ICANN.

Jonathan Cohen: Follow up on Paul's question and this gentlemen's question which stated that there was an upturning in government interest in ccTLD since the creation of ICANN, as the internet becomes more successful and more integral part of economy it would naive if anyone in this room to think the governments are not going to get increasingly involved in one way or another, how they related to particular ccTLD and what they do depend on the country and relationship between administrator and country and we don't want to be any part of it except to insure that whatever we required to be done to the stability of the internet be done. Therefore, to ignore that reality is just not being in the real world and I think that is why we are focusing increasing on issue of trilateral. And Andrew is absolutely right. Trilateral actually keeps us out of it and that is what ICANN wants. ICANN doesn't want to be dragged into politics of country code and administrators and their respected governments.

Michael Haberler of .at: our registry is running self-governance and currently the government is hand off and that is not necessary to remain this way. I could not visualize to get trilateral agreement. The issue that interests me is next step which is the conflict after that. Let assume we have a trilateral agreement and finally the government is more directly involved in policy which would result in change of the contract. What they are going to do if the local internet community and government are at each other? So far what I read between the lines from you is there is a slight tendency to give the government prerogative over the name space.

Andrew McLaughlin: The best way to think about the right answer to that question is my view is to think about what the situation is now and how it would be different if there is a trilateral agreement. Under the current situation, the IANA through ICANN is responsible for resolving their disputes. The current policy said the IANA asks the parties to work among themselves and take no action until they do work it out among themselves. In an extreme case, the IANA reserves the right to go in and do a redelegation if there is no resolution can be achieved. That has happened at a couple of points in the history of IANA. But the preference is always been to let the community work it out and that generally means with the agreement with the current ccTLD administrators if the government has a problem with the ccTLD administrator, they have two options: first, go to court or pass the legislation or follow the normal due process of law within the country. Second, go to the IANA because IANA doesn't like to deal with this dispute. It always has the policy that if you come to us, we will tell you to go back among yourselves and that serves pretty well. CcTLD news memo no. 1 included an updated to 1591 which said that IANA takes the views of government very seriously. It doesn't say it will automatically do what the government said but it recognizes governments are charged with protecting the public interests of their citizens. That is current situation. How is that different if you have trilateral agreement? If you have trilateral agreement, then whatever mechanism is defined in that agreement that is how the dispute is resolved. Presumably when we recognize that mechanism, it will be pretty comprehensive. We will shift to this oversight function from IANA to local community, there has to be a well-defined agreement on how the dispute will be resolved. Let me point to you if you want to look at the very recent example of Canadian, you will see details mechanism how CIRA and Canadian government propose to handle the disputes. Let me make one really important point from all of these, from my standpoint as a sometime lawyer and advisor to ICANN, I have an interest in protecting ICANN from requests from government officials that in essence represent innovation due process of laws within your country. We don't want to be in a position of getting a redelegation request from a deputy assistant under Minister for photo copying who said on behave of government I instruct you to redelegate. That is exactly the kind of thing we don't want to have to resolve. For the same reason we don't want to resolve what is and is not the country so we looked into ISO 3166 list. We much prefer to have due process of law within the country to solve the disputes. That means if you can get an agreement between ccTLD manager and government that you, the manager, are satisfied with and that is obviously preferable situation for us because we can then refer to the terms of that agreement and government can enforce it according to whatever the laws of your country allow the government to do.

Michael Haberler of .at: What is the case when you get the request from government to redelegate because some random body seems to entitle to do that. Or another word, what is the decision you are waiting for from a local conflict: is it a new law or is it a local internet community that government is foolish about it?

Mike Roberts: Presumably you have a trilateral agreement, there will language in document between the government and registry that obliges the government to act in a manner that consistent with the public trust for the operation of national registry. If in that situation ICANN got what seems to be clearly an inappropriate redelegation request from government, ICANN initial response would be to write back to the government saying you have sent us a redelegation request that on the face of it appears to be inconsistent with the written commitment you made under the trilateral agreement. Please advise.

Andrew McLaughlin: You should go and read CIRA document. In CIRA situation, if the government wants to do redelegation, they have to launch into a process of mediation and then arbitration and then they can actually go to court. But there is a face process to resolve the disputes including redelegation disputes between Canadian government and CIRA. And Canadian government was happy to do that because ultimately they have the legal authority and CIRA was happy to do that because they trust the government to abide by that agreement. So it worked out well for all parties concerns. But that is one model for how this might be resolved.

Muditha Gunatilake of Seychelles: Things are working well now but we have recently some ISP who has a good contact with government and these people will eventually get government preference when they apply for things. I am sure if we start agreement with government right now, the government would say NO and it would say that government should administrate. They don't know what DNSO means. Actually I help them draft the license and everything. We now will go with agreement and the other guy tells the government that it should not. We have only two ISPs so where will we stand in this situation. We have been there for years and so far everything is good. But we might get into problem if we ask government to sign the agreement.

Mike Roberts: If I understand your question correctly, you are describing situation that local situation is not ready for instant execution of the agreement. As I said as well as Andrew and board member, it is absolutely necessary that ICANN put this agreement together sensitive to local condition and we will.

Muditha Gunatilake: If we start signing agreement, government starts regulating things to great extents. Things are working fine. We may end up with more problems to the system that is already working fine.

Mike Roberts: We need to avoid getting into discussion about hypotheticals. We can't give answers to hypothetical questions. If you think you have special situation that does't fit under either of the models that we outline here, tell us that and we will come and visit you and see how we can work it out.

Andrew McLaughlin: Let me actually do hypothetical. Let suppose you got a government somewhere that is actively opposed to current ccTLD manager and wants to get involved somehow. There is nothing ICANN can do to magically solve that problem. The question is will ICANN go about this contract process in a way to create problems with government that would not otherwise exist. It's a perfectly good question. I think some objections that we had to the idea of sending a letter from ICANN to government asking for comment before we sign contracts. I think it persuasively argued that it causes more problems than it solves. But nevertheless, it is highly unlikely for ICANN to sign agreement in a case where government is out there actively opposing and seeking redelegation. It's very difficult to do in that situation.

Jonathan Cohen: What can ccTLD community do itself if you organize properly to protect yourself internationally. Maybe there is a solution if you organize properly.

Patricio Poblete of Chile: Why we need government to get involved? I believe we don 't need government to get involved. Unfortunately, they want to become involved and that is reality. We miss an opportunity of having develop a different model which didn't happen which is sometimes referred to FIFA model for Soccer Federation where the whole structure is based on geographic divisions or countries completely outside the scope of governmental authority. If the President of Chile doesn't like whoever the current president of Chile Soccer Federation, he could write a letter to FIFA headquarter in Switzerland asking them to be changed. That letter would be ignore because it's not the government's business to appoint the president of Chile Soccer Federation. That doesn't happen for us. Even though I would sympathize with some of you, people and Eberhard arguing that they have a private contract with the state of the late Jon Postel. And that's all there is. It's not the government business to interfere with that contract. I believe that doesn't match reality these days. So if there is some level of government intervention, what form will it take? I believe there will be different models, trilateral model which is used by Canada. In Chile we are trying to use slightly different approach where we, ccTLD administrator, on one side are talking to government and writing them a letter telling them we are about to begin a negotiation with ICANN and got back a acknowledgement letter from government. We then are submitting this letter to ICANN so it's not really trilateral. We are here in the middle, one side talking to government and the other side talking to ICANN. We hope that will work. We think it's a better model for many countries. In case the government really wants to get involved and talks directly to ICANN, I don't think that can be prevented. Most we can do is trying to guarantee to be done in civilized way which is of course to be true for most countries but not all of them. I must confess they don't really understand too well how this trilateral model works in the sense that the GAC principles included much more clearly than from what I hear from Andrew or Mike. From GAC point of view, ICANN supposes to do what the government tells ICANN to do, nothing else. No consideration of what has been done through a proper process or mediation, whatever. When questions about that in one of the meetings, Paul Towmey even said that there was nothing at the end that could be done against government's request and perhaps in some case the best thing that ICANN could do was to redelegate as soon as possible to save the life of current ccTLD administrators. Men with guns would do whatever they want and that is the reality. So if ICANN is going to be working in such an environment, we have to think very well about what we prepare to do as ICANN to deny the request the dictatorship in the countries and requesting the new delegate to be appointed or we wil deny or we will impose sanction on such government. What power we will have or we will not prepare to go that way, then perhaps we should say that we will do what they request to do. I think those issues we have to deal with them if we want to be completely sincere for the scheme we try to build here.

Mike Roberts: I think your comment confused some of the history of this. You have to distinguish between what is the intend of GAC principles and what happen in the pathological case. In the sequence of exchange you are dealing with, Paul Towmey was clear that the intended relationship and the language in the document burden all three parties to the trilateral agreement. The words are there. But what happened at least in the version of that exchange that I happened to hear was the individual in the audience kept pressing Paul "what happens, what happens, what happens in the worst case". And he finally said in reality the governments through the exercise of their police powers do what they want to do but that total unrealistic approach to what we are about. We are here almost entirely as representatives of democrat free government. That doesn't mean they always act in a lightning fashion but it means they mostly act in lightning fashion. And they have mechanisms for citizen representation and free expression and we have to take that for granted. We are citizen of our respective nations and we operate in that environment and I don't think it's productive to keep going back to "what if" and "what if" scenario because whatever that is it will play out regardless of what agreement is in place regarding this little piece of top level domain names business.

Amadeu Abril i Abril: When it's war, that's war. I do have myself problem with trilateral thing especially because of the language. I think ICANN knows what the domain is and know behind the domain there is an administrator. The relation of this administrator with governments from relation of internet community are different things. But ICANN has relation with NICs - domain name administrator, not the government because we are not an inter-governmental body. We do have GAC as an advisory body. Another things is how to take into account the role of governments. I agree with trilateral model even if it 's not really trilateral but it's trilateral game by the way. From ICANN point of view, it's only two players somehow. The other question is what you do in this case, why not say we would do whatever government wants. That is a solution but I am not sure it's a famous solution. Governments from time to time have a strange idea especially government officials. There are a lot of them. All of them pay to have idea and some of them react to those ideas. The best thing ICANN can provide to salute the internet is this cumbersome, expensive, tiring exercise, mailing list, going around the world forum, educating people including governments. The best thing we can tell them is come to the GAC, see how many people have the desire idea a year ago and we will tell you it doesn't work. Say why you should not take your ccTLD and give that to the chief of police because it won't work. If you really want to redelegate the ccTLD administrator to wherever you want, just have some process, try to explain it, try to justify to yourself, your local community, and perhaps all of ccTLDs, all of governments, ICANN will be at that point to explain you. But indeed at the very end, if there is a war, there will a war and I will be far from that war as much as possible.

Juan Carlos Namis of Belize: I feel some time I missed the bus maybe because this is one of my first meeting. Definitely, there is a big difference between ccTLD of developed countries and ccTLD of developing countries and I hope ICANN realized that. Once you come to ccTLD and domain names registration, if you are doing it for public service and internet community and that is fine like most university was doing it before. Maybe they are not charging for it. But the time you start charging for it, it becomes business. It's not a public service anymore. If you have private registry, domain names registration in reality is money-making machine for some ccTLDs. And definitely government wants to be involved. Any thing that will generate a huge amount of money, I don't think any governments in developing countries would not want to be involved in that particular process. Maybe the way the developed countries operate in which you can take your government to court, it's a different reality from developing countries. In our reality is that when governments most of the time want to do something, they get in done regardless of what the reasons you want to say. I hope when ICANN is talking about the revised financial strucure would never think about charging class1A $100,000 a year. That is if you want to become a registry that is going to be commercial, world wide and unrestricted. That will be biased to developing countries. If you have a small registry, you might grow over the years. But if you have to pay $100,000 a year, you might not even get to start.

Mike Roberts: If I interpret correctly, you would then favor zeroing out the fix component to reduce that burden in the absence of any activity. In another word, if you change it out and started as commercial, then we want to pay as you go arrangement either fully variable or close to fully variable. So you are saying let's simplify this thing and just do Variable. As I said and I want to keep saying that our job here is to reflect the consensus of your folks as to what is fairness way to portion the responsibility.

Eberhard W. Lisse of Namibia: Namibia is in fact a developing country but it shares with Chile one common denominator that I, as a domain administrator similar to Patricio, have been personally living in the country when there was a war. That is separate us a little bit from people who said war is war and don't worry about it. I remember having to jump into my bus when a mortar was rocked over my house in one afternoon. In Patricio country, former state President has been told it's not really a job of President to disappear people. I have a distrust of government that is very deep. My registry has been operating for 10 years as noncommercial and it has now forced in order to pay bills like this one to commercialize. So I fully support what he (Patricio) said. We should help the emerging registry by making them pay a fair share. I disgree with German registry said that they want to decide how much they pay. I am quite willing to pay a pure domain name tax one dollar or over per domain until I can afford to go the higher bracket. The bracket has too big of an increment. I put it to my customers and one dollar is not that much money. It's not against you guys. But the questions still remain "what is broken?" in my registry, for example, that operates for ten years on technically excellent basis, has capacity building program second to none in Africa, has escrow, has whatever you wanted. The facts remain that minor adjustments and all these things I want to do can be continued. What is the needs to go to my government to tell them we are signing agreement? You don't know my government. You don't know that the line of director, the ministry of trade and industry, she corrupts as you can be. What is the needs to actually change anything as long as nobody is complaining? If my government passes the law that said Dr. El you can go now, that is not much I can do. But in other countries you actually get killed when you try this thing. I don't really like it. If you come to me and sign 1591 now in the room, I will sign it now today. You have your agreement now today. Over ten years I have two complains. I have agreement with Jon Postel and it works very well. Now I consider it to be Jon Postel's estate. Why does ICANN want to come and change the rules when it's working well? There is no need to change anything.

Mike Roberts: I think it's the forth time that you beat up on that issue. Maybe we can bring that dialogue to a close. Let me stipulate. If the circumstances in your country are as you describe them, we will come and sit down with you and work out something that is suitable. It is painfully obvious and I am only the messenger that governments very large have decided they will involve. We are trying to be professional as we possibly can about working relationship that satisfy the needs of global internet and the needs of governments that state to us. I hope that we can keep discussion on that plane.

Fay Howard: I think it's very interesting to hear about these different problems in different countries. I have to say that the governments prepare to interfere in developed countries as there are in less developed countries. But I think it's encouraging some special arrangements can be made where required. But we really need to do is get on with this now. I can remember writing my first statement before ICANN can even be named. Registries want stable root services. They wanted agreements. Long time down the line I suspect some of the government intervention is used to delay. And we hear the ccTLD can't get organize, they can't get on dealing with this thing. The government has to come in and help. I think we need to get on with these now, get it sorted. If the light weight agreements are put in place for the basics ccTLD managers want, a lot of government might lose interests. If they don't, as Patricio said, at the end of the day, nothing you can do about it. People will come to you in the middle of the night and will not impress with the piece of paper you sign with ICANN anyway. If legacy is the default and there are processes in place to guard against the vicious governments win, show us some paper that might help people to be assured by this. Get documents out and let's get on with it. And then we can stop running around the world because most of registry managers will be happy that things are settle down and then you can get on the bigger issues if you required.

Jonathan Cohen: I agree completely. Let's get on with it. I don't see a problem with all of you. The issue remains is how much you will pay and what the model is. I think we are flexible on that. The other is some people fear they want more input into the board. We are willing to discuss that. So let's discuss it.

Amadeu Abril i Abril: And perhaps, Sabine brings out this morning understanding what is the real content of this contract. Fay: There is a document around now. It started by CENTR. It moved into ccTLD area. It's a draft contract. This document has a set of amount of consensus in ccTLDs. Please try to use some existing documents instead of starting from scratch.

Brian Cartwell of .cc: I have two questions for ICANN. What is the proposed governing law of this legacy agreement or trilateral agreement? How does ICANN determine what the relevant government authority is going to be and if no legislation passes or if even the legislation has been passed, how they are going to determine it applies?

Andrew McLaughlin: Keep in mind, trilateral agreement is not a three party agreement. It's agreement between ICANN and the registry, just like legacy agreement. The difference is that as to redelegation matters, there is some agreement within the country we defer to in making the redelegation decision. The trilateral agreement is actually a capital T where you have ICANN and then the two parties within the country that we basically acknowledge whatever relationship they come up themselves. So it's sort of capital T picture. As a governing law, option 1: US law, option 2: the law of the individual countries in which case ICANN will have 242 laws that would apply to different agreements. Option 3: use international arbitration priciples, the rules of international chamber of congress. Our NetworkSolutions agreement, for example, provides a ramp-up set of dispute resolution mechanisms. First thing you do, we give notice that they are in default. They have a certain of days to cure, then we first sit down and talk about it. If it doesn't work, we eventually go to arbitration. That is a way to not have to answer that question. You let the rules of international arbitration apply the law that is relevant as developed under the rule of international arbitration. Personally, I prefer that strategy because it avoids having to answer the questions of whose law applies and which court would resolve the dispute. Now you think you have jurisdiction you can always go to court within your jurisdiction or in the US but the better thing is to come up with something like arbitration which would avoid having to answer that very difficult problem. Actually, I talk in some length to Richard Francis about this. He even identifies some of the available rules. I know a good deal of international arbitration. So if you want to talk nuts and bolts we can do that. Your second question was how you decide what the relevant authority is. I think it's pretty straight forward. The relevant authority is the sovereign government for the territory. And it's pretty easy to figure out who the sovereign is.

Brian Cartwell: But you will get conflicting requests from some government agencies within the same government. How you are going to determine which one is correct? So if you receive two, then you just ignore both.

Andrew McLaughlin: We don't pick and choose. You put them in touch with each other and tell them - If they want to make request, they better figure out a way to get it consistent. In the mean while you smile and wait for them to resolve it. Once they figure it out, still that doesn't mean you would do it, now you know what that government thinks. This is not novel problem. This is a problem that people run across dealing with government all the time. We are not unusual in having to face that and we are not unusual in having to solve it on a very pragmatic basis.

Bart Cartwell: I am just curious if you thought about this thing in your head and what you are going to propose as a standard. Will it be a US by standard or is it going to be arbitration process, then US law and then foreign government law? Is it going to be a pecking order in which you do this?

Andrew McLaughlin: NO.

Amadeu Abril i Abril : The best thing we can do is to make this kind of conflict expensive for all the parties. We wrote ageement between ccTLD manager, government, local community, including all the parties. And then wait for the dust to settle somehow.

Andrew McLaughlin: The same principle would apply if you had two different conflicting statements from two different parts of the government, you first want them to work it out among themselves. You always have to have a nuclear bomb scenario that every thing failed, then what would you do? I suppose you defer to greater power but the principle of getting them to work it out themselves before they come to you is the one you would want to follow.

Calvin Browne: On the funding issue, this is relative simple thing to solve. You want to take services and split them. CcTLD has x amount of services that ICANN incur, they should be liable for that. You may have to have certain requirements for developing nations that you help along. Really the traffic on the root should determine who should pay for the root service system.

Stephen Deerhake of .as: My question is for board members and it's as follows - Basic matrix for any registry which is non-service price plan for domain registration, the average price they receive for registration over given period of time can be measured as percentage to the registration price. In the case of registry that I represent, if we would owe for the current fiscal year based on last year average price, based on the example in the proposed fix variable November proposal match the 30 % of that average price. In effect, the proposed ICANN fee is amount into tax of 30 % of our gross revenue. My question is as follow. Does the board consider this rate to be excessive and if so what ultimately do you consider to be a reasonable rate with whatever model for funding you?

Jonathan Cohen: Let me answer you indirectly. It's quite clear that ICANN board through its staff has been trying to get back from ccTLD community, its view of what's fair or what model is appropriate. I tend to feel myself the right solution is along the line suggested by Mr. Browne of South Africa and that is the fully variable method of funding or payment by ccTLDs, was some kind of floor level - I don't know what the floor level should be. I do like the idea of something that measures the price of ccTLD pay based on the level of operation, the number of domain name or fee per domain names. That way it should be the same percentage regardless whether you are developed or undeveloped except you make allowances perhaps at the low end to allow people some latitude. Nobody so far suggested any other model that I think will resolve in something fair. But no matter what model you use, at the end of the day, there is still a number to be achieved. Hopefully in that model it will be a percentage that is digestable and consider to be reasonable. I think the faster ccTLD community come to consensus that this is a model that makes more sense, the faster we can then work on what that number is and then you have your answer.

Amadeu Abril i Abril: We need to revise a system that basis in normal cases with exceptional rules, not base policy on exceptional cases because we will then need 242 different policies for funding. So we need some kind of general rules and then we will deal with all those exceptions which are probably 30, 40 or 50 ccTLDs, no more than that.

Andrew McLaughlin: On the funding question, the least important set of views in the room are us on this side of table. If we can pull this table into circle, it will be a lot more productive. The ICANN staff and board members are supposed to implement what the ccTLD reach through for your own consensus if that's possible. I am not sure that's possible. What is you all as a group concluded as reasonable? I have a request for agenda item. One thing that would be very usefull is to put some time aside for among ccTLDs, the question of what things should be specified as global requirements of all ccTLDs. I assume it's a small set of things. At the end of Melbourne I have to take Fay's ccTLD draft and draft we put together and try to put words on paper. It would be very helpful for me to know what people in this room think of the things that all ccTLDs expect to do.

Peter de Blanc: We have that in our Best Practices document we have spent a year working on it.

Andrew McLaughlin: Right, but for example, the root server operators have agreed to abide by RFC 2180. It is incredibly detailed, extremely demanding set of requirements for the root servers. My guess is that ccTLD administrators don't want to have similar model in place. You don't want a highly detailed Best Practices document that is mandatory to the agreement. 1591 spells out relatively some limited things. For example, providing name services for the TLDs as one of the requirements, does that mean simply resolving names or does it means you have to continue to accept registrations. But if the administrator said I am tired of accepting the new registrations, I am going to stop now. Would that be a ground to consider TLDs to be not performing a public service function? I don't know. I am wondering in the actual document in the conditions that would apply to every body, what are the few things that every body seems to think that should go in there.

Peter de Blanc: I would like to answer that by asking you, Andrew, to reread the one year in the making of ccTLD Best Practices which is pretty much the consensus of what we have already agreed to, or close to it, and throw in CENTR document we even manage to work that in there too. If you feel it should more in there, then submit in writing to us because we have spent a considerable amount of time on this document. It took us a long time to get to the minimum set of agreements that could very well be written on any contract.

Andrew McLaughlin: I am not advocating for more or less. I like some discussion today since we all here in the same room about whether people agree to that list is the sensible one, the one that every body can live with.

Stefano Trumpy of .it: I support what Mike Roberts said. I suggest we possibly stucture on topics because we don't have much time now.

Paul Kane of .ac: I would like to second that. We need to focus the discussion that is going on. One thing that Andrew mentioned which certainly would make me little happier is that if ICANN is doing a checking role to see if the name server is running I can use it as insurance should the governments come along and say you fail on duty to provide services. I can use the third independent services such as ICANN and say this guy checks my service every day, every week and every month and they say 95% of time or 100% of time I am up so it would give some reassurance to ccTLDs that concern that they might have a swift redelegation. One thing Peter mentioned in his presentation this morning that the cost of root server services, ICANN is in dialogue with root server opertor, could you elaborate on the costing if any of the root server services, more importantly, what liability protection there is to ccTLD operators? For example, we enter contract with our registrants to provide service if we have not been negligence in services effectively being provided in our service but the root has been given duff information by ICANN and we get sued. Is ICANN going to protect us or the root server operators going to seek protection from ICANN from which obviously we would be happy to pay. I try to get the feel of expenditure side and the cost side so we can discuss the budget in real detail.

Mike Roberts: I can give you a snap shot on where we are on that. I think we are all aware that we are both in cc community as well as in address community as well as in root server community. There is still fairly strong feeling about not monitizing the activity of the name and address system. The specific discussion that has gone on now for a year and a half with the root server operator that there is a strong feeling on behalf of those operators and organizations that their public service not be monitized. There is also the issue of liability. So long as this is a non-monitized service system and carry out primarily as a public service, then it is entirely appropriate for there not to be damages and not to be liability. ICANN accepted that if we have that in other area of our relationships and then also go to the general issue of damage with regards to our non-commercial relationship with cc organizations. The guarantees that the root server operators have made to the community have always been on a best effort basis. It turns out that in fact the performance of the root server has been one of the remarkable records of technical integrity that any body has ever seen. There is not a disposition among ICANN boards and staff at the present time particularly given the strength of that feeling among the root server operators to change that method of operation. With regard to our proposed relationship with you as a result of the relationships with root operators have with us, we also are intended for to be an exchange of mutual obligations and expectations with regard to operation of name server. The name server that you are responsible for and the root name server that we are responsible for through our agreements with operator. The issue I think of any liability that you may incur to your clients or registrants are the ones in which you would take legal advice depending on your individual situations to limit your liability. Fundamentally we are not in the position to execute an agreement with any of you that incur any form of liability for ICANN.

Fay Howard: By profession I am legal liability under ICANN and it should be just noted that your position is clear but a lot are not happy to take on risk. It put a financial burden on ccTLDs because if something does go wrong, they all get sued. That could cause a lot of problems for ccTLD managers.

Mike Roberts: The history of this is that all of this things started out as a completely nonprofit, noncommercial best effort basis going back in time. The creation of commercial environment that in which you as a matter of your business practice seek risk management advice and undertake insurance is a business decision that each cc makes for itself. Since the agreement is not put in place right now, you have no liability protection right now. And the consensus we feel exist in this community we should not monitize this relationship and we should not commercialize it and we should not create any liability relationship. I don't suggest that doesn't cause some of you difficulty but that where we are coming from.

Pilar Luque of .es: When does ICANN expect to have agreement in place with root server operators? And when does ICANN expect Department of Commerce to transfer the control of the authority of root server to ICANN? Could I please remind English mother tongue speakers to please speak slowly and clearly because it's quite difficult for non English speaker to concentrate all days.

Amadeu Abril i Abril: The second question doesn't depend on us directly but it depends on how fast we reach local agreement with ccTLDs.

Mike Roberts: Louis Touton was scheduled to be here and could give a detailed answer to that. There are three or four pages of the explanation of the current status of the root in the testimony I made last week in Washington which is linkable from the front page of ICANN home page. The operator agreement is in final draft and has been circulated to those organizations. There is some council to council kind of thins going on. To my knowledge, there is no substantial disagreement with the draft as it currently exists among any of the root organizations. We are anxious to proceed with that whether we are talking 60 days or 150 days within that time frame along with other priority. We wish to proceed as expeditiously as we can in that area. There is a whole program of improvement of the root system that was part of the white paper that the root server advisory committee made a series of recommendations. So this is not just a pursuit focus item.

Michael Haberler of .at: Have you ever hear a special case of two overlapping regional geopolitical domain? I am just curious how you will treat this. I smell the tendency of EU setting a precedent for its member states during the process of establishing .eu and probably contract with ICANN. Could you comment on that?

Mike Roberts: The commission is in the process of holding what I believe are consensus discussion within community now to determine what exactly that overlap consists of in term of registration policy and practices. I really could not comment on that.

Amadeu Abril i Abril: There is overlap everywhere. The .uk overlaps with many TLDs that being somehow open and also with .es or .fr or .at for those companies from .uk. But I don't think this will have a reflection the contractual set of what we are trying to discuss here.

Andrew McLaughlin: I am sure the European Commission is a completely democrat open institution that you are able to influence that to have an outcome that you are comfortable with as a citizen of Europe.

Per Koelle: We are talking a lot of funding here. What I would to ask the board and all the staff: what happen to this task force that were created 14 days ago and 35% of the budget is equal to 1.5 million dollars still the sum that ccTLDs suppose to come up with. And then you say we have to come up some plan of ourselves. We have for three years now talk to pre-ICANN and ICANN that we wanted some fixed fee bands that we could select ourselves into. We have asked the staff to come up with what do you need of the money and we wil provide it. A lot of registries will not accept some sort of taxation of domain names and now I hear Namibia come up with $1 dollar for each name because they don't have money down there. Well, maybe they should seek help more UN or somebody else, not from any other countries just because we have a little more internet connection.

Mike Roberts: The budget group for this year has been formed and it's met once to review our mid-year financial report. We posted today preliminary budget document for the new year starting next July 1 for comment by everyone. The budget group, which has 3 cc representatives on it, is going to meet in face-to-face meeting with the board finance committee in Melbourne on March 9. There is also in the budget document a whole calendar and series of questions invite people to examine the document and make comments on it. With the respect to specific questions of the name registries contributions, what the preliminary budget said is that is going to be work out in the consensus process of the budget. So the answer is still in front of us. The specific portion of total domain names revenue amount is not shown. So the answer to the 1.5 million number is not there because the assumption is some version of integrated structure that proposed in the staff paper will be adopted during the next 120 days or so. The important point here I try to make is firstly, you have a group of finance professionals in the budget group with cc representation on it. They are talking to the board and to the staff about everything to do with the budget. Secondly, you individually have lots of opportunity to examine the document and make comments to them through our normal comment process.

Andrew McLaughlin: The period of comment, it makes me want me to bang my head against the wall because I hear comments that say - how dare ICANN proposes the funding model, this is up to ccTLDs to figure out among ourselves and we will tell you what the answer is. And I hear other comment that says - why haven't you publish the funding model that just solves the problem? Can I take a quick survey? How many of you would like ICANN staff to keep publishing funding model until we get one that gets general support. This is serious question. Is it possible for this group, ccTLDs, to reach consensus? We heard some people said very strongly I want a variable model, some people said we want fixed model, we want self-selected bands. How do we get to closure on this? ICANN is so small and we do so many different things. What you are paying for is not simply a set of technical services. It is a transfer of management from the US government to international private sector organization. That is in part what is being subsidized. It can't be that specific. What you are being ask to pay for is a chunk of this overall process. And if we do our job right, we will lock into budget so there will be no danger of budget growth driving the organization to become larger in policy area. We have done that successfully in the registrars side in .com, .net and .org. We have a defined-fixed budget that does not increase from year to year. If the allocation among the various people paying the money changes according to which ones are more or less successful in that market place.

I need some guidance about whether I should be playing a strong role in trying to define a model, or not a strong role which I am happy not to play frankly. How is this community going to reach some kind of consensus when there are very strong differences of view, apparently irreconcilable.

Mike Roberts: The issue whether ICANN cost should be recovered through some sort of specific cost accountant methodology was discussed at some length by the task force on funding a year and a half ago. Their conclusion, not ICANN staff conclusion, was it was not even worth started. If you want to reopen the issue of trying to do cost accounting for the portion of revenue required for 4 - 5 million dollars. Then you should ask your representatives on the budget group which is where the finance professional people who know the complexity of cost accounting to bring it up.

Paul Kane: I am falling on Andrew's point. It's jus a brilliant idea. Why not just a brain storming session. Put it on the table, let's say one million or one and half million, go away cc, sort it out and come back, just pay the checks and possibly refer to some of the larger cc to effectively underwrite it. But no more of this name taxes, it's just turning into quick mine just where we were two years ago. You need the money and cc has some funds to give you but they want to be fair and equitable. So you just have an overall budget for the cc portion, break it up to a little unit. That is what you have done with gTLDs, NSI caps two million. Let's try to move over.

Mike Roberts: The difficulty with what you are proposing is that it is assumed it is easy to find a line that defines the economic interest of the cc separately from the historical gTLDs and the new gTLDs and from the address registries. Our experience over the past year and a half is not at all easy. In fact we don't have agreement on how to do that. That is why the new revenue staff paper said let's treat everybody in the same economic circumstances the same. And I think we are hearing various voices here today that saying - yes, and let's make it proportional to level of your activity. I am not taking that as conclusion. I am just taking that as evidence of voices we heard.

Oscar Robles: I just have one comment and one question. Following with this financial model, I have a comment that it should not be just complied with Jonathan comment condition that just mentioned. This model should not prevent ccTLDs to become more efficient. Let me explain. In order to enter this ICANN financial model, ccTLDs must become efficient just establishing fee or establishing registry-registrar model or get rid of domain restrictions because it is not cost efficient. One of these three simple solutions represents to jump to next fixed fee classification. We talk about efficiency. We couldn't be efficient 10 times or 100 times. It should not prevent ccTLD to pay more if they are able to pay more. My question is: does ICANN prepare to deal with countries restricted by the department of Commerce like Libya, Iraq, and those black listed countries.

Andrew McLaughlin: They are part of the root. They are part of ccTLD community. ICANN position is that we deal with the countries that are part of the root.

Anthony Bishop of .tv: We talk a lot about funding. We talk a lot about the local internet community. I echo Andrew's comment. I like to hear more about proposed contract between ccTLD managers and ICANN in trilateral arrangement. Contrary to Peter's statement, I think a lot of us in the room don't know all those provisions are that in the draft document. I think it would be helpful to talk about what basic expectations are.

Jonathan Cohen: Is there any way we can agree since we are running out of time to list the two or three items that are really most important and simple focus on them one by one. I came here representing the finance committee of ICANN and there are a lot of circular talks but I am not sure what to tell them you people think. I just have a feeling at the same time that we just focus one, two, three, we can actually make some decisions. There seems to be decision circulating the odd points that is difference. But I don't think you are all that far. You are discovering today ICANN board is not that far away from you on most issues. Can we do that, the Chair?

Mike Roberts: If you do the preliminary budget number of all the name registries, it 's about 4.3 million dollars. That's before the board look at it and before the public looks at it. That is the number going in for next year. If some people have said we will go pass 40 million registrations before the beginning of the fiscal year. So it's about 10 cent a name would cover it. What structure of the revenue contribution from all of the name registries does a better job and it's fair and easier to do than 10 cents a name.

Agathoclis Stylianou of .cy: Hearing about 10 cents a name or $1.00 a name, I think it is not the issue I asked this morning and I was told that it would be answered this afternoon. According to me, the question is it's not how much money we are asked to pay but why we pay this amount of money. It' s not clear and I am not getting clear answer back. Just like the contract, ccTLDs agree to pay ICANN this amount of money and in return ICANN agrees to do service. I can not answer this question to my peers either. I am referring to quantification that Andre or Mike has mentioned earlier on. The second component of my question is that it seems the money requested from ccTLD community does not much involvement of community in the ICANN structure. Are there any thoughts being giving towards rectification of these?

Andrew McLaughlin: What you get back is an organization that coordinates, unique identifies the name spaces, IP address space, and protocol port and parameter space. Those three spaces require some form of coordination to preserver the uniqueness of identifier across public, nonproprietary, open domain names space. So that is the service you pay for. That includes the provision of an authoritative root zone file that distributed across root server. That means you get IANA staff that provide free protocol port and parameter numbering services. It means you get the staff that tries to provide basic policy coordination among the different names registries and above all, you will get the structure that tries to move management function out of US government into an international nonprofit in which you have a say both as to the budget, as to activities, as to the policy, as to the identity of Director and so forth. That is basically what you pay for. We can go down what each individual employee at ICANN does and I think you find that a lot of them do think that you would conclude we are worthwhile. Fundamentally, it's that coordination function of unique identifier that secure your ccTLD registries place in authoritatively unique root.

Agathoclis Stylianou: But so far question is about quality and liability and those were being sent back to us saying no such issue defining the quality and liability of service.

Andrew McLaughlin: That is not true. We always said that we are happy to commit to standard of services for the services that we provide. That is different from saying that it's the contract for services. It's a contract that includes services and obligations on both sides. The model we are pursuing here is a public service model that each of registry or organization in this universe is performing a public service. You are doing in many ways with the great diversity and approaches and structures and so forth. But fundamentally, it's a service for the benefit of Internet community as a whole and as such the contracts and agreements and funding models represent the notion of common enterprise to provide the service to Internet community which can include components with very specific service guarantee. We have never given any resistant on that point whatsoever.

Coffee Break

Start at 17:00

Elisabeth Porteneuve: Three items we will discuss in the last hour here: Contract, Finance and Representation. Fay Howard will give a brief presentation on Contract document.

CONTRACT ISSUE

Fay Howard: During the ICANN meeting in Yokohama, we were represented with the Status Quo document by ICANN Staff, document for discussion on contractual relationship between ccTLDs and ICANN. It came apparent during the meeting that this document may be not suitable to many people. Following that, Willie Black of .uk went away and thought about putting an agreement together a very lightweight document that we could build consensus on. He did this and it was taken to CENTR legal group and they refined a little bit more. We left some issue open which is applicable law jurisdiction and we struggle mightily with the redelegation issue and decided to leave it out for the time. So what we have is a lightweight contract that just basically set out the fact that this is the manager that is obligated to do task running registry and keep it going through technical requirements, ICANN few technical requirements there. It talks about remedies in case of breach of this agreement and it's also interestingly enough in relation to finance tied down the flat amount a bit of consensus building flat amount of fee to ICANN and quite a bit of consensus really where big constituency is hard to reach agreement and is hard to all get together. Maybe we should agree on flat amount. We should divide among ourselves and individual registry will commit to paying this fee for a set of number of years, it is in the contract, it is your business, nothing to do with percentages as long as you reach the total required. We passed this contract out to ccTLD constituency and the working group looked at it in LA, made a few adjustments and then posted among ccTLDs for comment for consensus building. We have actually passed this document to ICANN. We have quite constructive comments two weeks ago from Andrew and things Mike needs to be looked at. One thing I would like to say to ICANN is that in looking at new document, please use the document that has the amount of consensus from global community as basis. We also take on both the issue of suggestion that we leave the issue now of jurisdiction. A lot of people and lawyers and CENTR managers are quite happy with the idea of international arbitration and stages for getting there. I think we need to get on with it now. It would be really useful if we have one or two documents for people to read that might help get pass the issue of redelegation.

Andrew McLaughlin: Could you say word or two in your document about relationship between Best Practices Principles and Contract whether it would be mandatory or there would be variation across all of their principles. What would you basically envision required to be in every contract?

Fay Howard: We came up with Best Practices document which was set out high principles rather than intricate detail. Some CENTR members signed up to a slightly amended version because for one reason or another the construction of their registries, the legal requirement, they could not quite sign up to it. Basically principles are there. I think ccTLD document that we worked in LA, it's quite high principles, it's an outline, it doesn't get down to individual policy. I think it would be for each registry the manager agrees to work by Best Practices principles at the top in Appendices 1 and 2. It may be for some registries to tailor it slightly.

Mike Roberts: Let's me summary the state of play here a little bit. As I said this morning, we will hold back on actually publishing document with legal level language in it until after Melbourne because we are making progress on this agreement and we have a face-to-face meeting in Melbourne that is important to wrap up this cycle of consultation-type-style meeting. But I think it's fair to say that we see on the structure that it's been involving from this end and taking account on the homework that CENTR has done, essentially five sections or paragraphs agreement of which four are substances, the fifth is miscellaneous things that need be dealt with in legal documents.

The First section would deal with mutual obligation of ICANN and registry manager. This would be where the general public interest language would appear.

The Second section would deal with delegation and redelegation issues.

The Third section would deal specifically with the mutual obligation and expectation with regard to operation of name servers, both technical-administrator and operational elements of that due to general concerns not only to ICANN and managers but also to the broader community.

The Forth section would deal with funding support. The present feeling about that is that the language will abridge manager to participate in and abided by the consensus budget development and adoption policy of ICANN. In addition to that it would specify the specific amount of money for the initial year and a percentage limitation on any unilateral increases on year-to-year basis on part of ICANN.

I think we have some distance to go. But I want to give you that as an indication of the thinking about where we are right now. There is no particular need for this structure to be different between legacy and trilateral agreement. The part of the ICANN-manager agreement that exists in trilateral arrangement except that obviously the provision on the section of delegation would be different because of the transfer of public interest responsibility in that situation to local government. It possibly would affect some of the language in the first section on mutual obligations. This would depend a little bit as I think those of you who reviewed language in Canadian agreement on the specific of local arrangement. I hope that is useful too.

Sabine Dolderer of .de: You talked about what we should expect after Melbourne and you also present legal document. But I find interest about the draft we have produced because it's not only for CENTR but also for the whole ccTLD community. Therefore, I would really appreciate if there is legal response. What do you think about this draft if you can leave with that and if not, why?

Mike Roberts: I was recognizing that there is a section here I omitted and that is the section that deals with disputes and settlement of disputes and this would also include question of outreach. I only mention that because the disputes things deal with various issues of liability that are obviously attended subject between us and I didn't mean to infer that would just drop into Miscellaneous. So our intention here is to come up with a document that represents a middle road of where CENTR document was as edited by APTLD with letter sent to me in November and by the three meetings we have now. Plus by public discussion on the public forum day where we are in the process of putting it together the opportunity for both ICANN and cc community to present their views as part of public forum day.

Andrew McLaughlin: Let me just say to Sabine and Fay. I hear you about your document and how much work that has gone into it. I will do my best to try to respond to the document specifically and incorporate where possible.

Anthony Bishop: Can anyone summarize briefly area of disagreement between ICANN position or draft and ccTLD draft that's on circulation? Are there a handful of issues that separate the two drafts? Secondly, the one area that has not been discussed very much is the area of mutual obligations and I assume that is the Best Practices that they were referring to and maybe the brief summary of what some of those concepts are.

Fay Howard: Some of the things that you highlight when you look at the ccTLD contract a few weeks ago in Zurich might be useful.

Andrew McLaughlin: I have not done my homework of mapping out the document that does that. I hope that I can do that in a timely way. Just to try to summarize. The biggest problem is delegation/redelegation. There seems to be general agreement that the principles of RFC1591 are fine. Figuring out how to write that into a form of document is really hard at least as far as legacy document is concerned. It basically keeps IANA and ICANN the responsibility to do redelegation but we need to do so by giving ccTLD comfort. That is not going to be done arbitrarily. Give governments the comfort that their views will be taking seriously and give the community comfort that their governments will not going to have automatic veto power over it. It's very hard to do that in actual language. But from what I can tell, it doesn't appear to be. But the legacy agreement, there are a lot of disputes about those principles. For a more capital T agreement (Trilateral), I expect to have more variations and basically whatever the ccTLD manager and local government figure out what is acceptable among themselves as long as it's broadly consistent with these principles we have been discussing, then ICANN job is simply to acknowledge that structure, not to get too deeply involved in it.

One of the things that I have been denouncing earlier is whether you call it "Contract for Services". For the reason that outlined, I am not comfortable with that. I would like to think of it as relationship agreement that defines the relationship between these two organizations including obligations, including services, including even specific service guarantee. But it's an agreement whereby ccTLD and ICANN and the other ccTLDs and the other registries basically create a community, a policy structure, a way of resolving disputes, developing policy, allocating the responsibilities in the way that doesn't direct mapping of commercial services. Just to put my lawyer hat for a moment, a big danger with a commercial services contract is that it would turn ICANN into a trade association which look awfully like a cartel which can really attract the attention of anti-trust authorities. It 's not in anybody interest to do that. It is much better to demonstrate that this is self-regulatory effort by internet community as a whole and not a potentially restricted trade cartel essentially monopolist within their given areas. That is one thing I objected to. That doesn't actually go very much to the merit of the contract. It just goes to the label that you put on the top of it whether you call it Contract for Services or Relationship Agreement. By way of example, the agreement we have been working on with RIR, relationship agreement, we don't get into remedies or disputes because we assume either the relationship functions or it doesn't. If it doesn't function, basically ICANN process blows up. It's been very difficult for all of us to build some sense of trust in each other over these last two years. I hope we can achieve some degree of trust underline these document, otherwise it becomes very hard to nail and resolve every potential dispute in the document. It has got to base to some degree on trust.

These are the things that's on the agenda this afternoon. Those are the three big ones where there is not yet agreement. On many of the details, there is substantial agreement.

Mike Roberts: One thing I like to add to the footnote is that one of the things we need to avoid in term of any agreement reference to the best practices is the expectation that ICANN is some sort of policeman about routine operational and technical activity of registry. Under the present situation which as Eberhard points out, it isn't working all that badly. That is not going on. Certainly the technical contracts in TLDs are very much aware when one of the TLDs got sick for some reason and the words of mouth we deal with that one way or the other. Fundamental principle to be embedded in the agreement is that you are competent to operate a world class Top Level Domain Name Registry. If you are not competent to do that , we will hear about it from the third party. So there is no point in conceptualizing some policemen role for ICANN staff which is got plenty to do already.

On the issue of timeline, this is sort of tricky arrangement. We will talk about money here in a few minutes. Obviously we have not settled the arrangement for this year and we are already working on the arrangement for next year and we are not very happy about the arrangement for last year. What I have to do on this point is to concentrate on the work of budget group and whatever broader input you like to make to settle on next year. Then to adapt whatever that agreement is to cover this year. We want to forward looking arrangement here. There is no point of rehearsing last year water over the damp. We need to do now is to come to some agreement given a new situation, a new registry coming online, how we want to do it starting the year July 1 and then make whatever adjustment to that for the current year. Some one asked me during the break - are you going to require agreement being signed incident to anybody making payment this year and how you are going to do all these agreement between now and the end of June? It seems to me that obviously we want to do as many agreements as we can between now and the end of June. To the extent that we can't or there is some consensus about how to deal with not being able to do that, we would essentially use the same mechanism you all came up with last year and that is in your best judgment you undertake to make voluntary contributions.

Peter de Blanc: Mike, I may be a bit confused between three way agreement and capital T agreement, but it's the same. Early on, you and Andrew talk about in a three way government and one of the objectives was to get the government acknowledgement that would not be arbitrary or redelegation. Few moment ago, you said that in a three way agreement there would be a transfer redelegation to local government. That is what essentially would be happening. Is it the transfer redelegation authority with an acknowledgement of no arbitrary preaches. Where is the global internet community get involved in the dispute resolution process in a three way agreement?

Andrew McLaughlin: The idea is not that you transfer redelegation authority to local government. The idea is that you transfer redelegation authority to a structure that is agreed between ccTLD manager and government. That could take the form of contract as in the case of Canada. It could take the form of a nonprofit organization that has the ability to select the technical manager for TLD. That is also happened to be the case in Canada. I don't have any particular view of what form that should take. But the GAC principles describe one set of relationships that would clearly qualify as an acceptable structure to assume redelegation authority. If you read those principles carefully, you will find that they are attempting to limit the authority of government to take arbitrary actions, at the same time giving the government a role in protecting the public interest.

Mike Roberts: Let me one other point for clarification. We are not talking about a three-party legal document. We are talking about three different two-party agreements. Two-party agreement between manager and ICANN, certain of these sections has different language in them depending upon whether a legacy situation or trilateral situation. For instance, in the agreement with Canada, the section on mutual obligations and on redelegations will reference the umbrella agreement that is in existent that confirms the structure for the local responsibility.

Andrew McLaughlin: In both cases, the only agreement that ICANN signs is an agreement with ccTLD manager. The difference is the redelegation authority resides with ICANN which is Legacy agreement, or shifts to something inside the country which is the structure we have been talking about. It might look like GAC principles and you can look at Canada variations on that if you want to.

Amadeu Abril i Abril: One thing I want to address Peter's concern. Andrew and Mike have mentioned a lot of time the Canadian agreement as one model. It's A model, not THE model. There are other ways of drafting redelegation mechanism that is sufficiently good. The problem is we don't have the alternative model to present right now.

Peter de Blanc: You basically are saying that by coming to disagreement among ourselfves, should there be an attempt of arbitrary redelegation that we collectively both ICANN and manager bury local government and bureaucracy to the point where we find out if it's really a serious redelegation or just the situation where the new priminister wants his cousin to run TLD.

Andrew McLaughlin: If you do our jobs right, the agreement would be sufficient detail but it gives you comfort in writing that you know what to expect when redelegation request comes from government. Another words, what constitutes the request from government, and then what ICANN does when that request comes in. That is the thing to make this transparent and predictable.

Peter de Blanc: Then by extension, we would need to have in place either through our ccTLD organization or somehow a fairly elaborate redelegation dispute resolution system.

Andrew McLaughlin: No, the idea is to set up something inside the country so that local community makes that decision. If people who should decide this are every other ccTLD managers except the people in that country, you will get a lot of resistant.

Bill Semich: The question is I heard Andrew said only one agreemnt between ICANN and ccTLD manager and Mike said there were three agreements. Where are the other two agreements?

Andrew McLaughlin: The third agreement is a communication from government to ICANN. It 's not a contract. It's not agreement, just one-way communication.

Mike Roberts: If you examined Canada's, there is a letter from Industry of Canada to me. Read that letter. It will give you the flavor of intended relationship.

THE END OF GENEVA MINUTES IS MISSING

The straw poll on ICANN funding model "proportional some way to the number of domain names" vs. "self-selected" was taken, and the result was:

29/8/9, i.e. 29 in favour of "proportional", 8 against, 9 abstentions.