Attendees: ALSC Members Carl Bildt, Pierre Dandjinou, Esther Dyson, Ching-Yi Liu, Tom Niles, Oscar Robles, and Pindar Wong, and ALSC Executive Director, Denise Michel.

Participants: Danny Younger, General Assembly DNSO; Lars-Odin Mellemseter, Global Name Registry; Dora Gerber, ISOC/Geneva; Chun Eung Hwi, PeaceNet; Wilfried Wober, Address Council; Roberto Gaetano; Vittorio Bertola, .IT Naming Authority; Rob Courtney, CDT; Christopher Chiu, ACLU/Internet Democracy Project; Alan Levin, Afridns.org; Vincent Easton, RgNames; YJ Park, MINC; Izumi Aizu, ANR/ELOCOM; Adam Peake, GLOCOM; Masanobu Katoh, Fujitsu; Kwasi Boakye-Akyeampong, CPSR; Penny Karas, NeuLevel; Joop Teernstra, IDNO; John Klensin, AT&T, IAB; Wolf Kleinwaechter, Univ. Aarhus, ICC; S. Verhulst, Markle Foundation; Helmut Schink, ICANN Board; Joe Sims, JDRP; Hans Kraaijenbrink, ICANN Board; Andy Sullivan, Reuters; Rogue Gajliow, Antel; Cathy Lee, RgNames; Don Simon, Common Cause; Mary Hewitt, ICANN; Neeran Saraf, Saraf Software Solutions; Jorge Planto, Cabase-Argentina; S.H. Kyong, ICANN Board; Linda S. Wilson, ICANN Board; Claudia Franco, Zacarias and Fernandez; Mamoufani Zakari, Direction Inforna Tique- Nigeria;

Agenda: The outreach meeting focuses on the ALSC's draft recommendations on the concept, structure and processes relating to an "At-Large" membership. The draft report was posted on our website on August 27 for public discussion and comment. As indicated in the report, there are numerous issues on which further discussion and elaboration is needed. Based on the input we receive and further work, the ALSC is committed to presenting final recommendations to the ICANN Board prior to the third annual ICANN meeting on November 15, 2001.

Scribes Notes:

Carl Bildt: (see Uruguay Sept. 7 slides at http://www.atlargestudy.org/meetinginfo.shtml) Welcoming/opening remarks, including creation and purpose of ALSC and issuance of the ALSC's draft report. Comments encouraged by Oct. 26 and final report will be issued prior to ICANN's annual meeting on Nov. 15. Our draft recommendations are to create an ALSO, define members as IDNH and hold direct elections for regional At-Large Council members and 6 At-Large Directors. Our intention is to build a policy and decision-making structure within ICANN that can foster understanding and accommodation among stakeholders/users. We believe this is workable and the most promising compromise. We're not saying this will work for the next 500 years, but we do think it is workable for the next 6 years. At-Large has been unresolved for the last 3+ years and that can't go on. We were chartered to seek a consensus on At-Large - which has not been defined. I don't think we'll achieve a solution that everyone supports, but it is a question of finding a concept that is acceptable to a majority and is workable and stable for a couple years. That's how I would define consensus. We've had 6 months of lots of different outreach activities across the world - online and offline. We've found a number of committed individuals but we found that it has not aroused interest on a broad scale. Personally, I have not dealt with an issue that has had so little interest/engagement. Of course, the fact that there are some committed, interested individuals out there is an important asset for Icann and needs to be developed further. We find great value in ICANN's bottom-up process and in the involvement of users. We don't discuss this in terms of developing a global democracy, because I think that's a delusion and that's dangerous to ICANN and concept democracy. But we do see great interest in At-Large to ensure transparency of ICANN and other ways to provide input in ICANN - provide a voice from all parts of the world. It seems to the same people showing up at each ICANN meeting. We'd like to opportunity for additional, new involvement. We considered how to define At-Large members - email based sounds politically attractive. But the ALSC could not find a way that could give us a purely online system. Without it being all-online is problematic - it disenfranchises people in various parts of the world, and the financial burden of scaling this type of system. The capture problem is highlighted in our report and in other reports. Not a workable option with present technologies. We think private v. public interests is not the way to consider ICANN's Board. We see public portion having private interests and in private portion public interests are represented. We think its more productive to look at ICANN has comprised of developers, providers, users - as philosophical approach its more useful. Then we recommend one-third of the Board elected by At-Large (of current 19 member Board).

We recommend domain names as a practical way of identifying people and these people also have a stake in the DNS and provide a possibility of building and supporting a structure. IDNH would have the option of opting in and paying a fee to become an At-large member. We think its possible to build and finance this structure to provide sustained involvement and discussions. We would like to have outreach efforts to raise awareness, which will take some time to develop. That will have to evolve over time. We suggest a structure based on 6 geographic regions - each region elects 1 Director and 5 Regional Council members. Each Regional Council appoints 2 members to Global At-Large Council. We see the Councils fostering the At-Large structure. We suggest the Directors be elected for 3-year terms and we believe this election is technically feasible so Directors could be seated by year-end 2002. We suggest a review after the second election. ICANN is a novel organization and we should re-consider it. We think ICANN needs some stability for a period before a review is undertaken. We think this is practical, workable, fair approach to these issues. We need input on all aspects of the draft report. There are implementation issues that need to be addressed. We are here to listen and interested in getting online comments and these will be taken into account when we develop the final report. We are here to listen to your comments.

Alan Davidson: (CDT and NAIS member) NAIS report is in some ways similar and in other ways different than the ALSC's report. Commend ALSC for getting this report out for discussion before its finalized. Members of the NAIS team felt parts of the report resonated - including importance of public participation and public interest in ICANN and being an At-Large member is about more than an election it should be about sustained participation. One important difference is who is the impacted public. NAIS thought a lot about using IDNH for At-Large members. If one looks at statistics of DNHs its heavily dominated by companies and commercial interests and organizations, so we think this membership base would exclude a lot of people and be dominated by entities already represented. Another point was practicality. If you want to have one person - one vote its difficult to determine who that person is and it's difficult to prevent DN warehousers from having multiple votes.

Bildt: There are some points in the NAIS report on which our outlook is the same. One is that although ICANN tasks are technical in nature they are of wider importance and accordingly it's important to have a governance of ICANN that provides for public input. We see it in terms of transparency and acceptability of the process. Providing that it might be a limited part of the Net community that may want to participate its important that ICANN provide a mechanism for participation. Differences between the two reports include DNHs. IDNHs have somewhat of a larger stake in the DNS. With email address holder you're talking about half a billion people, so we're talking about a complex issue. Not a workable basis for building a structure that is financially viable and viable for setting up a sustained participation structure. If the election is email based, and the world has a billion email addresses in a few years and we end up with only 10 million voters (which is scandalous by democratic standards) and then delivering 10 million pin letters out is not feasible in terms of mail service and cost.

Pindar Wong: In considering public interest with respect to ICANN, which fundamentally is that the DNS works (things resolve). One has to recognize that DNHs have a responsibility that their domain names resolve and if everyone does that it promotes the stability of the DNS. On practical aspects - trying to capture domain name registrant data may be difficult, but we need to recognize IETF's work on EPP. We would like input on EPP for use in gathering registrant data. Registrars are using EPP for new TLDs so we're interested in whether this could be used for At-Large registration.

Esther Dyson: Addressing the question of whom we expect to "show up." We are interested in refining these issues over the next few months. The best defense we have against fraud or undue influence is transparency. Any entity that sends a message to a large number of people will send it to someone who posts it, and that's what happened in the last election. There's concern about creation of fictitious people - harder to create a fictitious IDNH, partly because you have to pay for the DN but also for At-Large membership. We're trying to raise the visibility and value of participation - who develops policy in the first place, not just who serves on the Board.

Stefaan Verhulst: I think IDNH model is problematic. You referred to problems of email holders. We don't suggest all email holders as members - you have to register and will limit the amount. I would like to see the total amount of total domain name holders out there - it may also be problematic. It's a question of scale. There are fundamental problems and one I see is corporate capture. 80% of domain name holders are companies. At-Large membership should be the legitimate interface between the public and ICANN. Corporate involvement is inappropriate. You also indicate that you want to restructure the Board by re-trenching from 9 to 6.

Bildt: There has never been 9 At-Large Directors. We did a clean sheet study and we weren't constrained by past activities. We did investigate the historical decisions that were made and we were not impressed by the clarity of thought and reasoning that went into this. Our task was take a fresh look and develop something. I fail to see this public/private 50-50 divide as practical. I see the developers, users, providers paradigm as more productive. There are important differences between email and IDNH. There's no way you can do a purely online election system based on email. We found significant implementation and security problems with email. With IDNH you can run election without mail component. We are dealing with imperfect technology in an imperfect world. NAIS report did a good job of highlighting national capture, which can be a significant problem. You can foresee a large corporation marshalling their employees to vote in a certain way. We think it's likely to be detected. We spent a good deal of time considering the capture problem and we think it's manageable.

Dyson: We're looking for advice on what geographic requirements would be appropriate.

Wolfgang Kleinw¾chter: I think capture is important, but there's also a concern about the entry barrier for email users if you require IDNH. It may not be easy for email users to get a domain name. You will need to educate people on how to get domain names and you should consider creating mechanism to make it easy for people to get domain names and register for At-large. Considering changing the At-Large seats on the Board. I have no fundamental opposition to 6-6-6 structure on the Board, but that must be coupled with a mechanism that prevents At-Large from being voted down by the rest of the Board. Need to rethink the Board's voting mechanism. Regarding membership fee that MAC found that it would cost more to collect the money than the fee is worth. We need more creative proposals besides membership fee.

Bildt: We don't know how many registered and participated in the last election were IDNH. That's not relevant. We want to stimulate the development of options so that those who want to register and don't have domains can do that. We don't want it to be a barrier to At-large membership.

... Jidou??: The entry point to At-large membership - participation is more important than voting. Voting just one aspect. Using domain name as entry point is very problematic. The Internet community will have people who register but might not participate if costs money. Domain name is a financial issue. In some countries you have to pay in US$ which is problematic in many developing countries. Dividing Board by developers, providers, and users is skewed to interests of non-users if IDNHs are the only At-Large members. Lets find other ways to create At-large members.

Roberto Gaetano: I think there's a huge problem between using email addresses and using domain names. In theory, theory and practice they're the same thing but in practice they're not. We should get the largest amount of people to participate; I wonder whether people who don't have an email address should be left out. In theory you can push this to a farther extent, but in practice we must say we live in an imperfect world and an imperfect solution is better than none. I did not like the way the elections went last year. We had a lot of discussions in Europe on mailing list. The point is not to have Directors who are assumed to be talking for the end-user. Instead we need to focus on user participation. For the 5 Directors that we have now, what is the control that At-Large members have on them to ensure that they are voting in At-Large members' interests? First I would like to see a participation mechanism so At-Large members can be involved in a sustained way. We need to go in the direction of Councils or some mechanism for participation. If we have a local structure that will help ensure the debate and help keep the Directors connected to their electorate. Local councils/structures is the type of idea that I would like to concentrate. I would be happy with an imperfect mechanism with IDNHs but lets do it - it's better than nothing - if we can put a local/regional structure in place. I would rather have 5 Directors that have a relationship with At-Large members than have 9 Directors that have a relationship only with themselves. If there is a financial barrier to entry, that would not be acceptable. Technically I think we can find a solution. Private individuals and corporation problem will change in the future. But someone who wants to participate in the election but can't afford it must be addressed. For example, a limited domain name that enables people to be IDNHs and At-Large members. I would like recognition that IDNH is technical issue, not barrier.

Joop Teernstra: This is a good draft report and the interim solution of IDNHs is OK, but it's not going to give ICANN stability for 5 years - that's too optimistic. ICANN was launched as a consensus driven decision-making body. Lack of consensus has been a big problem for the DNSO. Your report has not addressed this or the issue of the NCDNHs. It would not be OK if the DNSO was condemned to be a registration industry SO. Now it is the only forum in which opposing interests interact so its important that they have a forum in which they can work this out and give policy advice to the Board. I'm worried that if you put the IDNHs in their own SO they have no other stakeholders to work things out with. This will push decision making up to the Board, since it will get conflicting advice from various SOs. In the end ICANN won't last as a consensus-driven body. People will say they keep getting out-voted on the Board and there's no way to work out agreement at lower level. I want an ALSC commitment that they are talking about real individuals and that the DNSO will become the workable forum where the divisions are thrashed out. The DNSO should be the proving ground for IDNHs. Voters of the At-Large won't really know their Directors and the Directors should prove themselves in the DNSO.

Izumi Aizu: Was involved in the ISP process and was member of MAC and am member of NAIS. Regarding IDNH I'm concerned that it would cause more misunderstanding for ICANN. Besides stability, we have standards of openness and inclusiveness. These are addressed vaguely in the White Paper. To make global consensus you need openness and inclusiveness. The Net resources are managed by very few. So if you restrict At-Large membership its pragmatic but it may not be socially, politically acceptable. How do you identify one IDNH 1 vote? It may be just as unreliable/difficult as email addresses. Can you count on ccTLD registries? It may exclude AOL users and others. In MAC we were concerned about disenfranchisement. What is appropriate for developing countries? Piggybacking on domain name fee may not be appropriate. Did you consider autonomous fee system for At-Large membership itself, or did you think domain name system was necessary?

Eliott Noss: [TuCows] There's an important thread I've seen in papers and statements on this matter and in the debate that it's registrars (or other business interests) opposed to interests of the users. In some circumstances I understand where that's coming from but its important to not see those two positions as mutually exclusive. There are small vocal minorities who make their views known and a large majority of users who won't step forward and have their voice heard but who have an interest in the process. The one thing the evolution of this industry has shown is ISPs web hosting companies and DN registrars can be effective channels for the voice of the end user. To me the place where this most relates is in the discussion regarding the DNSO and if an ALSO obviates the need for DNSO as it exists or the GA within the DNSO. I will absolutely recognize conceded the need for some level of interaction between the SOs to inform policy and resolve conflicts before it gets to the Board. Have that be through the DNSO puts the DNSO in a different position than the other SOs. I fear as a registrar being painted as opposing end user interests. If you look at what's happened in this industry to this point it has been one of the greatest justifications that a process like this can reflect end user interest. It's reflected in prices for domains going down and smaller registrars taking market share away from larger registrar. Every quarter this industry from a structural perspective has a concentration ration that's going down that is a boon for user interests and it reflects that users interests can be affected by so called business interests.

Adam Peake: (Nais member) Question regarding IDNHs as At-Large members - most domain names held by corporate interests you're creating situation where all seats could be captured by business interests. In the last election we faced nationalistic capture but at least different countries representatives bring different perspectives to the Board. Also, some ccTLDs don't let individuals register domains. Question about fraud - please explain the terms of the figures taken out of the database. Can you explain 6-6-6 - how would remaining 6 be constituted - also seems like you'd be rewarding the DNSO for its disfunctionality.

Edmund Catiti: I want to add my voice to congratulations for the work done. ALSC is looking for practical solution - not a public system - solutions being looked for. I want to add my voice to the issues of capture. When you consider developing countries, corporations have larger portion of influence so it's a larger problem (they own more of the domain names), so I think we need to look for a different approach to get representation of users and their interests from those areas.

Zachari??: I come from a French speaking country so my English is not good. If the purpose of the At-large participation is to allow users to have a voice in ICANN activities, then using IDNH will be a problem for a large number of people in developing countries because of two problems - financial - costs too much, and in some countries even if you have the money its not easy to become a domain name holder. The main problem you have with At-large membership is a lack of information. Outside of the election period there is no activity and I want to propose that the Board has objective that when we elect the Board we have to define objectives to the Board because after we elect the Board they don't have guidelines for activities. It would help us evaluate the work done by the Board. If we can resolve this problem it would be preferable.

Alan Levine: (NAIS and other group member) I think the comments you just heard from my African colleagues reflects the views across the continent. I think you proposal adversely effects At-large representation. I'd like to give you some figures - in addition to the fact that its costly and complicated (complication hampered African participation) if you look at number of hosts carrying domain names Africa accounts for .2% of those connected to the Net of which 87% is in South Africa, if one excludes the top 4 or 5 we're talking about less than a thousand domain name hosts carrying less than a thousand domains in Africa. I would urge you to consider this. It smacks of legislating the digital divide, almost by narrowing the membership.

Carl Bildt: That is, with due respect, pure rubbish. I'm all in favor of debate but there are limits to the rubbish I can take. If you talk "legislating digital divide" that's pure rubbish. Close to half of the population of the world has never made a telephone call but I would not tear down the telephone system in the US because of that. I would do serious efforts to bring telephone system to work in Africa, but I would not design the telephone system of the US because of the state of telephone system in Africa. Have some perspective - we are trying to propose a system that is fair and workable throughout the world. We are not legislating the digital divide - there are limits to rubbish I can accept as chairman.

Levine: I think its very closely linked to Board representation I think equity of the Board is important. I think in retrospect your interruption to my comments and your favoring certain participants in the audience with regard to speaking is somewhat out of order.

Bildt: I'm sorry you were last on the list, but in the last meeting you were first on the list. I wanted to allow those who haven't spoken at the last meeting to have a chance to participate. I wanted to make sure to involve others.

Bildt: Regarding the capture issue - we've spent quite some time on the capture issues and they're always there. The corporate capture is there on the email address approach as well. If you take a Swedish company like Ericsson they have employees all over the world who have emails addresses and their numbers are huge. In theory they could ask all of their employees to vote in a certain way. In theory, the threat is there with any approach. The barrier issue was raised - its important to lower the barrier for obtaining domain names and if this will drive down the barriers, so much the better. On the cost element - cost is coming down but there's no such thing as a free election. There is a cost that will have to be paid by someone. And then there are two issues - to have a system that is not too costly - if we can take the mail and pin number out we are talking about a substantially less expensive election. The second is who is going to pay for it. You can have everyone pay for it, even if they don't want to be part of it, or have a membership fee. We ended up on the second approach that reduces the cost burden on everyone else. There's no such thing as a free election. The philosophical issue - the wider you make the electorate the more philosophically attractive it sounds. The caution I have for that then its not limited to email holders. You could say the stakeholders in the Net are far beyond email holders. If the Net breaks down it affects functions that affects lives of everyone in the world. If you follow the argument that anyone who is a stakeholder should be an At-Large member then you follow that argument to the logical conclusion and you end up with an inter-governmental organization and hand ICANN over to the ITU or UN because then you can really say that you have a structure for participation of "everyone." I won't argue about the proposition of governmental involvement, but if you say that everyone should be involved, that is where you end up. We considered all of this and came back to the practical aspects and looked for something that is actually workable (probably workable because there are important implementation details to be worked out) and we focused on participation - the ALSO - and IDNHs which gives us something that could work with reasonable costs and implementation timeframe - as most appropriate way of running it.

Break

Wong: Following a point that Levine made about the DNS being a cesspool and why you would want to join. Levine is correct that the DNS is a bit of an issue there needs to be some cleaning up but we are not using At-large to address the DNS issues, other than to make people aware that the IDNHs have a responsibility to make sure their domain name resolves. It's bringing the awareness issue and operational stability issues to a wider community awareness.

Joop Teernstra: I raised the point earlier about At-large becoming an SO and the danger of Board member elections but not having a forum in which domain name holders can thrash out their issues with registrars and ISPs... My thoughts are available on line. The DNSO would become no longer functional if the providers withdraw from the DNSO and I would like the ALSC to comment on this.

Wong: One thing that has come through that regardless of how the providers, developers, and users aggregate that there be a mechanism for them to interface with each other. That was true in some of the discussion that we're aware of regarding the ccSO - that there is a commitment to have a liaison with the DNSO to at least meet at the same time. It's unclear what level of detail at which we'd need to address this issue. It's unclear if this can be left to self-organization. The idea of a general assembly has some merit but more thought is needed. How do you encourage ideas to be debated and considered among the SOs? You don't want the Board to make the decisions you want the depth of arguments and discussions to be made below the Board level. You want the bottom up consensus development process to be there.

Teernstra: Has the ALSC thought about the role of IDNHs not only as part of AT-Large but also as part of the DNSO where they can be part of the Names Council?

Bildt: The ALSC is reluctant to get into that. We were given 5 questions by the Board and I would recommend that we not get into issues that are outside our charter because that my detract from our ability to complete our mission. I don't see our recommendations preventing other issues from being addressed elsewhere.

Wong: One of the things that's clear from the history of the DNSO is the difficulties involved with establishing a domain name holders constituencies and how that relates to how the individual users viewpoint need to find some mechanism to be expressed. We're specifically saying it should be expressed through an ALSO vie IDNHs and we want to foster pariticipation and that a mechanism is provided to encourage that.

Kleinw¾chter: This morning we talked about a philosophical approach and moving to a 6-6-6 structure and considering providers, developers, and users instead of private vs. public approach. I fully agree with this approach because it's practical and workable and makes sense. But the public interest does not disappear and you have to consider the contradiction between public and private interests and the need to balance these. Both in the developers and providers groups. We know that the code is the law in the developers group. If you go 6-6-6 you have to include at a higher level the need for balancing private and public interest. There's a risk that at the end of the day the private interests will dominate and the public interests will disappear.

Bildt: I agree the public v. private issue is present in all three groups. To complicate things further, its not entirely obvious. Look at the capture issue. Consider the mobilization of an electorate in Germany to get a German or European on the Board - that would be considered a public interest in Germany but a private interest in the rest of the world. So it could shift depending on the perspective. I think the public v. private doesn't work in the real world and you need to be aware of those different perspectives in every issue. Like capture, the issue is always there in any approach. I don't think we can have a further mechanism that recognizes public v. private.

Wong: It's also a question of degree. I would have you consider that the main interest of the wider Internet community is the stability of the Internet. IDNHs should at least be aware of the need for their domain names to resolve. That is a tangible measure, which we feel for the representational standpoint, is justifiable. It should be clear that we do view the real challenge as creating the mechanism for participation. We have not given prescriptive details on how the ALSO should be structured. We heard in the ccTLD meeting that they represent all these different worldviews and wouldn't there be a more natural match for what we're proposing and we welcome your views on this.

YJ Park: (NC rep for NCDNH) We had an interesting discussion with Michael Pelage. He asked Paul Kane about ICANN making policies or being a technical coordinator. The ALSC assumes that ICANN makes policies which is different than what ICANN was created to do. What was your starting point? The starting point for most of us is ICANN is a limited technical organization and not a broad policy making organization. The ALSC's recommendations indicate that ICANN is a broad policy making body.

Bildt: I wouldn't phrase it like that - that ICANN makes policy. The basic issue of ICANN remains technical but in view of the importance of the Internet there are policy implications and public concerns in these technical issues. I would not favor widening the mandate of ICANN to be the policy making body for the Internet, which would be wrong and potentially disastrous. Be aware of the fact that there is a public interest in the Internet and public policy implications of ICANN's decisions. Consider how you define a country. People go to war over that issue and there must be a way for ICANN to deal with that issue. Defining a country can be a technical decision but it has broad policy implications.

Park: I think there are a lot of objections to your recommendations of IDNHs because more of the public needs to be involved. Bildt: We think there is a strong case to be made for the involvement of At-large members. Then for a number of other reasons we end up with a third of the Board being elected by At-Large but clearly there is a need because of the policy implications of some of the issues and to ensure transparency of ICANN's process. But basic mission remains a fairly technical one.

Roberto Gaetano: This morning there was a proposal to use addresses instead of domains for identifying members. In my opinion this is not feasible. Someone who has an email address and is connecting by dial up with his ISP doesn't have a fixed address and you can't map the IP address to the person. I would like to go back to Kleinw¾chter points and the proposal of having 6-6-6 Board members.

Bildt: Our proposal is to have 6 and 12.

Gaetano: In the corridors what you hear is 6-6-6, but it doesn't change the spirit of my question. In one way or the other we are re-shaping the structure. We need to be careful of defining the scope of At-large's participation and avoid overlap with those already represented in other place in ICANN. I am paranoid. Having followed this process from the beginning I have found that what I thought was paranoia was simply looking in advance at things that were going to happen. Considering Europe there were 5 names put on the ballot by ICANN and at least 4 of the 5 were active in telecom or other parts of the infrastructure and were already represented in ICANN. The popular vote elected one of the others. The top two vote getters were not recommended by ICANN. We should not make the same mistake with the new rules. We should clearly define what part is covered by At-large and what part is covered by others. An individual speaking as an individual it is very difficult to "capture" the IETF or others so the problem doesn't appear. But it is very possible to have it the other way around - to have the telecoms or registrars to also act as "individuals" and occupy other slots and unduly influence other sectors. Define clearly the scope of the different parts and define eligibility of candidates to the Board to avoid this problem in the future.

Levine: Regarding explanation about parallel of email and domain names - I don't support either approach. As I understand it ICANN allocates domain space to a registry that then has a relationship with a registrar who then sells a domain to a domain name holder. In the same way ICANN will allocate IP address space to a regional Internet registry that allocates that to an ISP who allocates that to a user. Although a dialup user may not have a fixed IP address but the ISP holds information on who is the user of that IP address. Although that IP address may change the mapping of IP address to user is held in that authentication in the same way the owner of a domain name is held in the registry database. That was the comparison I was making.

Teernestra: I would like to give Esther Dyson a chance to comment on my comments from this morning - the fact that the At-Large will be made up of IDNHs it has consequences for DNHs being a part of the DNSO. Has the ALSC considered this and would they support DNHs involvement in the DNSO.

Dyson: The ALSC has carefully avoided specifying anything about the DNSO. That is not our charter. In my personal opinion, attempts to try to represent individuals within the DNSO have still not attained traction. Our goal in the At-Large is to make sure the concerns and point of views of individuals is represented in its policy deliberations at all levels. So we are not simply trying to build a structure for electing Directors but for getting involved in policy making. What we have found is voting happens every 3 years. Unless you're lucky and you have a Director with a lot of time and money your voice doesn't get heard when the decisions are being made. Its important to have a structure beyond an election system and we're looking first to create regionally organized councils (my hope is in the future it may be organized along issue based lines) The point is to be involved in decisions while they are being made and not after the fact. You can revisit history and think about putting this within the DNSO but the fact is that ICANN is more than just the DNSO and there is a need for some formal structure to represent the public interest and this is one way to do it. If the ALSO determined that there was an IDNO that was completely effective and putting forth all the views necessary it might be become less active. The ALSC has focused on a membership structure that works, that has a means of outreach, that is decentralized, that will meet with general consensus based on our discussions with a broad cross section of people - with individual users and with parties that need to be involved to make this happen.

Teernstra: Do you think it would be possible to have a general assembly of all the SOs. I'm concerned that there won't be the needed interaction.

Dyson: I guess I don't care what you call - what we're trying to create is an effective way to represent the public interest at the decision making stage. It's very important what happens before something becomes a resolution (and afterwards, rather than during)

Teernstra: But the other constituencies aren't represented within the ALSO.

Dyson: The goal is to get the At-Large members to communicate. Communication is not only a formal structure but also assumptions you make going in. We're trying to create a body that will be effective in discussions.

Teernstra: How will At-Large Council operate? Wong: Can't say. This is where the discussion begins. We've created a framework on which to hang more specific ideas - the wider community can discuss how it wants the Council to operate.

Dyson: We're not trying to create the texture - we're creating the skeleton for the public to use.

Bildt: Don't expect us to sort that out in the final report. We won't. We'll create the possibilities of the structure that make it possible for bottom-up approach to produce something. Exactly how that will shape up we don't know and we don't want to dictate it. There are a lot of creative people around that can do it. We create the possibilities and the framework and the possibilities for it. I can see that the regional councils may be very different in the way they operate in the initial period.

Kleinwaechter: I like the picture at the moment that we're creating the bones and the meat will have to be added. The final report will be the beginning. A point made in another meeting, the current At-Large Directors said they are missing interaction with a qualified group. Other Directors have lawyers on their side and it would be most welcome if we had a group that was able to vet positions. Katoh said he gets 300 emails a day and he can't reply to all of them. We need to recognize these realities and create some body in between. I'm more interested now in the financing of the system. I disagree with a membership fee because collecting the fee is costly and prevents participation of some groups in the world. You should consider creating a special fund by certain stakeholders so no one stakeholder group can influence At-Large. My understanding that the role of the government is to enable citizens to participate in society and if governments made a contribution to enable their people to participate in At-Large. I know this is a slippery area that causes concern but I think we should explore other avenues for funding.

Bildt: It's also fairly easy to integrate a fee into the IDNHs approach - when you pay for the domain name you opt in to membership and pay a fee. I think its wise to integrate this into the ALSO. I think creating funds is unworkable and the system would be open to influence - or there would be the perception that At-large would be run by or influenced by governments or organizations or whatever. I think its important that At-Large be independent and self-funded. We want to recommend a system that is not too costly.

Peake: Question about 6-6-6 division of the Board. What are these groups?

Bildt: We will not - what we have said is that At-Large would elect one third of the Board (6) and we have not specified how the other 12 would be elected. It's outside our core mandate and we have time restrictions. So the short answer is the ALSC can't answer this question, but it certainly needs to be answered. Our proposal is 6 and 12 and we are not suggesting details for the 12.

Christopher Chu: (Internet Democracy Project and ACLU) Regarding membership fee - a concern is that the draft proposal's fee and criteria for membership and they're troubling in many ways. Requiring a domain name and a membership fee before a person can vote. This suggests that you don't have an interest in the DNS operation unless you have the money to get a domain name or membership. I find it deeply disturbing. You can equate to old laws that only men who own land can vote. In some ways you can view your recommendations as making the same mistakes. Regarding the fee, I'm concerned that it will be egregious outside of the western world.

Bildt: We want to have a system that is open to all who want to participate. You can opt in by saying that you have an email or by saying that you have a domain name. We chose domain name because we think it's the practical way for conducting an election on line that is not too costly and people who have a domain name have a stake in the system. We also should try to reduce the barriers to people getting domain names. We need to make sure that costs are kept reasonable. There is going to be a cost - that's unavoidable. Setting up the ALSO and a global election will cost money and we think it should be paid inside the system - not by outside entities. We think there should be an integrated financing system and a membership fee for the ALSO if you separate the financing from the expenditure you'll end up with endless expenditures and an unwillingness to pay for them. Are there going to be people who are unable to pay for it. Yes. The world is divided among people who are richer and poorer. The ALSC hasn't solved that problem. The Internet use is divided among different parts of the world - it reflects the state of the world. We must make certain we create the conditions for everyone to be part of it. As I said in previous discussions, a substantial portion of the world's population has never made a phone call - will never make a phone call in their lifetime. That's the way the world is, I'm sorry to say. I'd like to change that, but I don't think this is the mechanism to do that.

Dyson: The domain name requirement we see not as "you have to have a domain name or you can't play" it is any one is qualified to play, the way you get onto the voter roles is by registering a domain name and that has the advantage of decentralizing the registration process worldwide and has the advantage of being the ICANN community closer together because you have different people interacting with each other. We are sensitive to not being able to deal with the digital divide and we are interested in working those organization that are addressing the digital divide and are looking to organizations and foundations to make domain name holding and membership fees attainable.

Wong: One of the virtues of our approach is - if ICANN is successful the barriers to holding a domain name will go down. This trend should continue. An example is .name.

Bildt: You could argue that with this system the poorest people of the world will be over-represented. If the poorest people in the world are in Africa - Africa is guaranteed a seat. But it will take quite sometime for Africa to be one-sixth of the world's Internet users or domain name users. The system as designed has an over-representation in it.

Levine: The concept of making it work - I really, really agree with that. I want to make domain names work. The fact that it's restricted in numbers and excludes a large portion of users has been covered. It's complicated - you have register a domain name and have infrastructure to handle a domain name. The issue of agreements with registries is a hurdle you also have to get past. The fact that domain names are expensive needs to be addressed. You have not yet addressed the authentication issue with the registries. There is no authentication in the vast majority of registries. In the majority of cases the only verification is payment. What difference is it in opening it up to emails?

Wong: Payment and collection mechanism.

Levine: I would support open to email address holders and make a membership fee. Rather than invite the UN in sponsoring membership and go through all the domain name system hurdles this would be better. Then we are at a point where we can start negotiating. If you're going to charge every email user then you have to address equitable representation at the Board level. I want to draw a distinction between registrars and registries. Registrars are competitive but registries are monopolies. These parties should be willing to pay to get a broader representation at the Board level. Equal representation on the Board and bearing the cost need to be addressed.

Wong: Perhaps I'm missing something. ICANN does not have a specific mandate to deal with the email address system, per se. It does have a lot to due with the DNS. For the email system to work the DNS must work. I find it difficult to link the idea of stakeholder to email address holders. I find it more appropriate and concrete to talk about IDNH since Icann deals with the DNS.

Levine: Pont of information - for an email address to work its not only the DNS that must work it's also the protocols and standards as well as the IP and address system that must work. Wong: I understand. I also understand even the standard bodies have in having the standards they recommend implemented. So we're talking about the stability of the DNS.

Bildt: An alternative approach that you've indicated is email and have those that opt in pay for the system. That may be a feasible system. The downside is that you end up with something that is very expensive - probably $10-$15 a vote because it's a costly election mechanism. Then I think you have smaller numbers for participation. The base for domains is smaller than emails but it is more integrated in what ICANN is doing. We can reduce the cost and integrate the system better than we can with emails. Cost is a factor for people.

Levine: I disagree with that logic. You're using the DNS as an excuse to reduce the cost. The only thing that's authenticating the user in a DNS is the payment. It costs the same whether you're giving someone a payment through the DNS or another system.

Gaetano: The difference is if you have an email address holder pay a fee - how will that be done? You'd need an infrastructure to do that. If you want to make a payment of $1 it would probably cost you $20. This feature is already built into the DNS because they can piggyback membership payment in existing system. I don't like the idea of moving from email to domains because it reduces the membership but I think we have to stick to reality. A real possibility that achieves results is better than some ideal that will be too difficult to implement. We must go from the ideal that we all have of reaching every person on the planet. You need to be effective. Otherwise if it's not feasible, even if it's ideally perfect, is not as valuable as something that is not perfect but is feasible. We have to start somewhere - if we don't reach a consensus proposal then we won't have an election next year and no one will have a stake in ICANN. It's not a question of who wants broad power and who wants to restrict it to "land owners." The difference is between people who want to start with a position that has some merit and will allow us to take the first step and being stuck on square one for years and years in discussing the best solution. We have a saying in Italy that the best is the enemy of the good.

John Klensin: You should be aware that it's impossible to tax email users and it's a more serious authentication problem because people having numerous email addresses. The other issue with email systems is that if you're interested in a stakeholder presence, you should note the number of email servers on the network is probably far lower than the number of web servers and web accounts. The next observation is that in looking at domain names as a qualification is considering that some countries are pushing deeper hierarchies and if you tie it to second level or third levels in countries that choose intermediaries you may distort the policies of countries that are doing things that are structurally beneficial to the DNS. Conversely if you look at registrations deep down the notion that people pay for domains on a per host basis is false. Because in company or country related structures the charging structures deep in the tree registrations are typically based on some other algorithm. Similarly looking at developing countries they are typically running their domain systems on little or no charges and adding any fee would put them out of business, which is bad for At-Large. A personal policy observation is that ICANN's mission is technical in nature and it is not to act as a disseminator and enabler of the Net to everyone in the world. I'm very concerned about a notion for the At-Large that ICANN should extend itself into that dissemination area because it's in that area where people have no contact to the Net but have an interest. If you stick to ICANN's original charter than the focus would be on current stakeholders and making certain new stakeholders can be accommodated when they show up.

Wong: One thing we would like input on is how deep to go in the hierarchy. We define it in the draft report as "A registered name within the domain of the Registry TLD, whether consisting of two or more (e.g., john.smith.name) levels, about which Registry Operator (or an affiliate engaged in providing Registry Services) maintains data in a Registry Database, arranges for such maintenance, or derives revenue from such maintenance." Does that cater for the deepness in hierarchy?

Klensin: I don't think so. .us is an example of the hierarchy of being organized by geography and individuals are registering at the 4th or 5th level and the registrar goes down about one level, typically. .us is not typical but models like this do exist elsewhere in the world. We also have situations where NGOs and professional societies and others are registering domains at the second or third level so the organization is visible at the registry but others below that are not and individuals are invisible to the registry. The cautionary note is that this is the way the DNS was intended to work so if you implement a policy to discourage it is a bad policy from a strategic standpoint for DNS.

Michel: We used the definition in the report that ICANN uses in its contracts.

Wong: We dropped the last line about dormant domain names. We'd appreciate input.

Klensin: When you start trying to reach those deep names you run into privacy concerns. Making individuals visible to a national registrar base raises new questions. ICANN's language was written on the basis of identifying people at the second level because of ICANN's concerns. Now if you're talking about reaching people as individuals, they may be much further down.

Dyson: On the privacy issue, you should know that we are still working on it and would appreciate input. The notion is that if you want to vote you need to be identified to some system so you can be checked-off when you vote, but that doesn't mean its made public. The separate issue is how to create a list so people can send you communication from people who want you to vote for them. Whether this should be opt-in or opt-opt-out, whether this should be an alias-mailing list is unclear.

Klensin: But as soon as you bind it to the DNS you're in another kind of business. The ability to resolve a DNS name involves the ability to pull information out of the DNS. There's been a discovery that if you do E164 based phone numbers on the DNS it provides more information than is available for unlisted phone numbers now. I'm concerned about binding it to a DNS entry because DNS entries are retrievable or they don't exist.

Dyson: Maybe it's a sacrifice you make if you want to vote - but these are good issues to address and clarify.

Teernstra: It's good that we're touching on the definition of an IDNH because it needs to be clarified in the final report. There's a concern that At-Large should be for individuals and not corporations. Make it clear that we are talking about domains registered by individuals.

Bildt: That is a problem that we discussed this morning. Corporations usually have multiple domain names. It's a problem derived from the proliferation of the use of the Internet.

Teernstra: If someone is fired by a corporation and is removed as an administrative contact that person should not be an At-Large member.

Wilfred Wober: (European Address Council Member, Vienna University) I'd like to share some of our experiences in the address space. You should try to find a definition that is independent from technology or criteria. For example, you can't define an ISP. I have an entity at the university and at the AC and my own domain. If I allowed my students to register under my domain I would be some kind of an ISP, but I would hope that I could still vote as an At-Large member and I would also be able to vote as an ISP (in DNSO). I would caution against wiring economic or organizational assumptions into a definition that will be hard to track or will not evolve or time as fast as the infrastructure.

Wong: Using the address analogy you are an IDNH if you think you are one.

??: One thing I found interesting is that you're building on the concept of using the DNS name as a token for identifying the individual. You're assuming it will stick to the individual much longer than an email address and you're assuming that it will be cheaper than other mechanisms, such as pin numbers, etc. This is the thing that I found interesting and why I think it's a good proposal, even though it has many deficiencies.

Bildt: Whether a domain name sticks with an individual - most of us have numerous email accounts and domain names are likely to be the same. This creates complications but reinforces the importance of dealing with many of the issues involved with IDNHs.

Dyson: I have an At-large email address and it forwards mail to one of my other accounts - I probably have 50 email addresses because people keep making them for me.

Teernstra: It might be of help if the ALSC looked at the "color of title" for a name - the power to shut it off. If you use that criteria you are talking about an individual. Look at the our old charter.

Wong: Following Teernstra's comments - how would you check the "color of title"?

Teernstra: There are a number of prima facia criteria - payment or registration contact - look at the URL, which addresses this point.

Levine: So as I understand it we've come down to economics and the pro for using DNS is cost. As I understand it, my point is that to register any domain, including hosting a domain, it will cost more than $15 and the estimate that was presented is the cost to authenticate email address holders would cost downwards of $15.

Gaetano: I would be happy if the final version of the report indicated that using IDNH is valuable it would be good if the recommendation of ensuring that some way of registering domain names for the purpose of At-Large membership at a low cost will be ensured. Specifically, investigate specific TLDs or using a special registry that would provide no cost domain names. Alternative to registering should be addressed for no cost IDNH that could not be exploited for commercial use (so as not to take business away from registrars). I'm concerned about keeping the bar as low as possible.

Bildt: So are we and we've been discussing alternatives. I don't see any technical problems is setting up an alternative that has a low entry threshold for those who want to be involved but can't afford an IDNH and membership fees. That might entail a domain name for just this purpose. We are looking for other approaches to deal with potential disenfranchisement. If we achieve low or no barriers to domains then, considering costs of operating the ALSO, you can link the expenditure of the ALSO to the fee.

Verlhurst: Can you give a sense of what the fee level will be? High fee means less participation, etc. If you don't ensure that there are low or no barriers, how will you subsidize? Will there be cross-subsidization?

Bildt: I don't see cross-subsidization as a good option because of the problems I outlined earlier. Concerning cost I see three components - first is getting a domain name (costs can vary greatly). I don't think no cost is achievable. I don't think we will set the fee by the final report in Nov. We also will not provide detailed specifications for the ALSO - that should be a bottom-up activity. But I agree that for the start-up period there does need to be an estimate and some seed funding will be needed. But later on it's up the members of the ALSO themselves to decide what kind of membership fee and activities will be associated with it.

Michel: For the big ticket items estimates will eventually be provided (general estimate for cost of election, cost of outreach, etc..)

Verlhurst: How do you project regarding the electorate? I don't see a difference between email and IDNHs - you still have a problem of projection.

Dyson: Deal with by having the variable cost covered by the membership.

Verlhurst: How do you set initial cost? Want to identify the dependencies.

Wong: Fixed costs established by registrar, registry infrastructure. Not like you're starting from scratch.

Michel: You would take a general estimate of costs and recommend a first year number and based on how many members register and pay fees, you look at expenditures and presumably this figure would change for the second year.

Dyson: There's not only domain name registration opt-in but also reaching out to existing IDNH - we're hoping that registrars will help reach out to them. A mitigation of the cost is the elections are only every three years. There's a big difference between participation and voting. You don't have to have a domain name to participate as a member of the lists - you can communicate, state your opinion - do everything except vote. So it may be the people only bother to become members every third year, which would be unfortunate. Vis a vis the open participation - there's an ICANN official list but as you know there's lots of other things. As long as you're not official you can engage in filtering and its not called censorship. Anyone can filter the official ICANN list and call it something else and anyone who wants to subscribe to it can do so. We can see what the community wants.

Teernstra: The point on open participation - there's a link between capture and open participation. There are certain individuals who are able to capture the attention of a list or discourage people from serious works - it's a matter of participation and not spamming.

Bildt: We don't have a solution to that problem.

Dyson: You can create membership requirements for lists. That's what I hope to see arising - people who are interested sign up for a list and people who don't act in the way the majority wants them to get kicked out and they can go start their own list.

Bildt: Thank you all for coming. This has been instructive and constructive for all of us. There are some most difficult and complex issues. Informal outreach will continue. Thank you.