ICANN|ARCHIVES

Internet Corporation for Assigned Names and Numbers

>>ALEJANDRO PISANTY: Ladies and gentlemen, this meeting will start as soon as more than ten people are seated.

So we can start. The meeting will be chaired by Roberto Gaetano. While he becomes available, Ms. Rita Rodin will chair the meeting.

The agenda, as you know, has been posted for public consultation opened by the Board Governance Committee of the ICANN board, on the subject of the GNSO review.

A number of questions and papers have been set out. And the purpose of the meeting for the BGC is mostly to gather a final record of input and especially to observe the discussion and the dynamic of the discussion in the community before further steps are taken in the GNSO review.

As many of you know, the GNSO review was started by a review of the GNSO Council, which included very important steps of self-review. It was followed by an analysis and input and feedback on that work. And then a team was -- an RFP was put forward. The London School of Economics team of political science of Professor Patrick Dunleavy was selected to conduct this view. It took many months of interviews in the community, of field research, document research, and so forth, until they came out with a number of recommendations and questions. These have, in turn, been subjected to -- I would call it massive community discussion, discussions within the GNSO Council, within the GNSO constituencies, and, in a broader sense, of the community.

Input is still coming in. I am about to read, and many of you will already have read, an e-mail consideration by Danny Younger about whether the GNSO should exist at all in the ICANN structure. It seems from the provocative title that I read that he considers that it doesn't anymore perform a useful function. So it apparently seems that there still is a very wide open discussion on the GNSO that has to be fed into the review.

So I'll hand over, after this introduction, to Rita and Roberto to decide which part of Sicily starts.

>>ROBERTO GAETANO: Okay. I would like to make first a quick update on how we are and what -- how are we going to progress. And then I would like to give -- to use this opportunity to leave as much as possible time to the floor, because the purpose of this meeting is more for us to get input than to do our usual presentations.

I only would like to say that we are here, the Board Governance Committee and the chairman of the GNSO, we are here mostly because the Board Governance Committee is in charge of the process for the GNSO review, and the chairman of the GNSO, well, you might guess he is somehow one interested party at least. But also will support us in the moment that we will have some discussion, also for the functioning of if we get into the functioning of the GNSO, then he can address the questions.

We are not -- in this meeting, we are not going to provide answers to suggestions that come from the floor, but we are only here to collect the input and work on it.

So what is going to happen next?

We have just finished, the Board Governance Committee, that has populated a working group that will deal with the next steps. And so that means that we'll analyze all the material, the material that has come from the London School of Economics that is the report, from the different constituencies of the GNSO who have provided comments, from the community. We had an address forum where we had inputs from the community. And some of the discussions that we had in the Board Governance Committee and on the board.

So now we have collected all this material and we are in the process of nominating a working group that is composed by board members and former board members that is going to be formally appointed by the board with a resolution of the board this coming Friday at the board meeting and will become operating right away.

So this is basically it.

There was in the call for this meeting the indication of three areas that are of particular interest that we think can be controversial and that will probably get some input. You have probably had on the Web what those specific recommendations are and what those areas are. But just for the recapitulation, one is the composition of the GNSO. There's this famous recommendation 19 that I think is going to be one of the most controversial in the history of ICANN that is suggesting basically to review the constituency structure of the GNSO. And that is the first area.

There's another area, which is the voting. And the voting, there are several proposals for changing the voting. For instance, the (inaudible), the weighted voting, and raise the ceiling for consensus to 75%, and, you know, these kind of things.

And there's a third -- Denise, what's the third area? It escapes me.

Yeah, no, that I said. So the second one.

Oh, policy development process. This is if the policy development process has to be changed, has it been effective. And that is probably the most important, because it really addresses the purpose of the GNSO rather than the -- so the what has to be done rather than how the GNSO is operating, with which structure and with which voting it has.

So I would ask if there are any preliminary comments from members of the Board Governance Committee here, if there's something that I have forgotten. And then I would let you open the fire.

Okay. Yes, so it has been pointed out to me that it will be good if I explain what happens afterwards.

So this working group will evaluate all the material and will come -- will evaluate the recommendation and will make a summary to the Board Governance Committee with their recommendation. The Board Governance Committee -- or ask for further input or further material if there are gray areas. The Board Governance Committee will evaluate this and make the recommendation for the board.

I don't know if we are optimistic or foolish or what, but the idea is that this working group will finish their work within two months' time frame so that we can evaluate all those recommendations and present them to the board, issue, of course, another public comment period, and be able to make a resolution of the board in Puerto Rico.

We are all aware that we have a tight schedule. But we are planning to populate this working group with hard-working folks. And I think that we can make it. And, in any case, we have been working on the GNSO review already for quite a while. And I think that it is time to drive this process to conclusion.

Any other preliminary comment?

Dr. Vint Cerf.

Can you -- can you say your name to the --

>>VINT CERF: Yes. I'm Vint Cerf, chairman of the board of ICANN.

When you mentioned driving the process to conclusion, I was only going to make a plea to make sure that we drove it to conclusion and not to distraction.

>>ROBERTO GAETANO: That will -- we will take care of this.

So the first topic is the structure of the GNSO and the constituencies.

And the proposal is to radically simplify the structure of the constituencies, and specifically to reduce the number to three. And one that is related to the technical, I would say, operators, which includes registrar/registries; one to the business users, that will include the current business constituency, the intellectual property; and then to create a third constituency that will be noncommercial user and civil society, with the possibility of including in that constituency also not only the current NCUC, but, for instance, the at-large community that is now represented by the At-Large Advisory Committee as a committee of the board but that doesn't have a specific voice on the GNSO.

So I'm sure that there are many comments from the floor and approaches to this. And I would invite comments from the floor on this aspect.

>>VINT CERF: It's Vint Cerf again.

One of the -- if I've understood the proposal correctly, it was to collapse things into, say, three different groups, which would lump all of those involved in registry -- registration matters into one common element.

I don't understand how that would be very effective, because I don't understand how such an element would come to much consensus, given that the parties who are interested in registration processes are sometimes competing with each other, have business relationships with each other, which is why we have separated in the current operation the registries and the registrars.

So although that sounds like it would make things simpler, I'm not sure that it would be actually, in practice, simpler.

So I guess, I would raise at least that question and ask if anyone has a different view.

>>MARK McFADDEN: Mark McFadden, and it turns out, I don't have a different view. I have the same view.

I think that this is a solution in search of a problem.

There's no evidence at all that simplification of this sort is actually solving a real problem of any kind.

And in fact, in one of the simplifications, in recommendation 19, there is the suggestion that you would have a constituency that combines suppliers and users. And those are naturally competing interests. And so one wonders how any bottom-up consensus process would work in an environment where the group, the constituency, inside the GNSO would have naturally competing forces.

I don't think that you would ever have a mechanism by which you could actually build consensus in a group like that.

So one of the things I'd say to the committee is, as you consider this, consider what problem is being solved, and there's really no evidence that there is a problem being solved here.

The second thing is, even if -- even if you were to do this, think about what would be in those individual constituencies and ask yourself what -- how would consensus be built in a bottom-up process I submit to you that because you would have competing forces in those so-called simplified constituencies, a policy development would be even more difficult than it is now.

>>RITA RODIN: Chuck, I am going to make one quick comment. I think this notion of having just three constituencies, Vint, is what was in the LSE report. The BGC is far from agreed that that is a good approach but was just interested in hearing comments from the community on that.

>>CHUCK GOMES: First of all, and you probably already read the registry constituency comments -- And I am Chuck Gomes. I guess you could see that there. Sorry about that.

And there was a lot of thinking among gTLD registries that putting registries and registrars together was probably not workable.

Okay?

Especially within the last couple years and what we've seen.

My own personal thoughts in that regard, it's less -- could be less problematic if we can separate contractual issues from policy issues, which hasn't been a real clear separation.

But again, registry constituency really suspects that that might not work.

The more important issue for registries, for me personally, is not so much the makeup of the constituencies as the policy development process itself.

Going back historically, the DNSO, the predecessor of the GNSO, really was trying to do some bottom-up processes, and some of them worked. The first introduction of new TLDs was a process that worked.

But it was very difficult. Bottom-up processes are hard. They take a lot of time. And there was frustration that resulted because of that.

And eventually, then, the DNSO, later the GNSO, went to more of a council/task force type approach with a lot of voting. And I think that the LSE got it right in another one of their recommendations when they said that we need to de-emphasize voting and work towards really building positions that everyone can support.

That is the more important issue, in my mind, than the makeup of the constituencies.

And in fact, we have actually seen some recent successes in that.

The new TLD committee that just had a session that ran over a few minutes ago and ran into this session, that process -- now, it hasn't been quick, I don't think it could have been quick, but to my knowledge, and I think I have been involved all the way through, there hasn't been one vote taken. And yet we've worked towards some positions that are strongly supported by the group.

That is the more important issue than the makeup of the constituencies.

And then just one final comment with regard to weighted voting. I keep hearing people saying that weighted voting doesn't work.

I think what that means is, is that they haven't been able to get their position approved as easily as they had under the previous position when users dominated the DNSO and the providers, who had to live with what was decided, had to implement it.

But again, I don't think that's the important issue.

Weighted voting or nonweighted voting.

The important issue is let's get away from voting and let's work towards building consensus.

And like I said, we have a really good example of that in the new TLD process. It can work. It won't meet the time lines that are in more complicated cases that are in the PDP as we see it, but it can work and we're showing that it works.

>>PETER DENGATE THRUSH: Chuck, before you go, I wonder if I could take issue briefly with your characterization of the 2000 round of new TLDs as a product of consensus of the DNSO.

My recollection of it --

>>CHUCK GOMES: No.

>>PETER DENGATE THRUSH: -- really was it was something the board took over because of the failure by --

>>CHUCK GOMES: I'm glad you ask it, Peter. What I was referring to was the process to introduce new TLDs that went through the policy development process.

There was a lot of disagreement in that at first. Some people -- But there was -- there was actually -- what was it? Working group --

>>PETER DENGATE THRUSH: Okay. My real question of you, Chuck, is this.

I'd like you to tease out if you could a little bit more the difference between the policy discussions that you think are appropriate and the contractual negotiations which you think might not be. And could you think of that in terms of the fact of the earlier discussion we had today. The contract mechanism is the primary mechanism for regulation in this structure.

So if you are going to take the policy issue somewhere and have different contractual negotiations, and the contracts are how we maintain regulatory order, can you -- how do you see -- why would you separate those two the way you have described them?

>>CHUCK GOMES: Well, I think in registry contracts right now, they are already separated. But they are not separated in the PDP, or in the bylaws of ICANN or the process that the GNSO follows.

And so there's a disconnect between the two.

And as was pointed out in previous ICANN meetings, I think Susan was the one that pointed it out, there is a disconnect there. And if that connection -- if that was more in sync, it might be more realistic for registries and registrars to work together. That's Chuck Gomes' personal opinion, though, not a registry constituency position.

>>PETER DENGATE THRUSH: Thank you.

>>ROBERTO GAETANO: Just one thing. And then we will probably have some more time to go over voting because that's the third big chunk of the discussion.

>>JON NEVETT: Thank you. Jon Nevett, chair of the registrar constituency. I want to echo the concerns that Dr. Cerf raised and the concerns that Chuck Gomes raised about combining a registrar and registry constituency into one. Based on the issues that were predominant over the last couple of years, you can imagine that the lack of ability for a registry and a registrar community to get together to come up with some kind of consensus on a lot of these issues.

There's also an issue between numbers. There are -- what? 14 registries and 885 registrars. How would we work with those numbers?

So that would be a problem as well.

And you have the supplier/customer relationships.

So there's just a host of issues. And like one of other speakers said before, at least from my perspective, I don't see anything wrong with the current constituency perspective between registrars and registries. We each have our own. If there was one combined, we would inevitably break off into some kind of subgroup or subcommittee to issue a statement here or there. So we are going to end up being separate anyway, so it sounds like it probably makes sense to keep the system the way it is now.

I know you mentioned we'll talk about voting later. From my perspective, the 75% voting system with three constituencies is not workable either because you are giving each of the three constituencies, which is a one-third group, essentially a veto.

So consensus would be much more difficult to arrive at as opposed to now, if various constituencies get together, no one has a veto power.

Thank you.

>>ALEJANDRO PISANTY: It seems that at least in this room we are not getting a lot of expressions in favor of bringing together registries and registrars into a single constituency.

Just for practical purposes, if there's anybody that does think so, could you raise your hand and maybe come to the microphone later? We could conclude that in this room at this time, there's not much support for that proposition.

>>MILTON MUELLER: Yeah, I would speak in favor of the other constituency restructurings that were proposed by the LSE. I agree if the registries/registrars were combined, the same thing would exist within the constituencies and if you had two or three, two or three would be registrars and two or three would be registries.

So it's not clear you would accomplish anything by merging them.

But with respect to the business user constituency, I think there was a very strong case that was made by the LSE report and there's a very strong case in practice.

It was evident from the beginning, and I did in fact argue this at the beginning of the DNSO, that trademark, as a constituency, is a subset of business users. And indeed, if you look at the composition of the business constituency, it is basically the same interests and the same positions, and meet together.

So why double the representation of one particular interest at the expense of the other groups? You simply are doubling the representation of trademark interests, or possibly even tripling them when you consider that the ISP constituency really, again, has met together, taken almost always the same position as the other two business constituencies.

And so that's the problem, to respond to one of the speakers, what problem is this trying to solve. That's one of the problems it's trying to solve.

And then to create a broader category of civil society could be a good thing also in the sense that noncommercial organizations are, again -- We've had to deal with heterogeneity. It's kind of amusing to hear the registrars and registries say they can't deal with people with fundamentally opposing interest. Within the remit of our constituency we have organizations like the Red Cross, that consider themselves trademark holders and frequently take the same side as the trademark constituency. We have developing country interests, we have developed country interests and so on and so forth. There's no homogeneity, so we have had to deal with that from the beginning.

Nonetheless, we have managed and I think these broad groupings with five representations per group could, indeed, provide a very good cross-section.

You shouldn't assume all these constituencies will vote uniformly, and indeed, that would be your only hope you would not deadlock.

The only problem with the current structure which I presume you are aware of is that it is completely deadlocked. It is entrenched. And if it weren't for weighted voting, we would be deadlocked on almost everything.

There's hardly a single controversial issue that the GNSO has dealt with that would not -- that would not fail to get a supermajority if it were not for weighted voting.

So one thing, one of my nightmares about these recommendations is you will not restructure the constituencies but you will eliminated weighted voting and you will raise the threshold to 75%, which means that let's just not go to meetings and let's not go to anything because the GNSO will never accomplish another significant act for the rest of its existence.

I just can't see -- You either have to take the whole thing or you can't eliminate -- you kind of have to stay with the status quo, fundamentally.

>>PETER DENGATE THRUSH: Sorry, Milton, a question from Pete. Aren't the three logical constituencies the registries, the registrars, and the registrants, and that the difference between the flavors of registrants, whether they are big or small or commercial or noncommercial is something that should be decided -- is an internal matter for a registrant's constituency?

>>MILTON MUELLER: That idea has some appeal, as I told you last night.

But the idea of putting me and Marilyn Cade in the same constituency --

[ Laughter ]

>>PETER DENGATE THRUSH: That's another good reason for doing it, yes.

>>MILTON MUELLER: It's a great idea.

For people who enjoy mud wrestling.

[ Laughter ]

>> Is that a promise?

[ Laughter ]

>>MILTON MUELLER: Somebody, somebody.

To be more serious about that, it's not just that the -- that they are, indeed, fundamentally different categories of users. To say this a multi-national corporation with a hundred or a thousand domains, professionally managed domain portfolios, trademark portfolios is in the same category as a public interest organization in Korea that builds noncommercial networks because they both rent domain names is stretching things a bit.

Fundamentally, yeah, we have often voted with the business constituency on issues regarding the regulation of registries because as consumers of domain names, we have a common interest.

But when it comes to the more important issues, things like tradeoffs between IP protection and freedom of expression, or between IPR protection and privacy, the noncommercial public interest groups have a very divergent perspective from the business groups.

>>JORDYN BUCHANAN: Hello. I'm Jordyn Buchanan, and I'm speaking still only for myself.

So let me make sure I'm talking about the right things before I start to talk.

It's okay to talk about the groupings and is it okay to talk about the PDP itself as well right now?

>>ROBERTO GAETANO: We were making this division in order to kind of structure the debate.

>>JORDYN BUCHANAN: I can wait a minute. That's fine.

>>ROBERTO GAETANO: But if you talk about one issue, that doesn't jeopardize your right to come back to the microphone on the second one.

>>JORDYN BUCHANAN: I understand. With regards to the grouping, listening to Milton and having heard some other things throughout the course of the week, it's not entirely -- it seems obvious to me that you're not going to be able to get perfect divisions of -- in the constituencies that are always going to sort of make sense for the given issue at hand.

And so we can either simply accept that and realize we're not going to develop a scheme of constituencies that are going to always sort of match up with the interests for any particular issue, which is okay. I think -- So the recommendations to have only three constituencies, I'm not particularly beholden to the idea of combining registries and registrars. But I also think there's a lot of issues where registries and registrars probably are aligned in their interests. And just like Milton said, sometimes business users and noncommercial users align. So you could group them together as a registrant group or you couldn't. Sometimes it would work, and sometimes it wouldn't.

I don't think you are going to find a particularly perfect alignment.

So choose one, I think most of the time, regardless of what you choose as long as it has some relative degree of fairness, I think it will probably work out reasonably well, or don't choose one.

Why -- It's not entirely clear to me that the constituency structure actually makes ICANN particularly more productive than not having it.

So in such a scheme, you would probably have to not have any voting at all, because I imagine there are hundreds and hundreds of registrars that could show up, and probably overwhelm the entire process relative to everyone else.

But if you had some scheme where people could just show up and talk about the issues, it might be a little controlled chaos, or more chaos than control.

But it's not clear to me that the likelihood of getting good results out of a no constituency structure would be any worse than what we get today.

So that's what I have to say.

>>S. SUBBIAH: Hi, my name is Subbiah, I am from IDNS.net.

At the risk of complicating things a little bit more, I'm not sure this is exactly the place to bring it in but I think it relates. As one speaker said earlier, that, you know, maybe -- she said whether it's a solution or a problem. I think it's a real problem that probably could be solved, I mean when reworking things. And I think this problem has probably not been addressed. Because we ran actually into it -- a number of us ran into it, and I think Bruce also knows a little about it and he suggested that when the GNSO improvements come up, this may be a topic to bring up. Which is it's a little complex, so let me point out that the -- in the old days, we didn't have this many registrars. Ten, 20, 50. Now we have 800.

Even today, because of the difficulty, the relative difficulty of getting ICANN accredited for people from the third world, smaller countries, even now out of the 800 there's a lot -- the representation of the registrars is very low in many parts of the world.

The representation of registrars in the rest of the world, poorer countries, is still very low.

So because of historical reasons, because back then, to become a registrar was even more difficult than now. And so a lot -- The only root for many people was to become resellers, essentially.

So today we have maybe 100,000 resellers. And for the historical reasons of having signed up with a registrar years ago and having stayed with them, and because today it is not that easy still for some people in some parts of the world to become ICANN accredited, we still have a situation where out of 100,000 resellers, a lot of the resellers are in poorer countries. The mix is such that a lot of the resellers in those places are resellers.

Under the current constituency structure, the number of resellers recently tried to join an issue they were interested in coming from the third world, which is the IDN issue, and they wanted to join a working group put together by the GNSO not several months ago. But when this group of people, different people, tried to, resellers, primarily, they were the largest resellers, primarily, in certain countries. Large countries they don't have accredited registrars, but a very large reseller, which in terms of volume could have 100,000 names or more, but they were unable to join any of the registrar constituency or later they were unable to join -- they were sent to the business constituency, and the business constituency said that its rules did not justify this group to be part of the business constituency.

Therefore, these people were left in limbo. And I suppose at-large is a possibility. But they were left in a limbo.

And if ICANN is a representative democracy in the sense that you've got -- there is a threefold system you pointed out just now, which is the registries, the registrars, and the registrants. Somewhere in there, there's a lot of resellers, which, you know -- and they are the people with the closest connection to the registrants. And if they are not able to participate, then that becomes an issue.

Secondly, as a related point of view here, and of course we found in my own situation, we are a technology provider. And we were told also that we do not fit into the business constituency and so on and so forth. So that was not a reseller, but we also didn't meet criteria to be able to.

So in the end Bruce said bring in a an observer status for a lot of us. And in the meantime a lot of resellers got themselves ICANN accredited because they went out of their way to do it.

I think it's a real issue. Especially going forward if I want participation in the IDN context, especially around the world, poorer places, you are going to find that that community is not easily represented. And as a conclusion -- and it matches partly with the resellers in these parts of the world.

So there's been a discussion about a number of us from other parts of the world about starting potentially, quote-unquote, a language constituency, if possible, whether it's on the GNSO or just generally. But the idea that this would give -- because in the future, there's going to be a lot of participation, I think, from the IDN context from around the world.

That's all I want to say.

>>ROBERTO GAETANO: Can I -- So the last person in the queue, Sebastien.

I would like to keep the queue on this, and we'll close with Sebastien. We go to the next topic after him.

So we have four speakers now.

And of course, if -- I'm sorry, didn't see Izumi.

And if we have time at the end, I just don't want to monopolize the debate on the first issue only.

Just as a warning.

Philip.

>>PHILIP SHEPPARD: Roberto, thank you, it's Philip Sheppard from the business constituency.

I think as a general comment, and this is certainly the case of some institutions I've worked with, it is often the refuge of an organization running out of ideas to operate management by reorganization. And I think we need to be aware of that as a first principle.

One of my concerns in reading this particular recommendation in the LSE report was the recommendation was there, but the justification seemed to be rather narrow behind it.

And in looking for that, it seemed to be things like simplification of structure and maybe also linked to that a reduction in number of council and that seemed to be perceived as a good idea.

And I think on both of those rationale, we would have some problem, certainly in terms of the second one, a smaller number of people doing a lot of work is not necessarily a good idea.

And certainly the first one, the mere simplification didn't seem to be a sufficient objective compared to where we are now.

And if you look at the constituencies today, they are essentially birds of a feather group. So areas of organizations with common interests got together and that's how the constituencies led into being.

And that helps to some extent a cohesiveness in terms of the way that they would operate in a sort of hierarchical policy development structure like we are today.

Now, the more you dilute that, I think the greater problems you have in achieving that sort of internal constituency to make decision-making in given time periods, et cetera, like we do.

So I think there are some tensions in the concept of taking the small number of groups and putting them together just in terms of a structural issue, let alone anything else that's there.

Second comment that I would like to make, and this may help also your time keeping activities, is that I think it's very difficult to look at this question and delink it, I think, from your third question, which is going to be in terms of the voting structure. Because I think a lot of the LSE recommendations do naturally link together and I think they probably need to be looked at rather more holistically. So I would warn against trying -- I know we need a structure conversation, but I'd warn against being too linear in the way we consider the recommendations, because I think certainly talking to the LSE reviewers themselves, their view, I think, was that a number of these things did naturally link together, whereas some of the smaller recommendations could be handled separately.

>>JIM BASKIN: My name is Jim Baskin. I have just a couple of hopefully short comments.

First, I would like the board and this new working group that's to be authorized on Friday to take a long and hard look at who are truly their customers, to go back to a term that was used earlier today.

The -- I've heard a lot in the last year or so about certain constituencies needing to have more influence because they are the ones who fund ICANN today, that the money comes from the registrars or the registries, and, therefore, ICANN needs to listen to them more seriously -- maybe that's not quite an accurate way of putting it -- but that because the money is handed from one party to another in terms of the funding, that those parties are the ones that should have the most say.

Follow the money. Where did that money come from? The money comes from a step back. It comes from the users, the domain name holders who are paying all that money in, and then it gets to ICANN. So I'll stop on that. But I think that you need to look, in balancing the input that the board gets, the input that comes through the consensus process, you have to think truly about who are the users, the real customers of ICANN. Not the vendors, not the contractors who run ICANN -- the infrastructure for ICANN, but who are the real customers, and balance those needs, those customers' ideas better against those of the contractors.

Second, and a second one quickly, and it does kind of lead into the other parts of the questions. But the idea of having three constituencies and having a 75% consensus floor to me does seem like, as some others have said, a formula for no action at all.

I repeatedly asked the LSE people during some teleconferences that we had with them, that various constituencies had with them, I said, how does that work? How do you -- if any one constituency can create a situation where you don't have 75%, how do you get anywhere?

They said, "Oh, from our experience, it works, it works, because no one can get their way -- anyone can block anything, everyone has to cooperate."

Well, if the status quo is good enough for one of the three groups, then there's no way they want to cooperate to move away from the status quo.

Thank you.

>>BRUCE TONKIN: Just wanted to make a very quick observation, because I don't think this is perhaps clear in the issues with the structure. And that is, there's two different types of issues, I think, that ICANN can look at at a policy level. So I'm just looking at this in a general sense.

A lot of what the GNSO has done in terms of issues are policies that actually directly affect the existing operators, whether they are registry or registrar. And this is sort of what, in the registry/registrar agreements, is referred to as consensus policy.

And so that's very much a sort of self-regulatory concept in that you're making changes that affect their business today, you know. It's kind of -- they become mandatory changes.

And then you have other issues like, let's say, new gTLDs, where we're looking at, you know, changes in the future, you know, new things that ICANN could do, let's say.

And I think that's very different. And I don't -- in terms of how you might structure the representation to -- and how you might involve working groups, but I just wanted to pick up on what you were saying, is why would the registries/registrars want to have a say.

It's because the decision directly affects them. And that's the fundamental reason why that balance was there for that type of decision.

>>JIM BASKIN: I agree that -- I was not saying that the registries and registrars should not have any say. I was simply imploring you to consider how much say they should have, how much weight in the overall process. Should they be giving advice that is listened to? Everyone's giving advice, but I think we may have gotten to a point now where we have given too much weight to the contractors and not enough weight to the customers.

I see a difference that the -- So I'm not saying ignore those vendors who are a big part of the process and have to stay in business, but just make sure you adequately consider the other customers -- the customers that are feeding the money up, so to speak.

>>RITA RODIN: I just have one short comment to that. I think should you feel assured that the board understands, it doesn't look at customers, per se, but it has a number of constituencies that it has to deal with and address. And to take Bruce's point, I think we're just, when you look at the business impact that changes here could have on registrars that also factors into our decision. But the board is clearly aware of all the diverse constituencies as they are customers, if you will.

>>ELLIOT NOSS: I wanted to speak about a separate constituency that I hadn't heard discussed yet. Before I do, I would say I always struggle when I hear that registrants' interests and registrars' interests are at odds. I think that the history of the GNSO has shown that with the exception of very specifically intellectual property interests, that registrars tend to be the ones who are advocating for registrants' interests. I think the record speaks for itself in that regard.

The constituency that I would like to discuss and that I haven't seen much discussion about, at least not to my satisfaction in the report, is the ISP constituency.

I would strongly suggest -- and I'll try and back that up with some empirical data -- that the ISP constituency as originally conceived of and formed is today an anachronism. Like many things in the Internet, you know, the market and the construct of the market that existed in 1997/98, you know, maybe through early 1999, at the time of the formation of the original DNSO, has changed materially. So what was then primarily a market where the suppliers of Internet services, access and other services, were called ISPs, has dramatically changed.

Today you have network operators who are primarily in data centers. You have other companies more often referred to as ISPs that provide connectivity, either broadband or dialup. You have a third class that virtually didn't exist in '97/'98 that are Web hosting companies.

You know, at Tucows, where our primary business is and has always been supplying these companies, we've seen a shift over this time period, where in 1998, you know, our mirrors and who the folks we were supplying were almost without exception ISPs, in quotes. Today, the single largest segment of our customer base is Web hosting companies. Still lots of ISPs, but more who self-identify calling themselves Web hosting companies, with the evolution of broadband and the maturity of the network itself and the network markets. A lot of companies have moved out of pure connectivity and into services.

I think that the point that Subbiah was making earlier is complete resonant with what I want you all to be thinking of very hard, which is, I think, that there is a more appropriate concept or umbrella that should be thought of as what is now the ISPC, and that should be companies that are suppliers of Internet services, that touch end users directly, and that have to work up the chain with this artificial construct that we call registrar.

One of the points that was raised earlier in discussion today -- I don't know -- I think, Vint, you might have been the only one that I see on the panel that was here for it -- was Steve Crocker pointed out that registrar is an artificial construct. And it is.

The company -- That's something that we, inside of ICANN, have created.

The people that supply Internet services are a broad range of companies, and they may or may not be registrars. But registrar is a concept that we have created.

And, you know, I would really urge you in this reform process to look at that current definition and look at the charter inside the ISPC. In my view, it has been much too narrowly proscribed. And, unfortunately, what that's done is excluded a lot of businesses who sometimes are called resellers, sometimes are called Web hosting companies, but who are critical, critical citizens in this whole -- in this whole set of subject matter that we deal with that cannot find a right place to get represented.

So I'd please like you all to think about that.

>>PETER DENGATE THRUSH: Elliot, just to be clear, in my original categorization of registries and registrants, if we were to include everybody between that you've just been talking about as registrars, and that would include resellers and Web -- would you be -- that's what you are -- that's the point that you're trying to make, that between the registry and the registrant, there's a host of service providers?

>>ELLIOT NOSS: No, between the registrar and the registrant, there's a host of service providers.

So the difference is that if you're a registrar, you know, you may also supply other Internet services, but you're a creature of contract that contracts with ICANN directly. If you're a Web hosting company, an ISP, you're what Subbiah called a reseller, which is a term that we at Tucows actually made up in 2000 because we didn't know what else to call our customers.

Then you're somebody who has to deal with, you know, these contracted parties on one side and deal with end users on the other.

And, you know, one of the ways that the market has evolved materially is that today the telcos and cable cos who today predominate inside the ISPC are virtually invisible when it comes to the provision of Internet services.

They -- you know, their interests have tended to move more towards protecting the brand. I mean, I had a really both surreal and epiphanous moment when I was a couple of years ago presenting a new gTLD proposal to ISPC, and when I finished, all of the questions I was asked was, "How will we protect our brands?"

I've been dealing with ISPs since 1995, and, you know, I visit them all over the world. We have customers in over 100 countries. It was the first time in that 12 years that I had heard an ISP worried about brand. They were worried about whether they were going to wear sock that is day or not.

I mean, it was shocking to me, and really, really symptomatic of how misrepresentative, you know, that label had become.

>>ROBERTO GAETANO: Thank you.

>>IZUMI AIZU: Hello. My name is Izumi Aizu.

As some of you may know, that we are working very hard for the regional at-large body or RALO. And with the last meeting in Bali, Asia-Pacific, 13 ALSs agreed to form our AP RALO. We are waiting for the board for our approval of the MOU.

With that, we had an election of officers. And instead of the interim member of at-large, now I became one-year term ALAC member. So -- but I'm not representing ALAC, but just speaking for myself about this issue.

My first reaction to the LSE recommendation especially on the three groupings of the civil society are somewhat puzzled. On the one hand, it's a good idea to recognize the civil society as a distinct stakeholder within ICANN so we can claim and actually demonstrate the existence and workings of the civil society as a multistakeholder organization.

But on the other hand, we are a little bit puzzled, because one is that ALAC, or At Large, has been sort of designated or working to sort of convey, if not represent, the voice of the individual users. And in contrast to the NCUC, who is, in theory at least, is representing the organization of the civil society, the NGOs, NPOs, and otherwise.

So are we going to merge into one? Or could we just stay as ALAC and then have another element, reorganize within the GNSO? That's an interesting question, which leads naturally to the question of how this GNSO review is related to the next review, which is ALAC or At Large.

Of course, on the one hand, if you wait too much, then you never reach the closure, because there's another review for the board and other ICANN whole review. So we cannot wait too long.

But at the same time, if we miss that ALAC review, then you make a solo decision about the GNSO, which may not be effective as a whole organization.

In the strategic plan, the first round, it listed how many issues per each constituency or the SOs or committee should cover. And if you remember, ALAC, or At Large, has had -- it was given 11 or 12 issue areas, the largest of all. I think it was one point more than the GAC even.

So, in a way, we are supposedly involved or have some interest in all areas of ICANN, because it's natural the individual users are affected by ICANN's divisions A, B, or C.

So if you were just to merge or put into a constituency within the GNSO, then this wider coverage will be gone.

Of course, I don't think the intent of this LSE review was as such.

But at least we need to consider both options. Both, I mean is where should ALAC go in terms of civil society participation or individual users' participation in ICANN's broad mission.

I don't want to record all the history of At Large. But for some of those who were not there in the beginning, there was an interesting debate about whether ICANN should be a membership organization or not. And we were given to implement the membership after it was recognized by the Department of Commerce of the U.S.

So we spent -- and I was a member together with (saying name) and others about a year to design the membership, which was called "At Large" and went to the election in 2000, and then to the review after that, 2001, and then the reform in 2002.

And so there were certain ups and downs of how to position the At Large or the users.

In the original design, we didn't make a distinction between the individual users or organizational users, but just put them all into the users. Pragmatically, we agreed that it was so difficult to actually examine each applicant to the membership whether he is just an individual user or he belongs to the IBM or AT&T, but also he's an individual user. So that's one way to put it.

But, anyway, that time, we designed to have nine board members out of 18 from At Large. And we went the election of selecting five or electing five, and review, which concluded almost a consensus. But then the reform came that says there should be no At Large, no membership in the proposal. But we ended up with the advisory committee as ALAC and started to form this ALS regional process, which took about four years, from 2003 to 2007. So that is another process to be reviewed.

So -- And I think what is not written there in the LSE report is regional element, which, of course, the GNSO doesn't have per se, while At Large have.

On the one hand, it is -- if it functions well, then it will be a good channel to bring all the different regions' interests and issues -- IDN is one thing, digital divide is another, we are in a global community. But at the same time, it is a certain kind of not necessarily barrier, but it's a big hurdle that we have a lot of extra work in the region. So that is not really considered in LSE review.

>>PETER DENGATE THRUSH: Izumi, I don't want to interrupt you, but can I just interrupt you by assuring you that we are not going to fuse the At Large with the GNSO review, that there is a separate review coming for At Large and all the ALAC structures. Okay?

>>IZUMI AIZU: But how they interact is sort of my natural question, or your question, too, I think. You cannot make a discrete decision of GNSO/LSE review per se, at least for the process; right?

>>ROBERTO GAETANO: Izumi, I think we have got the point.

Now, in the interest of having everybody speaking, if you can -- if you --

>>IZUMI AIZU: I stop.

>>ROBERTO GAETANO: No, if you have other issues --

>>IZUMI AIZU: No, this is end of story.

>>ROBERTO GAETANO: Just want to reassure you that we got the message.

>>IZUMI AIZU: All right. Thank you.

>>SEBASTIEN BACHOLLET: Hello. Sebastien Bachollet from ISOC France.

Just to say a few words where I stand.

I used to be member from my previous job as a member of the business constituency, and now with my current responsibility, I am with the ALAC structure.

And I think what is interesting with the ALAC structure is that it's a cross-structure. It's not deal with one of the silo of the organization.

>>PETER DENGATE THRUSH: Sorry to interrupt you again, but we are not talking about ALAC. Are you going to talk about the GNSO review?

>>SEBASTIEN BACHOLLET: No, sorry. You talk about ALAC because you say that you will make the fusion of the --

>>PETER DENGATE THRUSH: No, I said we're not going to make a fusion.

>>SEBASTIEN BACHOLLET: And some ALS --

>>PETER DENGATE THRUSH: I said we are not going to make a fusion.

>>SEBASTIEN BACHOLLET: You said not, but in the presentation you said it would be. It was in your presentation. I'm sorry, I am speaking about what you say at the beginning.

But if I don't understood, maybe you make clear for everybody.

>>ROBERTO GAETANO: No, Sebastien. Just a second to set the record straight.

It is, in fact, one of the issues raised in the LSE report. But I think that one other thing has to be said, is that that is the -- one thing is the LSE report and one other thing is what the board is going to decide after having listened to all the comments of the community and of the different parties.

So what is written in the LSE report is not as of today the position neither of the governance committee nor of the board.

>>SEBASTIEN BACHOLLET: For sure. Just because you ask for comments and I take what you say at the beginning. Sorry if it's outside of your scope, but maybe I misunderstood something.

Just I want to add one point.

One of the difficulty today to -- in the organization of ICANN, its work in silo. There's a lot of people who say that.

If you are to change the organization of the GNSO with just three constituencies, that means that you will create a silo within the silo. And then it will be harder and harder to work.

And I don't know why everybody needs to kill one constituency to have one new one.

Today, in the bylaws, you can create new constituency within the GNSO. Then let's try to create new ones before to kill, if we need to kill one in the future.

We don't need to have to kill someone to be alive ourselves, fortunately.

Thank you very much.

>>ROBERTO GAETANO: Thank you.

Before going to the next area, I would like just to notify to the assembly that we have received one contribution from the outside that is four pages long. And therefore, I'm not going to read it, but will be part of the public record and the input to the work group.

I just want to let you know what it is about. And it is no continuing purpose and assessment of the GNSO where the basic question is whether the GNSO is still meaningful.

So this will be put on the record for the working group.

And then Alejandro wanted to make a final comment, and then we go to the next group.

>>ALEJANDRO PISANTY: Thank you. This is Alejandro Pisanty.

I would like to quickly respond and maybe elicit further cycles of response outside this meeting so we can concentrate on the following issues, Roberto. To Dr. Subbiah, this is this idea of an IDN or language-based constituency which has already been discussed a little bit in the part of the community.

I think that if this is going to be further discussed, what we have to find out is what is in common among IDNs or language-based approaches to of domain names that has more in common and creates more cohesion than the gTLDs or ccTLDs separately. There's a task force, there's joint work going on between these two groups and probably what we have to do is intensify that and make more visible that joint work on IDNs in order to identify whether actually a new structure is required for a better functioning.

The same happens with technology providers. Most of the technology that goes into the DNS is represented true the technical liaison group and the organizations there, the IETF, ETSI and so forth, in setting standards, not in providing technology that's providing services based on a specific technology. And what happens here, of course, is if you are finding an unfair barrier to entry in the business community, which seems to be one of the complaints, that will be an element for the review as it goes further, to find out whether some of the constituencies are limited in representation, as has been claimed, and whether there are explicit or implicit barriers to participation in those that can be removed by modifications in the bylaws and so forth.

Or whether it's just the lack of a good reading of the bylaws which invite -- not only allow but invite the formation of new constituencies and establish very specific steps by which an initiative can be taken to start a new constituency within the GNSO and identify whether it is large, well differentiated, well supported enough and so forth to be established.

To Elliot Noss's question, or affirmation, statement, that there is a host of service providers between the registrars and the registrants. Again, one wonders why this group has not come together, come up with a proposal for a constituency. And an idea of whether it would be on the supply or the purchasing side in the domain name market to see how it would affect weighted voting and so forth.

It doesn't seem that there's been enough cohesion in this group to put forward this proposal. And I will be glad to hear more about it.

And finally, to Izumi, I think that there's so much water under the bridge that has already run with the discussion of what specifically the ALAC does, that the NCUC doesn't do, what part commercial and noncommercial individual providers -- I mean domain name registrants have and so forth, that this would seem like going back to '96, the year before the foundation of ICANN, to restart this whole discussion.

>>ROBERTO GAETANO: Thank you. Elliot, please understand me. I don't want to start -- If we have time at the end, we will be able to respond.

I would like also to avoid a lot of confrontation here.

We are here to listen.

We go to the second group.

The second group is about the policy development process.

I think by and large, it might be my misinterpretation of the LSE report, but it seems to me that what they are proposing with the different recommendation on this, in this area, is to allow for more flexibility and efficiency in the PDP, by greater use of task forces that has been -- have been fairly flexible way to use the resources.

And also, by now, not slight jacketing the work of the GNSO and of the task forces with provisions that are hard quoted in the bylaws.

This is my reading of the GNSO review, but if you have different readings, you are welcome. I mean, you are welcome anyway to whatever your reading is to provide your input.

I would only recommend that you keep your comments short. And I see already Jordyn in line on the microphone.

And if we keep the comments short so we can have more contributors.

Thank you.

>>JORDYN BUCHANAN: Thanks. I have chaired two task forces under the current PDP, so it's going to be harder for me to keep this super short than maybe for some people but I will do my best.

I do want to point to a couple of the flaws in the current PDP, because wring it's instructive to sort of see what's gone wrong when you think about how you might fix it.

I think two of the constraints that are severely broken in the current bylaws are the time lines actually don't make any sense at all. And so to the extent we're going to have time lines, they need to be significantly rethought.

I think Chuck made a really good point earlier, though, which is different issues are going to necessarily require a different amount of thoughtfulness and time to work through. And having a one-size-fits-all approach to time lines in the bylaws at all is probably a mistake. It seems like more of a issue that the GNSO should probably work through on a case-by-case basis.

A real significant problem I think in the current process is either it imagines a different -- entirely different approach to what the staff is doing, or something doesn't make any sense at all.

But essentially, the -- I think it imagines a situation in which the staff not only helps guide the process, but essentially leads much of the process, in that it imagines that the staff puts together I think an issues report that not only includes a description of the issues but also a lot of solutions as well. And then the various members of the task force just get together and say, well, you know, there's thing that the staff said that are good or bad or whatever and quickly dispose of that issue along with some discussion of how it might affect their constituency.

So I don't know if people think that's a good idea or not. It's certainly not what happens today, and to the extent that we are going to have the notion of a staff manager doing stuff like after the public comment period, all we do is have the staff manager copy down the public comments and send the report back to the council, that seems to imply that the task force representatives don't interact with the public comments at all on their own and so on.

So I think that a much more active role for the staff is imagined in the bylaws today than actually happens. And so there's a serious disconnect with making it actually work.

So a couple of recommendations, then, to make it work better.

The first is to -- and this has been talked about a lot, but a more diverse set of participants in the task force process would be -- would work, I think, extraordinarily well.

I think creating mechanisms for nonparticipants in the day-to-day workings of the GNSO, to come along and provide expertise on particular areas or even provide manpower in particular areas that they happen to be interested in would be really helpful.

And sometimes these will be people that are not represented by any particular constituency.

Today we have the opportunity to -- the council can theoretically appoint an expert, but it happens relatively rarely, probably because the council doesn't know a lot of these people and in some cases they may not actually be experts. They may just be interested in a particular topic.

So I think being able to draw new blood in from outside would be extraordinarily helpful.

And the last point I'll make is that, and a couple of people alluded to this already. I don't know if it was Jim Baskin or someone else. In the status quo, if you like the status quo, and there's a policy development process underway, your best strategy to -- achieve your objective is to essentially gum up the process as much as possible. And that's what happens a lot. And that's why the policy development process is so slow and so unproductive in many cases. And I think the reason it why that's a problem, and this is something Susan Crawford alluded to earlier today on a different topic, is there's a bunch of stuff that we call policy and has the same merit as stuff that actually has gone through the policy development process that actually was just made as a bootstrapping exercise when ICANN was first created.

So for example, there's this thing called the Registrar Accreditation Policy and it contains a bunch of rules about all sorts of things like the WHOIS, and how registrars operate things like this, whether or not ICANN can sanction registrars. These are all very important topics, and in order to change the behavior of any of those things we need to go through an entire PDP and get two-thirds consensus around any of those points. And it makes it -- anyone that likes anything that's happening today with regards to any of those points just has to sort of stonewall and they achieve their objective. So I think it would be much better, I really encourage the board to strongly look at the idea of saying anything that hasn't gone through the policy development process and actually achieved consensus should no longer be treated as such. And essentially sunset a set of policies that came into existence at the same time of ICANN without having a real policy development process around it.

That way, every time we are having a policy development process, the people that are happy with the status quo can't win just by blocking the process.

Thanks.

>>ROBERTO GAETANO: Thank you. Thank you.

Philip.

We are short of time, but doubling the speed of your speech doesn't help very much.

>>PHILIP SHEPPARD: Philip Sheppard, business constituency.

As one of those parties who had participated in the input to the policy development process that was subsequently adopted by the Evolution Reform Committee, I was one of those who was slightly alarmed to see that was subsequently enshrined in the ICANN bylaws.

Certainly our view in writing it at the time had been these were sort of guidelines that would need be revised on the basis of experience.

And the fact that a decision was taken to enshrine them in the bylaws just slowed down the flexibility for that process. That's not to say we strongly support the idea that of taking them out there, and have guidelines that are more flexible, makeing it more practical.

The second point is in relation to I think probably even the short time period that's elapsed between the LSE during their review and today, things have changed and moved on.

We have more specific task forces that were just council members and/or constituency members represented in some of our groups then. More recently, we've expanded that to wider groups, we call them working groups, not task forces. That was a recommendation that the LSE was making. We are already, to some extent, doing that.

Also, we have seen very much on the recent sets of policy development the expertise of our staff support (inaudible) for so long now being embedded, becoming thoroughly professional in terms of the output that it is doing, and I think very much helping also with the mechanics and the lengthy policy processes.

So I think there is on this particular question -- and the third point also is increase physical meetings. We have already started this idea of interim meetings of specific committees in between the normal three ICANN meetings.

So I think some of those three recommendations from the -- from the LSE review, two of those are already happening. The third one is getting those guidelines out of the bylaws and allow us to be more flexible in the way they are developed.

>>ROBERTO GAETANO: I have Peter who has questions.

At the end of the line is Milton, are you the last one?

>>MILTON MUELLER: I hope so.

>>PETER DENGATE THRUSH: Quick question.

Philip, thank you for that. Two questions.

Would allowing the -- talking about embedding the time lines in the bylaws, and I can quite understand why the bylaws shouldn't contain that level of detail ideally, it should be a principle that there be a process and it should have some milestones. But my question is, if the bylaws were to be amended to allow the GNSO to self-select a set of time lines for each issue, would that itself just not allow another opportunity for someone who wanted to gum up the works by delaying that decision and then -- and by creating a completely unsatisfactory set of conditions?

>>PHILIP SHEPPARD: Potentially, yes. I mean, that danger clearly is always there. But I think it just makes more sense. You look at an issue, you judge its size. And therefore, the beginning of the process you can say our estimate of our completion is likely to be this.

And it's -- And therefore, you have a -- you create a continual set of relative and relevant measures rather than a one size fits all, which will naturally lead to failure and has in the majority of cases.

>>PETER DENGATE THRUSH: So the risks of gumming are actually worth running because of the advantages?

>>PHILIP SHEPPARD: I think so, yes, because then you can set your measures of success at the beginning based on the nature of the issue.

>>PETER DENGATE THRUSH: And the second question was really, are these PDPs getting enough staff resource? And if the board was to facilitate more resources for PDPs, would that make much of a difference?

>>PHILIP SHEPPARD: We look at the GNSO policy development as being one of the more important executive arms of the organization and you compare the number of staff devoted to that compared to the number of staff devoted to other areas within ICANN, I think that may be something the board should consider.

>>CHUCK GOMES: Chuck Gomes.

Peter, if I could respond to that last question that you asked Philip.

First of all, having just chaired the reserved name working group, one of the things I learned, it is really necessary, and I have communicated it to Denise and she has responded favorably, is that it would be really helpful to have administrative support.

There's been a good increase in higher level policy support with policy experts, people that can do research, et cetera. But bottom line, there's a lot of administrative work in terms of document preparation, et cetera, that would be greatly helpful to task forces and work groups. And I have communicated that and Denise is well aware of that and I'm sure that they will plan for that. An area that will really help. Because when you are leading a task force or a working group, it is very time consuming. And if you have to do the administrative tasks as well and you already have a full-time job, it's really tough.

And I'll be brief on the rest.

Make greater use of task forces. To me, that's a no brainer. It does not scale to have councillors doing all the policy development work. It just doesn't.

You can't schedule things, because they are conflicting with everything they are doing. They don't have enough time, they are busy.

So I think that one is just clear as a bell that that should be the case.

In fact, I think this is one area where the ICANN bylaws and the PDP process and in particular the responsibilities spelled out for the council are to manage the policy development process. Not to develop policy.

Now, we can argue whether it was meant to be that precise, but I think that's really good. And as registries, we proposed maybe two, three council members on task force working group. And get other people involved. And I totally agree with Philip. We're seeing that.

In the past few months, we're seeing a lot more people getting involved in working groups and task forces. And I think it's working.

>>ROBERTO GAETANO: Thank you.

Mark.

>>MARK McFADDEN: I want to support this as well, and kind of echo a few of the things that Chuck just said.

The idea of having task forces and working groups do more of the work is something that is already happening.

I was on the IDN working group. That work is effective. But I do want to warn the committee about one thing, and if they take on this suggestion from the LSE, which I hope you do, that one of the things that is needed here is when there are task forces and working groups, there have to be fairly crisp definitions of who are part of those, the membership of those working groups and task forces.

If the task forces and working groups get to be large, and a lot of people will have self-interest to join working groups simply to be observers of the activity, to watch what's happening so that they stay in touch, not necessarily contributing any energy to the group but actually just sort of lurking there, being a passive participant, that is something that is not helpful at all to those working groups, especially as they get very large, as they move out of the teens into the 20 and 30 members. It's a very, very difficult thing to do.

One of the things that happens in those cases, you need a very, very skilled working group or task force chair. And in our community we have those, but we have a limited supply of them. And as we take greater advantage of this, what I encourage the committee to do is make sure that the definition of who is part of those task forces, who is part of those working groups is defined pretty precisely.

>>PETER DENGATE THRUSH: Mark, that's not really a board issue, is it? That's a housekeeping issue for the GNSO, how it sets up and how it populates its committees? You surely don't want us telling you how to set up task forces or working groups.

>>MARK McFADDEN: Peter, what I want is a crisp definition of what the membership can be like. I'm not asking that the board define who the members are, but say categories of members; right?

If I'm going to push down, and I think this is a good idea, Peter, if I'm going to push down policy development into smaller, more mobile, nimble groups, I want a good definition of who can be in those groups under certain circumstances.

The council, I think, is the one that does the population. But the board is the one that does the definition.

>>MILTON MUELLER: Keep this brief. I wanted to echo Jordyn's comments about the importance of leveraging the staff. Would you consider that a board issue? Something you want to hear about?

>>PETER DENGATE THRUSH: Absolutely, yeah.

>>MILTON MUELLER: I think this is a very -- potentially very productive move to make, but also a risky one in certain respects.

The up side is that the staff will know the process, and so I have always been a little bit concerned about the fact that task forces and PDPs are chaired by people who are in a constituency, and then you always have a political issue as to who is biased. So that maybe you should be looking at staff actually managing the process.

The potential down side is that those people are not accountable to anybody other than the board, who may not know what's going on. And we don't want them to become a magnet to for lobbyists who then try to short circuit the process by trying to get staff to do things that are biased in a particular way. Or we don't want staff substituting their own positions for those of the constituency.

I think we learned in the TLD PDP that you really need the staff to push on issues that are contentious, and that if you don't have an active staff person pushing the process forward, you can really spend enormous amounts of time.

And so there's an up side here and a down side to be aware of. You need to think of some sort of accountability mechanism that would prevent autonomous decisions being made by a middle level bureaucracy, which is the way a lot of international organizations have evolved, by the way. This is not something that's new.

The staff in a lot of these organizations are generally a lot more powerful than ICANN's, but can frequently become an independent force unto itself.

So....

>>ROBERTO GAETANO: Thank you.

We have a final comment by Bruce, and then we move to the third and last part.

>>BRUCE TONKIN: Just a very quick comment. When I hear the discussions about how to structure the constituencies at the council, it sounds very similar to the new gTLD discussion, which is basically, we want from the top down to define what the categories are and whether it's registries and registrars and ISPs or whether it's registrants and registrars or whatever. So you're defining categories from the top and then defining who gets into each of those categories.

Or are you approaching it more from a bottom-up point of view.

And certainly -- and I'm just sort of speaking factual rather than trying to direct you -- that the bylaws at the moment assume that there is a bottom-up process and new constituencies can be formed.

So one of the questions is, why hasn't that happened?

And I just want to reiterate also what Philip and Chuck said, that if you -- it's worth the Board Governance Committee looking at what's happened in the new gTLD work perhaps as a case study, because that is evolving. We're trying to take into account new approaches to doing things. And one of the things that has arisen is that as we started to go into some quite tightly defined topics -- one was IDN new gTLDs, and the other was, most recently, sort of protecting the legal rights of registrants -- they have attracted totally different people into the process, and new people, which has been great. But we didn't get those new people through the constituency structure; we got those people by, you know, opening those groups up for more open participation.

So I think it's interesting to just see what happened in recent dynamics.

>>ROBERTO GAETANO: So I realize that the bar is open, so I would understand if somebody will go to the drinks. However, I would like to have a short third part to address also the voting and the representation.

As has been commented already, this is partly overlapping with structural changes. A little bit we have addressed already the question of the threshold for consensus. There's also one other issue, that is make the council smaller, which has also to do with structural changes and the reduction of the number of constituencies.

But I would like to also throw one additional comment on this that came at the board/GNSO dinner yesterday. There was one provocative question, which is, do we need to vote at all? And if so, when?

Throwing this as if you didn't have enough elements for contention on this and keep you away from the drinks for a little bit more.

So who wants to open the fire on this?

I see somebody.

>>VINT CERF: Vint Cerf again, ICANN board chair.

Just to be very brief, my perspective as a member of the board is that many of the issues that arise are sufficiently controversial that achieving consensus is quite difficult. It serves the board well to have input which expresses the real diversity of opinions and positions on some of these topics. And to force an attempt at consensus, which might take forever or not actually ever be achieved, may be less useful than having a very clear expression of what the different views actually are.

Although that leaves the board with the problem of trying to come to conclusions on some matters, it seems to me better to try to make progress in the context of a well-informed set of diverse views than it is to have the process gummed up, to use the technical term, and not ever reach, for example, a board decision.

So I'd like to suggest two thoughts. One of them is that any parsing of the communities of interest in the Internet is bound to be wrong. That is to say that there is no pure parse of people's interests.

More critically, depending on what the topic is, you might actually find the interests breaking up in different ways and having different groups of people aligned on one matter and not aligned on another.

So the focus of attention on trying to get a precise parsing and categorization of membership seems to me less important than getting clarity with regard to the diversity of views and their rationales.

So I'd like to suggest that getting consensus, if it's there, is really helpful. But in the case that it isn't there, I would, as a board member, rather get some results, even though it does not represent a consensus, than to get nothing.

>>ROBERTO GAETANO: Thank you.

Susan.

>>SUSAN CRAWFORD: Just very briefly, the role of the board actually in this construct is to judge whether consensus exists. And a failure to reach consensus is not a failure; it just means we don't need a global rule in a particular instance.

The idea is that everything not prohibited should be permitted. There should be lots of local diversity in how registries do their jobs.

So I just want to be clear about the model that ICANN represents. Actually, deliberation, consensus-reaching is the basis for the rules, not the board's decision about any particular rule.

We are actually charged with trying to figure out whether the global Internet community wants a particular rule in place or not. And to place the burden of choosing from among positions on the board itself puts a tremendous pressure on the board that may not be appropriate.

Thanks.

>>AVRI DORIA: Avri Doria. Speaking on the consensus, one of the problems with trying to reach consensus, and I think this is going along with what Vint said somewhat, is not only the gumming up of the works, but you end up having to find the lowest common denominator. Now, sometimes that's a good thing, and sometimes the board may be the one that looks at the positions when consensus wasn't reached and finds that lowest common denominator.

But I think that in somewhat of trying to reach consensus within GNSO is a good attempt.

If it can't be reached, explaining why it wasn't reached, where the differences of opinion are, and moving on. The other problem with trying to reach consensus is, you can add years to the process while moving down to that lowest common denominator.

So I think it's a great goal and I think it's something that we should strive for. But when it can't happen or in those places where it can't happen, some of the work in the task force that I've been working on a lot is -- on some things, there is consensus; on some things there isn't and never could be, given the structure of the constituencies as it exists. Maybe I'm a little pessimistic. Maybe it could happen in a couple of years that we would reduce things to a low enough level that we could agree.

So I think allowing a vote and allowing a vote with a strong majority, not even necessarily the two-thirds -- or I guess that's a three-quarters majority -- is viable for concluding a process and moving on.

>>ROBERTO GAETANO: Thank you. Jordyn.

>>JORDYN BUCHANAN: Thanks. I'll be brief. I just wanted to reiterate something that I think Chuck had mentioned earlier.

And that is, there's a lot of talk about problems with weighted voting.

I don't know that there's a lot of data to support that view. Where are the policy outcomes that we're unhappy with as a result of weighted voting?

And I think that -- I understand the sort of knee-jerk reaction against it. And in some cases, it's probably frustrating to people who are on the opposite side of a weighted vote. But that doesn't necessarily mean that the policy outcomes that are resulting are bad.

So I think you should very carefully look for data supporting that position before you accept the conclusion that, therefore, we need to get rid of it.

>>ROBERTO GAETANO: Thank you.

Is there anybody after Philip that wants to speak on this issue?

Okay, thank you.

>>PHILIP SHEPPARD: Thank you. Philip Sheppard.

Just, again, two quick comments.

The first is, I think, again a decoupling suggestion. The idea of the weighted voting and the threshold were part of one recommendation from the LSE. I think they have different arguments behind each, and I wouldn't necessarily see they need to run together. We'd have more concern, certainly, with the raising the bar for, like, 75%. I think, again, that may factor into a reorganizational type concept. So I would recommend decoupling those.

On second, on the weighted voting, I am somebody for my sins who has experienced both of the previous organization as a councillor and the new organization, the names council and now the GNSO Council.

And certainly the characterization, I found, of operating under the new structure, in contrast to the aspiration of change, which was to try to re- -- which was premised on the basis of an assumption of too much conflict and trying to resolve it, has actually been the reverse. Because I remember at the time in looking at some of the -- the figures there, the number of times, actually, on the names council in the previous balanced voting where we had direct conflict was relatively small. And more recently, there is a perception, certainly, and a feeling of perhaps entrenched conflict and concern and (inaudible) status that was never there before.

So I think that change has itself structurally led to greater perceptions of role and right, and just in terms of the way that it functions at the personal level.

And I think the second comment I would make is one to do with trust. My first comment is to do with public trust. And ICANN's public trust role, I think, is ill served by the suppliers' side of that, the contractual parties being seen to give greater weight in the way that they can help write the contents of their own contract, compared to the users' public out there. And that leads on to the question, the other trust question, which is antitrust, and just how close does that come to that second concept.

>>ROBERTO GAETANO: Thank you very much.

So that concludes the third part.

I have just a last question. Is there anything that you -- any of you would like to add and that has to be stated on the public record and is not best resolved in private discussion at the bar?

One, two, three. Gone.

Okay. There's really no conclusion. This was only a phase in which we were here to gather input. And rest assured that the input that you have given is going to be on the public record and is going to be taken into account by the task force that, as we have said in the beginning, is going to be made with a board resolution on Friday.

Thank you all.

[ Applause ]

(6:15 p.m.)