APNIC 29 APRICOT 2010 Banner

Transcript: APNIC Community Consultation


Due to the difficulties capturing a live speaker's words, it is possible this transcript may contain errors and mistranslations. APNIC accepts no liability for any event or action resulting from the transcripts.

1400 - 1530 (UTC+8), Wednesday 3 March 2010

SRINIVAS CHENDI: Sorry, we just take five more minutes to start the session. Thank you.

OK. We're able to start. It's not a good start for me. I apologize first and I do apologize for the delays. Good afternoon, APNIC members, guests, ladies, and gentlemen. I'm Sunny Chendi from APNIC and on behalf of the APNIC Secretariat I take great pleasure welcoming you to APNIC 29 here in KL. So before we start, APNIC would like to thank the generous support offered by our sponsors to this event Telekom Malaysia and PHCOLO. Ladies and gentlemen, please give them a round of applause.


We appreciate your support and we look forward to receiving the same support in all of our future meetings. With every APNIC meeting we are providing live video, audio, and chat. As well as, as you can see on my left, we're also providing the live transcript. Remote participants who are not here in the room, cannot attend the meeting, can participate in all the APNIC sessions through our remote features. And I'd encourage all the remote participants who have joined us to utilize these features, to ask any questions or comments through the Jabber chat room. And now I'd like to welcome our Director General, Paul Wilson, to introduce the session. Thank you.

PAUL WILSON: Good afternoon, everyone. Thanks very much, Sunny. Thanks to everyone being here at the session. It's a good turnout for what I think will be a very interesting and a quite important session. I'm not going to go into a detailed explanation of what is to come because that's going to become clear and, of course, there are some details on our website as to what this session, this Community Consultation session is all about. But I did just want to say a couple, make a couple of points. The session, as you probably know, is designed to provide an opportunity for input from this community on some proposals and some considerations, deliberations of the ITU, which are posing with the ITU to become involved in something that's very close to our hearts, in the management of IPv6 address space. This is something that we've seen coming for quite some time and you'll hear all about it.

But I think the point of the session is that we don't generally, in this community, see enough of our friends at the ITU and in their meetings, they don't see enough of us. And I think this session is really a bridge between those two communities. It's between communities and protocols - two, if you like. So the point of the session is to bring the voices of this community out on some issues that we'll hear about. And to give the people who are concerned with those issues and proposals the chance to bring those forward.

So there will be a formal interaction. There will be a presentation of the results of the session to ITU meetings that are coming up very soon in the coming weeks and beyond. I would strongly encourage those who have something to contribute to do so. But I'd also ask you to do so in the spirit of mutual respect and of trust in all of our abilities to understand each other and to come to agreements that will actually make sense for everyone. So I'm sure that our able chairs will help to ensure that's the case. But I really would like to ask everyone here to see this as a bridge between two communities, who need to actually get to know each other better.

As to the session, I'm very happy to have two co-chairs, distinguished co-chairs here for the session. The first is Masato Yamanishi. He is the Deputy General Manager of Technical Planning at Softbank BB in Japan. He leads the team responsible for IP addressing and peering at Softbank. As you know, that's a very major provider in Japan. Yamanishi-san has been actively contributing for several years now to the APNIC policy process, and at the same time, to the ITU's processes through the NGN and IPTV standardization processes. Yamanishi-san knows both worlds. His company, Softbank BB, is a member of both APNIC and ITU and he's really in an ideal position here to help with the chairing of this session.

The second is a man who almost needs no introduction. I think most of us would have known Sharil's work for many years. Sharil Tarmizi is the Chief Operating Officer of the MCMC, the Malaysian Communications and Multimedia Commision, a government agency here in Malaysia that does very critical job within Malaysia. He is also the immediate past chair of the ICANN Governmental Advisory Committee and served in that role very successfully for quite a number of years. This is amongst, in case of both our co-chairs, amongst many other achievements.

In the interests of time and moving along in to the real content, I'll pass over to the chairs of the session to take it further. Thanks very much.

MASATO YAMANISHI: Thank you very much for the kind introduction from Paul. Thank you very much for participating in such an important session.

Before starting presentation from the panellists, let me explain a little bit what is ITU. Because somebody, of course may know what is ITU and their current activity, but somebody may not. Let me explain what the ITU is and the connectivity about IPv6 allocations.

Could you show the chairs' presentations? Actually, many point about the ITU are duplicated with the first presenters' presentations. Let me skip a little bit. OK, go. OK. So ITU is the International Telecommunications Union and it is part of the United Nations. And under their umbrella, they have three sectors. The first one is Radiocommunications and the second one is Standardization and the third is Development.

The current topic is Standardization sector and also the Development sector. WTSA is a top-level meeting of ITU-T and it is held every four years and the last one was held in 2008. The resolution is talking about IPv6 and you can see the last item it says, "study the question of IPv6 address allocation legislation for interested members and especially for developing countries". Then, based on this resolution, ITU Council, which was held in 2009, decided to conduct further activity towards the implementation of this resolution.

The ITU IPv6 Group was made based on the result of the ITU Council last year. The terms of reference of this group includes drafting global policy proposal for the reservation of last IPv6 block and also possible methodologies and implementation mechanisms to ensure equitable access to IPv6 resources by country, and also the possibility for ITU to become another Internet Registry and a proposed policy and procedures for ITU to manage and reserve the IPv6 block. And the last one is the feasibility and advisability of implementing the CIR model for those countries who would request national allocations.

And next step is - sorry, I missed one information. The first meeting of this group will be held in 15 and 16 in March, and the conclusions of this group will be reported to next ITU Council which will be held on April 13 to 22. Then it goes to WTDC and ITU Plenipotentiary Conference. Now we're standing very important status. That's the reason why this session was planned.

So, before starting presentation from panellists, - I'm sorry, I think many people is also attending through Jabber. So could somebody volunteer for Jabber subscriber for Jabber attendance? OK, thanks.

Then another request from chair is please discuss, in this meeting, please discuss ITU and not to discuss or say comment for ITU itself because time is very limited. So we would like to concentrate to ITU's idea.

Another request is, as I said, time is very limited. If you will have a question for clarification from each presentation, please ask just after each presentation. If you will have comments or discussion, I will take enough time after all the presentations. OK.

Also, I'd like to request each panellist, please finish your presentation in 10 minutes. I'm very sorry I'm limiting the time. But time is 1.5 hours only. So please limit within 10 minutes. OK.

First presenter is Xiaoya Yang. She worked for China Telecom and MRI for China and now she is working in TSB which is a secretary office of ITU-T. And she will, she will explain about current IPv6 allocation statistics and some concerns from member states of ITU and also ITU current road map.

XIAOYA YANG: Thank you very much, chairman. My name is Xiaoya Yang and I'm representing the International Telecommunications Union. I appreciated very much the introduction from the chairman, and also the introduction from Paul Wilson. He mentioned that I'm presenting some discussions not happening in this group. So I beg your understanding, if there is something that you are need further information, I'm quite happy to answer your questions at the end. And I apologize for a lot of jargon used in this presentation. I will briefly talk about ITU but I guess I will skip the slides.

I will talk about these IPv6 issues, center of the ITU discussions. And talking about the mandate we received from ITU membership why we are doing this and a little bit about activities because they were also covered by the chairman and I will come to the conclusions in the end.

So I will skip this one. Just adding that we have a membership of 191 member states and over 700 sector members.

And also a little bit about the vision for the future of ITU because we see important private public partnership in our work. And we are planning to have a new membership category for academia university and for reduced fees for industry members from developing countries.

OK, so talking about IPv6 issues. Any talk on IPv6 has to start with IPv4. So we all know that the IPv4 address allocation is imbalanced worldwide, mostly due to historical reasons, but also we believe that is contributed to our IP resource allocation policy. For example, in this figure we see the status as of May 2009. The United States holds more than half of the allocated IPv4 addresses.

And the first-come, first-served policy has an early adopter's reward; therefore, the US is continuously aquiring one-quarter of IPv4 addresses allocated each year. This is the figure for 2008.

Therefore, we are currently in a situation that, for example, if we look at the figures here, our Internet users in the United States have about 6.6 IPv4 addresses. Well, this number in China is about 0.67. And the same number in India will be 0.23.

Looking at the figures, different people come to different conclusions. Some ITU members, mostly developed countries, believe there is no problem. They believe IPv4 address distribution correctly reflects the development history and the current usage of the Internet.

And coming to the IPv6 deployment, taking into consideration its slow start, but the justifications from these people are that this deployment is driven by market forces and it's currently picking up at the economically optimum rate.

However, the other groups of ITU members, mostly from developing countries, have different feelings. They feel that they are the ones paying for IPv4 depletion because they have paid higher prices for IPv4 so far, and they don't have extra IPv4 in stock. Therefore, when IPv4 reaches its depletion in the near future, they will be forced to deploy IPv6 when they're not yet ready. And they might have to pay even higher to get IPv4 resources necessary for the transition period.

And also their fundamental concern is will IPv6 policy involvement repeat the history of IPv4? Will we have a scarcity issue again for IPv6 in the future? Because we're having generous allocations right now and because we're also following the first come, first served based on the immediate-needs policy.

If the answer to this question is yes, here I will quote a saying from APNIC meeting. "From a public policy perspective, there is a risk to create yet again an early adopter reward and corresponding late adopter set up barriers and penalties."

The developing countries feel it's very difficult to get their concerns heard and understood in the IP resource policy making process.

Their requests for IPv6 management are focused on three areas: They want equitable access rights to be guaranteed to IPv6 resources now and in the future. And they want governmental involvement in the policy-making process. In addition to that, they want assistance, international assistance to them, to help them to raise awareness of the urgencies of IPv6 deployment and to help them to build up their capacities, in terms of human resources and also infrastructure capacity.

Our mandate comes from two resources. The WSIS Principles and outcomes and the decisions of ITU Membership. The WSIS here for those of you who are not familiar with it, is the Work Summit on the Information Society. And I will just mention some key words from this mandate.

The equitable access, as a principle for IPv6 management is enshrined in these outcomes.

And governmental involvement, this is also requested by the outcome of WSIS in paragraph 68 and 69 of the 20th agenda for the Information Society which recognizes a need for enhanced cooperation in the future to enable governments on equal footing to carry out their roles and responsibilities in international public policy issues.

ITU has a lot of resolutions, made at all level of ITU organs for us to carry out working in this area. These organs include those I've mentioned here: the Plenipotentiary meeting of ITU, which is a host organ; the World Telecommunication Standardization Assembly, which is for the standard sector; the World Telecommunication Development conference, which is the organ for the Development sector; and the Telecommunication Policy Forum, which is a summit for the regulators worldwide.

I will skip this one. It's on resolution 64. The chairman has already been mentioned the tasks we were instructed to carry out.

This request was further specified in the ITU-T Study Group 3 meeting. The reason why ITU was requested to be an additional registry of IP resources was raised. We were requested to evaluate this proposal and to report its advantages and disadvantages to the ITU Council in October 2009.

Our activities, I would briefly mention the two sides of our experience in the resource management. ITU-T and Telecommunication naming/addressing resources management, which is already a function of ITU with more than 120 years of history.

And we have key recommendations for telephone numbers, Internet and signalling points, and international mobile numbers. Here I want to especially mention ENUM as an example of ITU and RIR cooperation.

In this ENUM implementation interim procedure, RIPE-NCC is taking care of the registration of ENUM implementation, where ITU plays a role in verifying the holder of the E.164 numbers. We have good cooperation experience with RIPE I believe.

And the second example is ITU-R management of the geosynchronous orbital positions. Here is just an example of how ITU are managing spectrums and orbital positions with national allotments to guarantee equitable access to resources.

Someone who is interested in this, we can discuss later. I will skip it here.

And all this management, roles, and procedures are defined in regulation, which is a binding international treaty.

Work to implement ITU resolution, WTSA Resolution 64 includes the two studies, ITU interest to external consultants. We have done two studies. The first study was carried out by Nav6 USM Malaysia, which proposed a Country Internet Registry model. Professor Sureswaran Ramadass will talk about it later, so I will not talk anymore about this study. Because Professor Mueller is not coming for the second study, I will briefly introduce his idea in his study.

The reason to have the second study is because it was proposed that marketing economical factors might help to improve the efficiency of resource management like pricing, transferring, and marketing this kind of mechanism. Professor Mueller is an economist and he has done a second study and he came out with the proposal to have our transferable address block lease model, which will be a set of provider independent blocks, ranging from /48 to /32, which will be allocated to our ISPs, according to at our recurring annual fee, but there will not be any need assessment in managing this resource.

Also, as requested by Resolution 64, we're initiating a project to help developing countries, according to their request. The objective of this project is to understand their needs, raise awareness, encourage the deployment by participating and the strategy of this project will largely rely on partnership with private and public sectors in national and the international level.

Also, to implement Resolution 64, we have discussions with the ICANN and RIPE-NCC Secretariat and we have reached agreement to joint effort to help developing countries in awareness raising and training and capacity-building. And during these discussions, a suggestion was raised to have a global policy proposal, which will be drafted by ITU but it will follow the RIR policy procedures. To reserve an IPv6 block for future needs of developing countries, which are not provided through their RIR's process at this moment because they would like some long-term planning.

The decision of ITU Council is to set up this group, to do this job, as the chairman has already introduced. I will just mention that this group is open to ITU-T and ITU-D Membership, which includes ISPs and all industry members. Its first meeting will be in two weeks and all five RIRs, either Members or non-ITU Members are invited to this meeting, to this group as well.

So I will close my presentation with the observation and the conclusion. I think there is a willingness from governments to participate and contribute to IPv6 deployment. The individual-based RIR policy deployment process is widely perceived as open, transparent, and is bottom-up. And what ITU can bring to this process as of added value is ITU could help to get all 191 countries involved.

To conclude, I believe that Internet governance needs inclusive vision to address concerns of all stakeholders. International organizations like ITU, ICANN, and RIRs can cooperate and contribute according to their respective role. ITU IPv6 efforts try to find a meeting point for governments and Internet community in international Internet policy discussion. And we look forward to cooperation from RIRs and the whole Internet community.

In this slide, some additional links provided to those who want more information.

I guess I have to ask you to keep your questions.

MASATO YAMANISHI: Thank you very much for the introduction about ITU's concerning point, and also ITU's activity; and as I said, previously, is there any question for clarification? OK. Go ahead. You can go to my microphone without permission from chair.

BRAJESH JAIN (Inspectorate, New Delhi, India): This isn't my personal capacity. What you're suggesting that both Internet community providers today, and ITU keep allocating separately?

XIAOYA YANG: I think the detailed proposals, including the two studies that have been done by the consultants haven't been discussed in detail, in ITU. It will be discussed firstly in the coming meeting in two weeks. And I think a lot is still yet to be done by this group. So to your question. I think this is the first chance, and now we're discussing hearing APNIC and we will look for intense discussions here. I'm not going to I cannot answer your question right now.

MASATO YAMANISHI: Sorry, next presenter will talk about what is the proposal in ITU, so please do not ask question at this stage. And time is very limited. So if you have questions for clarification, please remain at the microphone.

PAUL GERMANO (Google): Real quick question in the beginning of the presentation, you used the IPv4 allocation as an example. I'm a little confused, given the much greater address pool size we have with IPv6 that you're using historical representation as an example that doesn't really apply in this case. It's very unlikely we would run in to the same sort of early adopter issues that we had with IPv4 in IPv6. I'm not sure that really, I'm not sure how that backs up the argument?

XIAOYA YANG: I'm presenting some opinions from one group and I think I hear another opinion again from you. I will have no comment on that.


LORENZO COLITTI (Google): Near the beginning, I saw concerns from developing countries that prices for IPv4 addresses would go up with exhaustion. Are there any details on that? As I understood it, the fee structure for RIRs, but I'm not an expert, was relatively flat. It depends on how much address space you have?

XIAOYA YANG: I understand you are referencing the current status. Our question is, will there be a scarcity issue of IPv6 in the future? And will that, the resource management policy will be tightened up and the price will rise because that's already happened in IPv4. That's all what we know for this moment.

LORENZO COLITTI: We're talking about I understand that I misunderstood but we are talking about something that will happen when IPv6 address space is exhausted, so whenever that is.

XIAOYA YANG: I think yes our concern is about the future of IPv6

LORENZO COLITTI: Because the solution to this issue is to move to IPv6. Yes?

JOHN CURRAN (President/CEO, ARIN): It actually follows the other two questions but it's a clarification of your slides. You know the public policy issue could exist if an IPv6 scarcity issue arises and we've also seen the ITU is willing to do studies you only mentioned two studies. Are there any studies with the ITU regarding the probability of that scarcity of IPv6 arriving, either studies or staff assessments? Because that would be fairly important to know about.

XIAOYA YANG: I think that's a very good proposal, because our work at this stage is, as I have reported in this presentation, and this work just started last year since February and we reported to the Council in October and the Council decision is to have this group to further study the issue. So if there will be any further activities, I think it would be a good proposal to propose to this group in two weeks and we can continue.

I think it will be as that gentleman has just asked, all the discussions should be based on real research and this kind of predictions, whether there is a council or not. We're not sure. But we have this concern at this moment.

MASATO YAMANISHI: Sorry for the interruption, there's still four people standing at the mike. May I limit? Any more people?

SANJAYA: This is coming from Jabber chat and he said it's question, not a clarification. So it would be more appropriate for later.

BILL WOODCOCK: Something that I'm afraid I don't understand from your presentation - you said that developing countries are afraid that the price of addresses would go up in a market environment, as scarcity becomes monetized. And I believe that you presented this as a problem - is that correct?

XIAOYA YANG: It's a concern.

BILL WOODCOCK: At the same time, you also present Mueller's proposal that proposes to do exactly that in the face of the actual bottom-up constituency-drived policy that everyone pays the same price based on need. So are you saying markets are the solution, the problem or both?

XIAOYA YANG: For clarification, first of all, these studies haven't been discussed in ITU Membership yet. But indeed in last year in IGF there was a workshop and this proposal was discussed. My understanding, if I can provide more information from you of that workshop is that this study, this proposal, was discussed and it was suggested more influence on the routing possibility of this proposal should be added and it should be further studied. That's my understanding. So, as I said, I'm not proposing any solution for the moment, especially that we only have these two studies done by consultants and is not discussed in ITU yet. There are two proposals and we wait to hear from the community how to improve, whether it's applicable and this kind of thing.

BILL WOODCOCK: I think one thing you might hear from this part of the community is that the ITU would be taking a higher moral position if it came out and said that pricing developing countries out of address space by imposing markets on them would be a bad idea.

XIAOYA YANG: I will not really argue with you about this. It's from the economist's study and he believes that marketing mechanisms will provide motivation for conservation. This is his idea.

MARTIN LEVY (Hurricane Electric): In your slides, you talked about the structure and how the ITU would deal with some of this. But I'm wondering, as in our case, a large operator of an IPv6 network, we have a lot of influence via a voting mechanism within inside this organization, within inside the equivalents to this organization, whether it by ARIN, RIPE, etc. The policy that comes out of those and the discussion has enormous input from the operations side of the business. That's vital. The Internet, unlike the telephone system, has a global impact. Anything that occurs on the Internet is essentially dealt with at a global level. I'm talking about routing and IP allocation.

MASATO YAMANISHI: Martin I'm sorry.

MARTIN LEVY: The clarification is, I wanted to understand how you would be as open to the community versus the Members that the ITU has at the moment? Because that's different. The Internet is a different beast than the telephone system in this regard.

XIAOYA YANG: I'm not sure whether we should open this - this is a question for clarification, it's a question really for discussion. I'd like to answer you that ITU its own membership structure. And I think, although many people here in this room may not like governments but governments are still there to represent our benefits from each of the countries. So if you have any concerns, I believe that the first channel you can channel that to ITU is through your Member State. And secondly, we have membership from private sector. We have industry players all the big telecommunication companies are our industry members. Any concern from them, like operational concerns, I think it can also be brought to ITU.

MARTIN LEVY: I will leave that for a separate question.

MASATO YAMANISHI: Please do it at the end of the session.

JONNY MARTIN: The 21 countries you mention that don't have IPv6 allocation at the moment s that of concern to you, as in the ITU?

XIAOYA YANG: Sorry, I haven't got your question? Did I mention any country haven't got it?

JONNY MARTIN: Sorry, there are 21. Have the ITU done anything to help those without allocations get allocations now? Because it's not hard.

XIAOYA YANG: I think that's what we are here for, and we want to promote its awareness of the urgency for IPv6 deployment. But before we're doing that, we have first to address their concerns at the same time. They are asking us.

JONNY MARTIN: If their concerns are that they don't have any address space, that's something that can be solved right now. They can go and do that. It doesn't need much discussion or any policy changes. It seems a bit backwards the way things are going.

XIAOYA YANG: Anyways, I think this is not a concern I raised in my presentation.

MASATO YAMANISHI: Let me interrupt at this point. Again thank you very much, Xiaoya. Our next presenter is Sures, who is the Director of National Advanced IPv6 Center, and he will present about a proposed, which is a proposed model in ITU. OK.

MASATO YAMANISHI: While we're waiting and swapping PC, let me say as a request or question. Please, 1, will you ask a question for clarification. Please point out which part is the target of your question. Otherwise, it is quite long. So please. Do you want to use your laptop?

SHARIL TARMIZI: Whilst technicalities are still being sorted out, can I just have a quick check in the hall. How many of you here work or represent some part of government in any shape or form? OK, one gentleman in the back. So it's quite interesting, when I go to an ITU meeting and I ask the same question, I get completely the opposite response! So, we're talking about communities. How many of you are sector members of the ITU? Anyone here? About ten. Less than ten, actually. Quite encouraging, I would say. See, in times of war, you have the court jester dancing up and down making sure that the people are kept entertained while this is happening. But, can I ask a more serious question. I think there were several points raised by Xiaoya just now, and as my co-chair just mentioned.

There were some points during the session that I would definitely urge you to keep for the discussion at the end. But I would just like to point out we are talking about different cultures, different experiences and different people looking at the same issue from different lenses. If I can use that term?

There's a government lens, there's a private sector lens. There's a civil society lens, and there's a business lens. So, in our discussion here, I hope that you also keep those various perspectives in mind so that when you come to the question time, this is something that we can engage in a meaningful debate and discussion. Sures, are you good to go? The machine doesn't like you!

Maybe I should also, I think being the only government guy on the panel, as a co-chair. I had a short stint with the IP community thanks to a lot of friends like Paul Wilson, you know, Ray Pawlzak and those guys, and also, a relative stint with the ITU side of the community. So I have some benefit of having been in both worlds. Sometimes I cannot help but notice that it is a bit like the story about cats and dogs trying to talk to one another. Again, you know, I don't like to use the term, which is why I keep reminding people about different lenses for different people and sometimes for different purposes.

Governments are quite often used to the idea of a first-come, first-serve policy as well if you would like to know. In the area of spectrum allocation, for example, at the national level, very often it is done on a first-come first-serve basis. Other forms of allocation of scarce resources are also first-come first-serve basis. One interesting point that was raised by, I think John Curran, was the issue about studies on the actual scarcity. That could be the elephant in the room that's there, but no-one is looking at it in a more careful or scientific manner. Are you good to go?


SHARIL TARMIZI: I'm running out of stories!

SURESWARAN RAMADASS: OK, thank you to the two co-chairs for keeping us entertained. I'm going to talk a little bit on the expansionary approach to the IPv6 address allocation model. And as all of you know, the Internet has evolved from a research-based closed network to a social network used by everyone. It has grown now to become the largest economy in the world. USM was appointed as a consultant to conduct a study to look into expanding the current IPv6 resource allocation model.

The researchers have been researching an expansionary approach scheme that provides greater choices to the Internet community, but still maintains the integrity, sustainability, and the routability of the Internet. Based on our studies, we find that the present system of IPv6 address allocation can actually be further expanded to meet the growing demands of the Internet community. Our model proposed to maintain the IANA/RIR, but further expand the system as well to give ISPs a choice of who their address block provider will be. This will potentially help to reduce the cost to the ISPs and in return ISPs can invest that money to further promote IPv6 connectivity, to potentially further expand closer participation in the IPv6 activities and policy development by local ISP Internet communities.

To meet the local needs of the ISPs and industry, especially by providing local language content, system, and trading. To better achieve conservation of IPv6 addresses as local entities and local ISPs know the local requesting organizations better, and to provide better support and awareness programs to help move the IPv6 agenda, especially for the developing nations.

This is the current model, and this is the current expansionary model. The peer entity could be at the regional or the international level. The peer entity should be a multilateral, multi-stakeholder international body that would ensure close coordination between the CIRs and the RIRs. This is the proposed hierarchy model, from an hierarchy viewpoint.

The CIR model really talks about new entities. Would serve in parallel to the current RIRs, thus providing ISPs greater freedom of choice for obtaining IPv6 address allocations. The policies followed by the CIRs would be in close co-operation with the leadership of the local ISPs, specifically to meet the interest to satisfy the local needs of the users. It would adhere to the technical aspects of the Internet, address conservation, aggregation, and registration currently practiced.

The CIR model does not disturb existing infrastructure, nor does it introduce any form of new infrastructure. There would be no additional fragmentation as our research and studies have shown this. Overall, numbers of prefixes added to the core routing table would still remain the same as well. As such, the expanded RIR model would not impact nor threaten global Internet stability or routability.

The CIR being closer to the user. The CIR as an organization would potentially be set up by an organization of local ISPs. It would be able to better satisfy the local needs of the local user, for example, multilingual local language support and localized Helpdesks. The ISPs could then obtain potentially cheaper, even free allocation of IPv6 addresses and this would really help the developing countries in the region grow their Internet. The CIR, proposed to be headed by the ISPs, would value-add to the RIRs and benefit the Internet users by differentiation of services. The CIRs would have equal participation in the policy formation and resource distribution so that Internet resource distribution and decentralization are more balanced, especially within their own countries. Implementing CIRs would facilitate a more equitable access to Internet resources, especially for non-English speaking countries, providing greater access to the Internet for everyone.

In conclusion, the RIRs have greatly contributed in the early growth of the Internet. This is valued and appreciated. But we have to move with the time, and in creating the next generation of the Internet an Internet that will eventually become more open and non-monopolistic; an Internet where ISPs have a freedom of choice and potentially, a cheaper Internet for ISPs and all users. The proposed CIR model will only work if, if openly and correctly discussed and implemented in the greater interest of the Internet ISPs and the netizens.

And what do we believe in? I have a little video that shows what we in USM believe in. I would like to show that to you all.

(Video plays)

Thank you.

MASATO YAMANISHI: OK, if you have a question for clarification, please come to the microphone.

OWEN DELONG (Hurricane Electric): I have two questions. On slide 10. On slide 10, you claimed no impact to the routing table or fragmentation of the address space. But I think that that conclusion is very subjective and very dependant on the nature of the CIR implementation and the policies that each CIR adopts locally. Do you have a comment on that?

SURESWARAN RAMADASS: We agree and we've clearly stated that the CIRs have to work hand in hand with the RIRs; they have to follow the baseline models.


JOHN CURRAN: On slide 9, I thought that I saw a contradiction regarding the same question if you could bring up slide 9 it would be helpful. I guess the question is, specifically, your policies adopted by the CIRs are claimed to be local and more focused on the country, and yet it has to follow the RIR. I'm trying to understand from your proposal what happens when a country proposes something outside of the RIR framework. How will global buy-in of the global routing impact be obtained? It is not clear from your paper or the presentation.

SURESWARAN RAMADASS: OK, the paper and the presentation talk of what is the fundamental policies that we should be following now. As to the actual implementation and how things are going to be done, that depends on the body that takes it on to implement. We just make recommendations purely technical recommendations.

JOHN CURRAN: I'm sorry; I didn't hear an answer to the question I was asking. My question is, are the Country Internet Registries free to make policy within the scope of the RIR that they're working with, or can they make policies with the routing impact outside of that scope?

SURESWARAN RAMADASS: Not outside of the scope. It should be within the scopes that have already been set within the RIRs.

MASATO YAMANISHI: I think that's a very important point, so let's discuss later again.

STEPHEN KENT (BBN Technologies): I was looking over your report and I see that the ITF is characterized as a research organization there, could you clarify why you think it's a research organization rather than a standards development body? Just trying to find out if it was a really large typo or a perception issue.

SURESWARAN RAMADASS: Okay, it's a standards development body. Typo.

DAVE CROCKER (Brandenburg Internet Working): I've always understood that a precept in making changes to an operational system is to only make changes that are essential. And I apologize if I missed this, but I don't think that I understand what problem needs to be fixed that your proposal responds to?

SURESWARAN RAMADASS: That is something that you need to ask the ITU. We are a technical proposal group and we wrote a paper based on requirements of Resolution 64. So we don't get into the details of that particular question.

IZUMI OKUTANI (JPNIC): I'm speaking as an individual, not representing JPNIC. My question is related to the question of the gentleman who spoke earlier. I'm not sure. What are the reasons that those countries who feel concerned that they don't, that they're not able to receive address space. I think we should do more analysis on why they feel that way and if it is just simply because they misunderstand about the procedure or is it the language issue? Or maybe if it is a criteria? I think we should have more information about these things and analysis before we come up with this as a solution.

MASATO YAMANISHI: Sorry, it is a comment, I think.


SHARIL TARMIZI: Perhaps you can hold that one and put it for a question for further study.

IZUMI OKUTANI: It's a question, so I'm interested to know, what are the reasons?

SURESWARAN RAMADASS: Chairman, can I answer that quickly. I think that the question is very good and something that should be further studied. I agree with you.

MASATO YAMANISHI: OK, Lorenzo and John.

LORENZO COLITTI: There's something I don't understand around slide 9 or 8. There is a diagram. Oh, before that still. At one point you said in an answer just recently that the CIRs would operate in their own country in accordance with RIR policies?

SURESWARAN RAMADASS: Well, we believe that the CIR can be a country or a community. It could be a cluster of countries.

LORENZO COLITTI: So if it is operating within the RIR policy of the RIR system, why there has to be a separate peer entity in the RIR system if these countries can be naturally reconduced to the RIR which they are already a part of?


LORENZO COLITTI: So there's no particular reason, just the way we would choose to do something.


SHARIL TARMIZI: Just to clarify, I think your question is probably concerned about double dipping, but maybe we can come back to the double dipping issue.

LORENZO COLITTI: No, just a question of, if the policy naturally flows through a particular body and then is received by downstream entities, so to speak so if the CIR is a part of, as regards to policy flow or conceptually inside an RIR, why they have to "report" to a parent entity outside.

BILL WOODCOCK: There's a confusing extra box on the slide. I think that was the point.

JOHN CURRAN: I have a question regarding slide four and am having trouble reconciling the bullet with what you said earlier. This will potentially help to reduce the cost to ISPs and in return, ISPs can reinvest money. I want to know about the scope of the study you conducted. Is it a technical scope? Because as a cost statement, and I guess I'm wondering, is there an assumption that it is lower cost and if it is, where did that come from?

SURESWARAN RAMADASS: You're right it's an assumption that the cost could go lower.

JOHN CURRAN: And the assumption is based on the CIR model not having anything to do with the RIRs that it's relying on for global policy?

SURESWARAN RAMADASS: No, it shouldn't. It's an assumption that basically says if it opens up to different basically by opening up and allowing other people to allocate address, competition may come up and this may help to reduce the cost.

MASATO YAMANISHI: Sorry, let me limit at this point. You can ask your question, but no other people. Go ahead.

BILL WOODCOCK: I had a question that I'm going to defer for a moment, because that last extreme so confused me. So, you're assuming that if a third party, a governmental body sets a price for a service

SURESWARAN RAMADASS: We didn't say governmental body.

BILL WOODCOCK: OK a third party. A non-RIR body sets a price for a service that they will set a lower price than the constituents of the service set themselves? That is if I and other ISPs collectively decide what fee to assess ourselves for IP addresses, you assume that that will be a higher fee than another party will decide to charge us?

SURESWARAN RAMADASS: Actually, you're wrong. I said that the CIR could be a collection of ISPs. And should be a collection of ISPs. If you look at it right there. The CIR headed by ISPs.

BILL WOODCOCK: OK, that's what RIR is?


BILL WOODCOCK: What is an RIR, if the RIR is not the collection of the IP address stakeholders?

SURESWARAN RAMADASS: RIRs represent regions, CIRs can represent countries and communities.

BILL WOODCOCK: No, RIRs do not represent regions; they represent IP address users within regions.

MASATO YAMANISHI: Bill. Let me say my comment. Also the difference between NIR and CIR is not so clear in my understanding so let's discuss that later.

BILL WOODCOCK: The question I originally got up to ask was, if you're saying that there will be no additional policies outside of the existing framework, that would cause problems like, for instance, impact the routing table, then it seems that this is a no-op, right? I mean, what will this do if you can't actually do a policy that affects allocation, other than what allocation normally is already.

SURESWARAN RAMADASS: It is a policy to increase competitiveness. That means if I am an ISP, I should have more than one organization to go for to get addresses. If I have, very simple if today, the five RIRs can allocate cross-boundary addresses, this problem is gone. Because you could go to any one of the five RIRs and get an allocation of addresses. And you want to know pricing? If I go to AfriNIC and got my IPv6 addresses it would be free today.

BILL WOODCOCK: Does your study

SURESWARAN RAMADASS: Let me finish. Since I have to go to APNIC and get my addresses, I have to pay them an annual fee. Now, why the difference? Why can Africa give it for free and why is APNIC charging me for it? And you know what, I'm one of the people who helped to create the IPv6 standard and we paid money to help create it. And now, I have to pay money to get back a block of addresses that I helped to create? Does that make sense? Anyway, well, that's a personal opinion. Next.

MASATO YAMANISHI: Sorry, may I go there.

SANJAYA: I'll leave it to the chair to determine whether this is clarification or for discussion later. For clarification of Tim McGuiness if the CIR provides all of the services you mentioned, wouldn't those costs need to be recouped in fees to RIRs, therefore pushing cost for IP address administrations higher for those in developing countries, in relation to the current model of distribution?

That's one question. That's one question. And then another question from Matthew Moyle-Croft: The proposal seems to try and soften economic problems. If the issue is the cost to developing countries, why not fix that as funding to the RIRs?

Third question from Elliot Leer: Please clarify what is the difference between CIR and NIR? Thanks.

MASATO YAMANISHI: I'm not sure that the second one is a question or not?

SURESWARAN RAMADASS: The first one was answered by John when he asked me, are we doing an economic study? We're not so the economic part is out. The second one, I don't even remember. My memory is bad. The third one was the difference between a CIR and NIR. Is that correct? It's pretty close, both the CIR and the NIR. It's again how it is to be run. But not much differences between the two.


SKEEVE STEVENS (Intelligo): My question was actually based around the third one, what would you say are the differences between the CIRs and the NIRs. But, given the bottom-up policy structure that the NIRs do, what are any of the disenfranchised countries' problems with creating their own NIRs at the moment? And if they've got a problem about which region they're getting allocations from, the whole RIR process is designed around policy and working together and being able to change those policies within this community. So why can't you just work within the current framework at the moment to create your own NIRs. If you want to create a NIR in India that talks to AfriNIC, you can propose it. And if the community accepts it, there's no issue there.


SKEEVE STEVENS: So why are you doing this, then?

SURESWARAN RAMADASS: We are a technical study. The technical study clearly states you could do exactly what you just said. Earlier, other people were saying that you could go to another RIR and get a block of addresses, you could possibly screw the Internet routing table. This technical study shows that that doesn't happen. So thank you, you actually clarified, made the statement that I was trying to make very clear. If we open up to allow IP addresses to be obtained from any of the five in other words, if I could go to John and get my IP addresses from ARIN, then, you know what, you have then a democracy. You have a non-monopolistic environment. Because currently I only have one. Yes, John.


JOHN CURRAN: Just to clarify, you said a moment ago, that the study shows that it is possible to open up ISPs to additional sources of address space and not destroy the routing table. Is that the summary of the technical study?


JOHN CURRAN: And in the case where you create an abundance of Country Internet Registries, and current existing multinational ISPs that currently go to one Regional Internet Registry, or they go to five. Because they're truly global in nature, so they have to go to all five. The case where the routing table isn't impacted is the case when those global entities don't participate at all in the alternative scheme you're creating. Is that correct?


JOHN CURRAN: Thank you.

MASATO YAMANISHI: Thank you very much for Mr Sures, and thank you very much for the cooperation of all of the attendants.


MASATO YAMANISHI: And the third presenter is Adiel who is the CEO of AfriNIC and he will present about the developing country's approach for IP address management.

ADIEL AKPLOGAN: Thank you, chairman. Thank you very much. I will briefly talk about my experience of running AfriNIC, which basically serves the biggest concentration of developing countries and give my perspective on the issue of IP address management and the Internet development in general. I will start by saying that IP addresses are important for the future of Internet development. That is a fact, and that is something that governments and policy makers in developing countries start to understand. And that naturally raises some questions and some concerns and a willingness for them to understand really what is going on behind the IP address management. So for us, it is something natural. But what we have also noticed by working very closely with many governments in our region is that the real problem is their awareness of the Internet IP address management system.

That makes them anxious about their future. They don't really understand what they're doing and nobody tells them really what we do. What we are doing here today, discussing IP address management and discussing policy development. How they can, as governments, propose the policy. How they can voice their concern within a process which exists and that's what we have. We have noticed. And we believe that the main way of solving this is to work very closely, as technical community, as an RIR, to work closely with Government to collaborate with them and help them to understand how policies are developed. How they can participate. Hold their hands and talk to them the way that they are used to. I think Internet is a culture. It's a culture based on some technicalities.

But we need, when we are talking to governments, to adopt a language to the political environment so that they understand what we, as technical people, what we say and understand easily. So for me, I think this whole question about CIRs bringing IP address management close to people by creating more local registries is not really the problem. Because I bet to go to many developing countries, government websites. Regulatory websites, where you can find as much information that are available on the RIR websites.

The real problem is to bridge the gap, to create a link between government concern, government anxiety about the Internet future and what is really being done. Nothing prevents governments today and that's something that we're doing with several regulatories in our region to replicate information about IP address management on their websites. On their local information dissemination campaign so that operators, which are not already LIR, can understand so that they can support already ongoing training and awareness campaigns that RIRs are doing today. We need to put more emphasis on collaboration and cooperation than competition. I hardly understand how competition in the allocation of the IP address system will not harm the stability of the Internet. How competition can allow the community to be in charge of policy, to be in charge of defining the rules for IP address management in general. So I think we need to refocus on really Internet for development when we're talking about developing countries in general.

Policy can not be the same in a system favors competition. As soon as you open the door for competition, it will be difficult for you to maintain the same policy because when you talk about competition, you talk about getting to the point where one offers some advantage that the other can not offer, and as soon as you are going down that road, you're deviating from some common sense in terms of policy and the stability becomes subjective. Everybody can see stability from their own lands, can see policy from their own approach.

At AfriNIC, we have, we engage ourselves into our awareness and collaboration with government and regulators. And what we have noticed, as I said before, the huge problem comes from lack of awareness, lack of understanding of our process, lack of understanding of multi-stakeholder bottom-up approach in the policy definition. We have been asked several times, the question about not repeating the early adopter issue for IPv6, and we have a very simple answer for that. Today all of the five RIRs have received an allocation of a /12 equally. AfriNIC, the smallest RIR, got the same amount of that allocation, which we have. Which is already there. So the risk for that to happen is limited, because the community itself is putting in place policy that allows you to solve the issue.

And if governments still have concerns, they can participate in that policy development process, and now is the time for them to do that, so that all of their concerns are addressed. We have not even touched that /12 yet. It has not been allocated by IANA based on the current need of our region, but it is a policy that says, this is a new protocol. We need to give all the RIRs the same amount of IP addresses. Now, those who exhaust those /12s will go back and receive more, if in our region, we have not even started to use it, it is very difficult to go and request more.

The major concerns for governments are sometimes valid. But the solution doesn't always come from a fight for control or a top-down approach in managing the resources. We are committed in our region to continue to work very closely with governments; to give them all of the tools that they need to be more active in our community; to continue to share information to create an environment for them to feel at their ease within the technical community, so that together, we can address their issue and they can also understand the issue which are purely technical so that this Internet that we are concerned about, that's becoming more and more an economical tool stays stable, continues to grow, and continues to be a tool for development. Thank you.


MASATO YAMANISHI: Thank you very much for a good presentation. Are there any questions for clarification? It seems not. OK.

Let me go to the last presenter. Last one is Save. He was a policy development process manager in APNIC, and now he's working as ICANN's representative in the AP region. So he will talk about relations between ICANN and RIRs. Go, please.

SAVE VOCEA: Good afternoon, all. I will just read out one of ICANN's comments. I think most of this you would have heard already. Where does the demand come from? What we are hearing here is that some countries are concerned that late adopters will find there is very little space left, so I want to make sure that the country has sufficient addresses for its network and that they are not controlled by anyone else. The reality is that to this day, the IANA...

MASATO YAMANISHI: Sorry, are you trying to show something?

SAVE VOCEA: No, no, I'm just reading. The Internet Assigned Numbers Authority has only allocated 0.146% of the total IPv6 address space to the existing RIRs. So there's still a lot of IPv6 addresses to go around for many years, that's like 99.854% left to be allocated. What we believe is that there is no evidence that there is a lack of ability to get IPv6 address. For instance, in the APNIC region, its IPv6 allocation policy allows any Member today to receive IPv6 address space. A Member just has to send in a request and no proof of need is required at this stage. The main purpose is to provide incentives to get people, entities, companies, to use IPv6 and to promote transition into it.

In the ICANN's multi-stakeholder model where governments are free to participate in the Governmental Advisory Committee, or the GAC, the Number Research Organization and the Address Supporting Organization do provide update reports and statements to discuss IPv6 matters and that is always welcomed by the GAC.

The RIRs manage the consultation on the global policy which then comes to ICANN. So essentially, they will manage the consensus process. The multi-stakeholder model with its open and bottom-up process affords the opportunity for all to engage and participate, including governments. Existing policies state that if a new RIR is to be recognized, ICANN must allocate a /12 to that RIR. The mechanisms by which a new RIR is recognized are well documented. ICANN believes the bottom-up policy definition process is the most appropriate way in which Internet-related policies are developed.

In the international discussions about these matters, it is important to establish objective diagnosis in the case where there are findings that there are some concerns or even problems. It is important to learn if they can be solved with existing mechanisms, and if the current mechanisms allow for change. ICANN maintains that the current system of allocation has demonstrated strong resiliency and that it is adaptable to change. If there would be a problem in the future, there seem to be enough mechanisms to discuss possible solutions within the current system.

The ITU could play an important role here and should be encouraged by its members to learn more about these processes and to participate in them. Any means of improvement to any system should demonstrate value or additional positive contributions in order to be implemented and prevent any non-intended consequences.

So in ending, ICANN is grateful to have the opportunity to participate in this community consultation and shows clear support for the RIR system, and it will support the RIRs as they discuss these issues with the ITU. Thank you.


MASATO YAMANISHI: Thank you very much Save. Are there any questions for clarification? It seems not.

LAWRENCE HUGHES: There were two concerns of voice. One is whether we may run out of IPv6 addresses some day and look at unfair allocation in v4. The other is a desire to do jurisdiction shopping, like you're looking for a judge or a jury that may be more amenable to his case. And the second, I don't think is a legitimate concern with the ITU. The first, I think we can address with a quick calculation. Even in the /3 that's been allocated currently for giving out one eighth of the concern address space. If you take the standard allocation block, a /48 and the number of people alive, there's some 5,000 allocation blocks.

MASATO YAMANISHI: Is there a question?

LAWRENCE HUGHES (Infoweapons): There seem to be two concerns voiced here: one is a concern as to whether we may run out of IPv6 addresses one day, and run into the same kind of unfair allocation that has happened in IPv4; the other seems to be a desire to do some kind of jurisdiction shopping, like a lawyer looking for a judge or jury that may be more amenable to his or her case. The second I don't think is a legitimate concern of the ITU. The first, I think we can address with a quick calculation; even in the /3 that's been allocated, for giving out 1/8th of the current address space if you take the standard allocation block, a /48, and the number of people alive, there's some 5000 allocation blocks for every human alive. So my question is, has anybody made that calculation yet, are you aware that there's really no danger of a shortage if there's 5,000 /48s for every human alive, are you still worried about running out?

MASATO YAMANISHI: I think it's not a question for Save himself. So let's discuss later. And the chair of the next session and the APNIC Secretariat kindly agreed to extend this session a little bit. So I would like to take a ten minute coffee break, and after that coffee break, let's resume discussion and... the discussion time is still limited in 30 minutes. But please do not leave here. I still have some comments. Can you show the chair's slides.

OK, I think the CIR model, which Sures mentioned actually contains several points so I made some slides of what is the concerning point and also, what is the major discussion point of CIR model. So during the coffee break, I would like to show that slide. So my idea for the discussion is dividing the time for each topic. I would like to do so, if you would agree. But anyway, let's consider based on my slide.

Also, when discussions will start, I would like to ask how many people would like to ask a question or would like to say comment. No, not now. At the beginning of the discussion! I would like to ask again. So please raise your hand at that time and then I can, I'd like to try time control. To do time control. Anyway, let's enjoy the coffee break. And ten minutes later, please come back here.

MASATO YAMANISHI: When discussion starts, I'd like to ask how many people would like to ask a question or would like to say comment no, no, not now at the beginning of discussion I would like to ask again, so please raise your hand at that time. I'd like to do time control. Anyway, let's enjoy coffee break and ten minutes later, please come back here. Thanks.

(10 minute break)

SHARIL TARMIZI: Has anyone seen Professor Sureshwaran? Where is Adiel? He's on the way somewhere. OK.

MASATO YAMANISHI: Even though one panellist can not yet come back, since time is limited, we would like to restart the session. Now it is 4:52. So we would like to finish this session until - anyway, until 4:22. I'd like to, I'd like to try - people can say their comment as much as possible, but maybe some people cannot say, cannot take a chance to say some comments. So for such case, you can send your comments to IPv6@apnic.net before 9am of this Friday. So Adiel has come back. So let's go back to the discussion.

Sorry to show my personal view for this topic, but I think there are two concerns exist as a background of this topic. First one is concerns for current IP address allocations, as Xiaoya presented. And the second R second one is concerns for economics of IPv6 address. And then proposed CIR model includes also five major points, I think. The first one is the allocation policy. So the current way is request-based allocation, but the proposed one seems to be a growth-based address partitioning algorithm. And the second point is, NIR or CIR as we talked in previous session the difference between them is not so clear. So I'd like to discuss this point in the session. And the third point is RIR scheme can only or competition between parallel schemes, or parallel policy, or parallel organization. Both aspects should be included in this topic. And the fourth one is ITU will become one of the Registries, so people made concerns for neutrality from regulator. So I think this is one topic which should be discussed. The last one is policy decision process. Actually, both side stays open, but actually the Member State is a little bit different. The discussion in RIR is community based. So everybody can attend a discussion, even though their organization is not a member of RIR or NIR, like that.

However, discussion in ITU, yet if discussion in ITU is Member State or sector member based. So, if your organization is not a sector member, you can't attend a discussion of ITU. And also in some meetings, only Member States can attend to the discussion. So, actually, policy decision process is a little bit different. OK.

So is there any major discussion point? If you have, please come to the microphone. OK, go ahead.

OWEN DELONG (Hurricane Electric): I don't have an additional discussion point off the top of my head, but I do think attempting to run through these points in some sort of order may artificially constrain the discussion and prejudice it and I think that could be to the detriment of the community and the discussion.

MASATO YAMANISHI: Very useful comment. What I'm trying to do is, since we have 30 minutes, so if you will agree, I would like to spend 10 minutes for background concerning points. Concern for current IP address statistics and concern for economics at the basics address.

SHARIL TARMIZI: Thank you. I think as my co-chair said, this was sort of a very quick attempt at trying to distill the issues. But I think it certainly is not exhaustive and is not intended to be. But if the participants, all of you are keen to have a more free-flow type discussion, with hopefully with some measure of a structure at the end, I think we're also quite alright with that. I think the idea is to get a sense of what this community feels about the proposals that have been put forward and also a sense of what it is that we think is lacking in the current thoughts around the issue? Whether, for example, we heard quite a few commenting about whether there really is scarcity being a real issue or is that a 50-year-away kind of question as opposed to something that needs to be looked at immediately?

Perhaps maybe we can invite some of the panellists to comment on some of the other presentations that have been made here to sort of kick-off the discussions? Many of you had very, very good points, I think you raised. So I think we have to stop you there because the earlier part was about clarification. But now is the time for the discussion, so can I invite any of the panellists who have initial reactions to some of the earlier four presentations, please?

XIAOYA YANG: I think Save, I'm not sure I'm pronouncing your name correct, talked about the ICANN document submitted to ITU already as a contribution. Actualy, it talked about it touched the question whether scarcity is an issue. Also Sharil has mentioned this really is the elephant in this room. The difficulty I perceive now is it is very difficult for anyone to give a yes or no answer for this moment. This is just my observation. And I think as John's proposal, and also the other gentleman has proposed, it really needs to be further studied. If anyone can come out with a concrete study and convincing the other side of the opinion that whether there is a problem or there is not a problem, would be a lot of progress to solve this whole mess here.

And, also, I have another comment about Adiel's proposal. And Adiel, I think he's very good in summarizing his experience in AfriNIC, and his cooperation with the governments. And I think he mentioned that in the policy process is very important to have dialogue between the governments and the Internet community. And also the chairman mentioned in the slide, the last point, there is the issue of policy process and there is a kind of counter confliction between the two camps and between the way they work. But I would just like to call to the attention of the meeting here that each has their own unique constituency. And even though the governments, we might not like them for this and that reasons at national level, in the international level, governments are still an important entity to represent the concerns of the country. So that's very important. I also want to mention that a fact that it's very difficult for government representative to participate in the Internet policy discussion, because, as you could acknowledge if I'm representing our government, I cannot speak on my individual behalf.

And, also, when we challenge, there is a gentleman's question about how we can get expert participating in the ITU process or something, a question like that. It's more for opinions, than a point for clarification. It's also about a point that although ITU have private partnerships experience for so many years, there is still a lot to be researched to open up those sovereignty-based voting procedure, to individuals and to the civil society.

We have to really to be innovative work together to find a solution for that. And I think that again comes back to the same conclusion as Adiel, we have to keep in dialogue and you have to have the willingness to cooperate from both sides. Thank you.

SHARIL TARMIZI: Any other reactions? OK. Thank you, Xiaoya, for that comment. I just wanted to sort of kick-off and start some discussion by maybe sharing with you some of my own perspectives. I now work for the government. I'm the Chief Operating Officer of a regulatory body, which is independent but has policy direction from a Minister, within the Ministry. The functions that we undertake, we manage scarce resources on a daily basis. In the traditional sense of governance, which is managing spectrum resources which generally I think is accepted as scarce managing numbers (E.164 numbers), which is also generally accepted as scarce. And we also manage the local coordination of, as well as international coordination, of orbital slots, together with countries that we have to have bilateral arrangements with through the ITU and so on and so forth.

That's one part of the experience I bring to the table. The other is I also have been engaged with the Internet community since the year 2000, almost from the early days of ICANN, right up to the chairing the GAC for about four years. The earlier point I made about looking at things from different lenses and different lenses for different people is something that I need to articulate again because, let me share with you a problem I faced as a government official during one of the World Summit processes. I'm an accredited Government official to those meetings. I had occasion to deal with one of the UN reps, I will not mention his name. It's related to the IGF at that time, if some of you have participated. And I said to this person, the GAC represents the opinion of a large number of governments on Internet-related issues, particularly the DNS.

The response that came from this government official was, "Oh, you people in the GAC, you are technical." Now I can assure you nobody in the GAC was technical. But they said, this person who is a diplomat, actually, said "Oh, you people in the GAC are all technical. And we, the people at the UN, understand this issue of public policy." So you can imagine what my reaction was when we had an eclectic collection of government officials in the GAC. Of course we had an interesting mix, which is nowhere else compared in any part of any governmental organization anywhere in the world because we had government people who were scientists; we had government people who were economists; we had government people who were lawyers, who came from different Ministries.

Contrast that with an organization like the ITU, which largely, traditionally have people from the Telecommunications Ministry. You may have heard comments made by the keynote speaker to the APRICOT meeting about IP versus telecommunications. A lot of it is, again, approaching maybe the same coin from different sides.

So coming back to this official from the UN who said to me, "I wasn't government enough," from his perspective. This is what I mean and I try to describe by the layers of complexity of governance and the different lenses that people apply.

Coming from a developing country myself, Malaysia, switching hats now, a lot of thinking that we have here in Malaysia is actually around the areas that Adiel mentioned, which is engagement, capacity-building. Governments don't know what to do. And they fear what they don't know. How can this community, for example, help that? Help them overcome some of that? Adiel at AfriNIC have done a bit about that. And Paul Wilson around APNIC has done quite a bit of that. He attends government meetings and Adiel attends government meetings, in and out, just to build capacity. From my own personal perspective, it's not question of whether it's a CIR, NIR, RIR scheme, ITU, this, that and the other, the more pressing need for developing countries is if we need to move from a v4 situation to a v6 situation, how can we get there?

So let me just sort of try to kick-off for the discussion, for the debate, just so you're all full of it. But now it's a bit slow.

MASATO YAMANISHI: Thank you very much for your useful comment, Sharil. I think it's good chance to hear voice from the community. So I'd like to take a time as much as possible. Is there - oh, before starting, could you, I would like to count how many people want to say some comments for these topics? So could you raise your hand if you have some comments? Actually, it is less than I expected. It's good. Maybe less than 20, I think. Let me limit one minute for each person. And also there are many non-English native speakers, including me. So please make your comment very slowly and clearly and simply. OK.

So I'd like to separate discussion time a little bit because there are many discussion points. So I'd like to discuss these concerning points which were shown. Is there any comment or question or something for these concerns behind this topic? OK, OK, no - I can't use the pointer. As I said, I think there are two concerns behind this topic. OK. First one is concern for current IPv4 allocation statistics from developing countries as Xiaoya presented. And the second concerning point is economics, or pricing, about IPv6 address which somebody mentioned in the previous discussion time. OK. Let's go to microphone freely.

DAVE CROCKER (Brandenburg Internet Working): The two questions you present are reasonable but might not be the best ones we could pursue. Let me explain why.

There is nothing that is going to change the IPv4 allocations in enough time to be useful. And it appears to be clear to me, at least, that that history of IPv4 allocation speaks to the past and that IPv6 allocation has enough differences to it, so that that past could be misleading for the future. The question on the economics clearly is relevant to the future. But I frankly think the single most important suggestion that we've heard was the one we just heard, which was that what people should be focusing on because it is urgent is how do we get to v6? And it has special challenges for people from developing countries. Based on my own experience, what I do know is, if something is urgent, that creating new infrastructure administrative mechanisms does not help. And, so, perhaps the question that ought to be on the table is within the existing structure of RIRs, how do we facilitate getting developing countries up to speed with v6 as quickly as possible? This is a completely different topic than this session has been designed for but I suspect it's the only important one.

MASATO YAMANISHI: Completed topic means out of scope, you mean?

DAVE CROCKER: I believe it's in scope for what we should be talking about and out of scope for the way the topic has been designed. And I think I understand how the design of the topic has developed over the years. It has a very interesting history but let me suggest that we may need to change the nature of the discussion.

OWEN DELONG: The greatest value of the Internet, in my opinion, comes from its ability to democratize communications across all levels and types of societies. A bottom-up process for policy development is key to that nature and that ability and that value.

Second point, the view that CIRs would be run by the country's ISPs is not how the ITU has operated in other areas traditionally and my concern would be that the CIR process would devolve to be much more similar to the ccTLD process of history. And I think that if we do that to the address space, it's pretty ugly.

Third, regarding scarcity, everything that's been currently allocated is from 1/8th of the total address space roughly. If we run out of that in less than 50 years, the simple change of eliminating stateless auto configuration and going to a 96-bit standard prefix for the remaining 7/8s of the address space would multiply the number of networks by four billion in the remaining database. I don't think there's a scarcity issue in IPv6. It's completely artificial. F-U-D: fear, uncertainty, and doubt.

MASATO YAMANISHI: OK, thank you very much. Before asking answer for the panellist, it seems that many people have a comment for the whole picture, so I change my mind. In this time, please ask the question for whole picture, OK, not each topic. OK. Is there any comment from panellist for previous comment? No? OK.

SANJAYA: Question from the Jabber room by Tim McGuinness: Given I as an IP address user can participate in RIR policy development in an open way, how could the ITU guarantee that I can participate in CIR address policy processes without paying ITU sector fees? Second question is probably related to the economics of IPv6 from the same person as well. Given that the TABL study specifies a higher price than RIR costs at least initially for TABL blocks, could you clarify why this mechanism would be useful for developing countries?

MASATO YAMANISHI: Xiaoya go ahead. About 30 seconds. Go ahead. OK. So you have a comment for second question?

XIAOYA YANG: Yes, I think the second question is about the TABL model. And I think we need to further discussion it as it is discussed in IETF. It's a concern that the routing sustainability issue should further be studied, and also the pricing model. The price we initially set is also a question. I cannot remember the details of that study but I believe that the answer to your question is a need to be a trial period to find out the right price and for developing countries, this is also mention that's more for conservation of IPv6 resource for the long term and it might not directly solve the problem of developing country issue, for solving some of their concerns.

MASATO YAMANISHI, Sures, if you have some comment from the first question? OK Go ahead.

STEPHEN KENT (PBN Technologies): This is a directed question to Ms Yang, and it's really a clarification that I should have asked in the previous session. You cited ENUM as an example where the IETF and ITU have worked together and ITF transferred responsibility for ENUM to ITU. Just again for sort of level setting, do you characterize the ENUM work and deployment as a success? So far now that ITU is responsible?

XIAOYA YANG: I don't think I will say any comments about ENUM implementation itself. This is indeed the initiative from all over the world, from this community. And I'm saying that this is a good example for ITU to cooperate with one of the RIR. This is experience. I think ITU is satisfied with this experience in collaboration with RIPE-NCC.

STEPHEN KENT: I will be satisfied with the "satisfied with the experience," OK.

JOHN CURRAN: If I understand the chair, we're now on the general topics, not the specific items. I don't know the time limit but I'll keep my comments as succinct as possible. People have spoken about conservation and I do believe a study would help there, or at least someone to show up with one hypothetical example that indicates that a conservation issue is even theoretically possible, which it may not be today. I actually believe that Sures referenced in his paper, a reference of 1% utilization over 50 years using very, very generous numbers. I haven't seen anyone show up with a theoretical usage of IPv6 that gets us past single-digit percents in any time frame. This is about as theoretical as it gets until someone can show up with a counter-case. That's conservation.

Regarding policy and useability, we're not talking about addresses, because addresses are just numbers. The most important part about addresses is the ability to actually make use of the addresses in the Internet. And that means they have to actually be routable. And it turns out address become routable through a magic process that everyone here knows but we didn't really discuss. The ISP community trusts the RIR community to implement policy in a participatory, wide-open fashion, not a membership, no requirement, any participation is allowed. So that at the end of the day the ISP community is willing to accept the consequences of the globally developed policy in each of the RIRs on the routing table. This has huge implications and it's not assured by any other process. We do not know whether or not the CIR local policy process, or an ITU process, will be seen as participatory enough by the ISP community.

And the fact of the matter is addresses assigned by a CIR won't necessarily be usable for anything unless this is the case. No country could regulate or community and force global ISPs to follow their policy. So essentially we've set up a system where in order to make good policy, you have to have global interoperability, bottom-up policy setting, which was expensive.

Last-item cost: the cost of doing that policy discussion, which happens among the RIRs, including between the RIRs, is very expensive. In order to have addresses usable, on the Internet, that discussion will need to occur. And it will need to be borne by all the registries that are involved. To this extent, when people are looking at the cost, of alternative registries for IPv6, I don't know what number they're putting down for those registries' participation in the global policy process but they'll need to participate if they wish to be routed and that participation will have cost. If all we do is duplicate this process, it will become very expensive for everyone and there will be no lower costs. I don't think we have lower costs coming with this, I don't think we have aggregation coming with this, I don't think we have conservation coming with this and those are the three principles of the RIR system. Thank you.


MASATO YAMANISHI: Thank you very much again. Again, please make your comments shortly otherwise you'll miss the social event today, OK? Go ahead.

RAPHAEL HO: I'm a concerned individual and I'm just very happy that there are people on stage looking after the concerns that we're going to run out of IP addresses in the very distant future. I, as a concerned individual are very concerned about running up E 164 addresses. Would the ITU consider delegating this to a CIR-type model that we can assign the E 1 64 addresses ourselves?

XIAOYA YANG: My answer to you is that E 164 doesn't have a scarcity problem, It's a flexible lens.

BRIAN CARPENTER: I live in New Zealand and I'm speaking as an individual. I first heard about this idea that we've been discussing, I think, in late 2004. In early 2005 I became chair of the IETF and my very first discussion with Zhao Hulin, who was director of ITUT at the time was about this exact topic where I explained to him a number of reasons I didn't think it was a necessary idea. And I thought the idea had gone away until quite recently, by the way. I simply observed that since he told me it was an essential mechanism for the developing countries in 2005, AfriNIC has been created and has been given a /12 and I don't understand what the problem is?

KUO-WEI WU (NIIEPA): I'd like to mention one thing. Since last year we're, we started monitoring the CCTLD new server, the v6 operations and I can give you a number. More than 100 ccTLD new servers do not have v6 turned on. So, how you can make the v6 running? And second of all, even for loads of the turning the v6 on the ccTLD server, some of them is not stable. It's up and down. So I think we don't like to make a v6 assessment for this and ITU can play around but if they ask those member states, make sure your ccTLD new server turn the v6 on.

PAUL WILSON: Hi, it's Paul from APNIC. If it's OK with the chair, I'd just like to address the NIR versus CIR question, with some clarifying remarks.

MASATO YAMANISHI: Paul, actually, I'd like to take a time for that topic later.

PAUL WILSON: Sure, OK. I'll come back.

AXEL PAWLIK (RIPE-NCC): I have a somewhat uneasy feeling. We are having a great session and I really like it. Lots of brain cycles go in to discussing quite concrete proposals, studies, papers that seem to address a problem which I don't understand. I would like to take a step back and ask the question, which problems are there that we are trying to solve, and if that is not clear, then we should do some research into that? If we are seeing problems, then I think everybody in this room and the whole technical Internet community and loads of people are very committed to help keeping the Internet run smoothly. So I'm sure if we see problems, we are happy to help, by whichever way. But the question is what is the problem? If there's no problem potentially, then I'm all with Dave here, and Brian basically, saying then we need to talk to each other more.

And we do as RIPE NCC, we do regular round tables with governments and regulators. We do feel it's difficult to get to them, to find them, the people who have the concerns. So maybe there's some way of interaction between the ITU and ourselves? If you go and get your regional offices to get the regional governments together who have some concerns, then we're happy to fly around the world and do outreach and capacity-building. Absolutely. That's what we really want to do. Thank you.

MASATO YAMANISHI: Thank you very much. That's the reason why I asked the comment for background problems. Anyway, may I limit at this point? So there are three people remaining for the general topic.

KENNY HUANG (EC Member, APNIC): It seems like we're heading in to two different directions. One direction is we are going for the research and the other direction is the action for calling policy development process. Basically, the two directions required two disciplines and two totally different practices. If you are going for research, it will probably require validity and significant testing. That's regarding to doing the research, what people doing research in general. If we are heading to making policies, that somehow the proposal itself needs to demonstrate significant interest from the stakeholders of, even coming from the stakeholders. That wasn't shown in the research report, that's what I saw. Thank you.

MASATO YAMANISHI: OK. It seems Adiel has a comment. Go ahead.

ADIEL AKPLOGAN: My comment is trying to refocus what we're doing here. I think this whole meeting comes from the fact there is a working group set up by ITU which will be meeting in two weeks from now to discuss about a few issues among other, for the outcome of the two reports which has been presented today and the possibility of the ITU to become RIR to take care of some of the concern of the developing countries. We're seeing the two studies don't actually solve the issue of the developing countries. CIR, now the trade-based allocations of those issues. So the issue is more about, for me, from what I hear, is more about creating an environment where the technical community and the Internet community and government from ITU can discuss concerns.

Discuss issues seen from both sides. So how can ITU help in that? In one of the presentations of ITU, it says that how ITU can help bring the 191 Members, the government members, how concretely can we do that? It is something, probably that can be discussed and bring into that meeting.

Saying that, the issue is not about fearing about scarcity. The issue is about raising awareness more efficiently. Bringing, building the bridge between the two communities. How can we do that efficiently? Because if we will spend time here addressing different aspects of IPv6, which is why, and we know from the technical community we have a lot of issue for adopting IPv6 for deploying IPv6. And that is, and that's another huge issue pointed out by many speakers. We would rather focus from that and trying to solve something that doesn't really exist, we are wasting more time. And that's my concern.

I will come back again to the awareness issue. In our region, we have had several government representative regulators asking us during a meeting if in their country, there are IPv6 already in use, why in their country there are many more than ten, even ISPs already having IPv6 allocation or even announcing and having trial on IPv6. That means there is a disconnect again between regulators looking at telecommunications side and the IP side. If we take solid numbers, there are 5,000 more IPv6 address allocated in the Africa region than IPv4 allocations in the region, even though those IPv6 allocations represent 6% of ISPs having their own IP address. So there is stuff happening but people need to be aware. They need to be aware of the right information, and they need to be put in the right environment where they can address the issue.


DESI VALLI (India): Most of us are advocating the fact there is no issue of scarcity in IPv6. Coming from a developing country which is India, I think there is an issue of economics. Even on IPv4, we have raised a concern as Members, as well as from different association organizations, about the economics of the IPv4, taking the same existing IPv6 also and there must be a study to be done based on the ground reality of individual countries rather than having a uniformed pricing model and most of the time the reason which is given is APNIC has a cost of operations which is based on Australia's economy. So I think these cannot be a reason for setting the price of IPv6 or any IP, whether it is IPv4 or IPv6. So economics strategy must be done before setting up a price model for IPv6 which is based on the ground reality.

LORENZO COLITTI: Speaking as one of the few networks that have deployed IPv6 routing tables, I have an idea. We have measurements to show how much IPv6 is out there. We know that it's very small but at the same time, having looked at the internal economics of the transition, I can say that getting the address base is a trivial part of the transition. It's not getting the address space that is the cost of the transition. We obtained address space in 2005. We didn't do anything with it for a couple of years and then we started work. So it's a tiny amount of work. Also, I'd like to follow-up on Axel's question. Axel was asking what the problem is? I also don't understand what the problem is. I understand concern but one thing that might be useful is if we somehow were able to agree on what the problem is or more likely if we can agree on what would be a problem.

For example, if we think, well, if we have given out one 256 of the address space which is what's currently allocated to Global Unicast, or if we have given out half of that, then we need to start thinking about whether that might be a problem. So if we could agree on what would constitute a problem, I think that would be step forward.

MASATO YAMANISHI: Thanks, Lorenzo. So, even though time is a little bit over, if you have some comments for each point in the CIR model, I mean, yeah, as Paul had some comments for NIR or CIR. Or if you have comments for request allocation policy or competition between different schemes or ITU will become worldwide Registry, or about the decision process, please come to the microphone. OK, go ahead Paul.

PAUL WILSON: Thank you very much. The question has come up quite a number of times as to whether or not the proposed CIRs, to the extent that we understand them, are the same as NIRs and I wanted to just clarify a couple of very important points about NIRs to answer that question or at least to clarify it.

The NIR system in APNIC has evolved and been the subject of community policy-making and consultation over quite a number of years to ensure that it is a system that's consistent with the technical and operational needs of the network. And the model has arrived at a couple of particular features which do support that. So in particular, the NIRs in the APNIC region are absolutely part of the one regional community, and in particular in terms of policy, NIRs contribute to policy making as partners in policy-making in the APNIC policy process and they also abide by policies. That's absolutely fundamental. So there is no question in the APNIC NIR model of a policy divergence or fragmentation between the regional registry and the NIRs or amongst the NIRs. That's number one, the compliance with a consistent regional policy framework.

Number two is the fact that NIRs do not receive a pool of addresses. They draw when they process and approve address space allocation requests. They draw the address space from a common regional pool of addresses and they do that with full autonomy in most cases, so the systems are actually automated. Once an NIR has decided to make an address space allocation, that's their decision but the address space comes from the common pool. That specifically is designed to avoid fragmentation of the address space. So the NIRs under our model avoid the two types of fragmentation which I think have been observed about the CIR model, as I say, to the extent that we understand it, which is the danger of fragmentation of policies and fragmentation of address space. So I think it's important to understand that those differences, it's important to understand that NIR model is a fully supported and operational model in the Asia Pacific region.

We field fairly regular inquiries about the establishment of NIRs. The number of NIRs that are established is only at quite a slow rate. And that is according to decisions and choices of the communities who are expressing interest in the first place. The latest NIR that has been approved by the APNIC EC is in India. And that understandably, given the size and scale of the operation there was not a rapid process, but I'd like to stress the establishment of an NIR is a serious question for all concerned in the economy country, or economy, and at the APNIC level.

While I say that NIRs are actually designed and actively designed to satisfy the technical and operational needs of the network, I really think it's not clear what CIRs are being proposed to do exactly in terms of the problem that they're intended to solve, what the details of the CIR proposal or the actual implementation of CIRs could be is really unclear.

What I'm afraid of is that the CIR model, as it's described, would satisfy one thing that I've heard numerous times in discussions, particularly at the governmental level, which I will say this is not, on the surface, it's not an unreasonable desire or requirement. What I've heard multiple times is a country wants its own address space to do with what it chooses. That being, one, to characterize, very bluntly, one motivation that I've heard expressed. And I'm afraid the CIR model satisfies that directly and I'm absolutely sure that CIRs, if established, would be in most cases run responsibly. But the risk is of any responsible or divergent behaviour coming out of that model through that idea of having our own address space to do with as we see fit, which could mean any number of different things, of course. Thanks.

MASATO YAMANISHI: Thank you very much. Good explanation. Bill, you have a comment on NIR or CIR?

BILL WOODCOCK: Yeah, a very quick one, which is that I'd like to dispute the notion that the RIR system fails to represent some people, particularly in developing countries, that are represented by the ITU. So I just went through the ITU's website and checked the list of countries that are Members of the ITU and found 18 generally diplomatically recognized countries that are not Members of the ITU that are served by the RIR constituency. And, you know, this is hundreds of millions of people. I also checked and many of those have received IPv6 allocations already. All of them have IPv4 allocations. OK? So, the idea that the ITU can represent developing countries better than those countries and their constituents were able to represent themselves in the RIR system I think is one that can't stand unchallenged.

MASATO YAMANISHI: You have a comment for NIR or CIR? OK.

DESI VALLI: Based on experiences of going for an NIR in India, we took almost two years to decide upon the principle approval of making an NIR in India. I think these process delays, and is creating some sort of doubt upon the functions of the RIRs. I think if these processes are transparent, must more transparent, and the speed is has been implemented or addressed, based on the needs of the local community, the need of CIR may not be arising. Thank you very much.

MA YAN (BUPT SureNet): Look at the history. We know Internet really takes so important a function and that became the global and social infrastructure and we also look at the history and the government take a very important role in developing the social and economic and the order for all countries, not only including the developed, but also the developing economies. When we talk about the CIR and NIR model, I'd also like to see making a policy that will be sustainable development for the current and for the future.

Say the policy then, we will not only talk about the possible approaches in the policy but also more concrete, more feasible, including the technical process, how the technical feasibilities, and what's the impact, and the current ones and the future ones, so I appreciate if the research report could be publicly available, then more comments would be more helpful to target the problems and solving what we have been really heading to. That would be more helpful. Thank you.

MASATO YAMANISHI: Thank you very much. Any comments for NIR,or CIRs?

MARTIN LEVY: I will speak slowly. This isn't to do with the CIR issue and it's simply some statistics that are as useful at this point as they would be at any point. 26 countries or territories presently within the RIRs, collective RIRs IPv6 allocations that are not listed, they don't have allocations today. And of those, six are ITU Members, the rest are not. Now there are a few that aren't listed that don't actually make sense. That's just a country code disconnect between the UK versus GB. That's a simple one. And I apologize for that one. So there are places that I will email this in. Yes, sorry, this is a disconnect from Woody's, but there is a reason. Woody's list is very good. This contains such places like Ascension Island or Antarctica, Bouvet Island I'm just jumping down the list here. But we get to places like Guinea or Somalia, Chad, etc, that are Members.

So I want to make sure we understand what the scope of this requirement is, because not taking any territory or country and saying it's insignificant and shouldn't have addressing, I'm not saying that, but there are certain countries that are only 40 kilometres long and just off the coast of Antarctica, so the statistics have to be taken carefully. I'm just stating in reality out of this list, these are places that, at the moment, sometimes don't even have v4 allocations. As I said earlier, some of them don't even have penguins.


There aren't that much, yes, sorry, I was talking about the live version of penguins. So just make sure that when you look at this, you understand the true scope of this, because most countries are already connected or have allocations ready for connection. That's all.

MASATO YAMANISHI: It seems you're going back there, I'm afraid. Not actually related with NIR.

MARTIN LEVY: Yes, you're right.


ADIEL AKPLOGAN: I have one comment again on CIR versus NIR. And I just want to make a recommendation to Nav6. I think their proposal in that regard needs to be clarified because it's very confusing. The gentleman from India said something about if it is a CIR, the process will have been faster than if it is an NIR, but we have spent a couple of minutes explaining that in the policy level, the policy will be the global policy and if I understand well, the NIR set of policy is a policy defined by the region. So it has to be clear how the CIR will be set-up, will give them the legitimacy, how will that come from? Because if it goes through the policy development process which is bottom-up, it will go through the same model as the RIR are using today. So I think that the information in this study or the outcome of this study is confusing in several aspects. Even we saw it in the graph presented today where we have an additional RIR serving only the CIR in the graph. Those information are the way people are seeing the outcome of this report. If that's not the case, I think it needs to be clarified, because the whole work of the working group set up by the ITU will be based on that. The conclusion may be wrong from what was explained today.

RANDY BUSH (Internet Initiative Japan): I've spent over 20 years working in some of the countries we speak of in the third person in our techni-colonialist view from Europe and America. And in my experience, there has not been one case where they have been unable to get an IP address allocation that they needed. But there have been many cases where they, if they got an IP allocation, the colonialist monopoly, PTT, who is an ITU Member, would not route it. OK. This is the digital divide brought to you by the same people who brought you the analogue divide.


MASATO YAMANISHI: Sorry to interrupt you, but Sures has a comment for Adiel's comment. Could you say?

SURESWARAN RAMADASS: It's in reply to Adiel's comments. They were well taken and Nav6 has been open to comments on the paper that has been put on the ITU website. It's probably been about six months now and we've not received a single comment on it. So we welcome comments on the technical aspect. Please don't ask us comments on the political or governance aspect because that's not us. Anything on the technical aspect, anything regarding technicalities, you're welcome to give comments, you're welcome to give feedback, you're welcome to visit us, sit with us and discuss. Some of the researchers are in this room, and if you wish to meet them after this, you're welcome to sit down with them and have a discussion with them. But, unfortunately, till now, I've not seen anything in that manner coming to me.

Everything seems to be all on governance, which that's that side of the table, that's ITU. OK. So we are purely on the technical side. If you have a technical issue - for example, a good thing that Adiel raised, one of the clarification points, we can sit down and go through that and it will be nice. Thank you.

JOHN CURRAN: I have a question for the panellists and I believe it's a technical question but I have to admit I don't know regarding what's intended with the CIR proposal. In the paper, is routing considered technical or governance? It's technical? And the paper also discusses the ability of a country or a nationality to have control over the resources - is that technical or governance?

SURESWARAN RAMADASS: That's closer to governance. It's a possible model.

JOHN CURRAN: It's a policy point. So, when you're talking about the address space issue and you're talking about the viability of CIRs, and you're saying they're viable without routing impact, are you presuming the CIR model will allow addresses to be issued that are globally usable and routed globally?


JOHN CURRAN: OK. So back to the simple question, now that I know what's in scope and not in scope. Your paper doesn't outline how it is that those allocations get globally routed and that would be a useful clarification for the community to understand.

SURESWARAN RAMADASS: OK, John. Send me an email on that.


MASATO YAMANISHI: OK. Actually, we already exceed 30 minutes more. So I'd like to make your comment as last. Then we would like to discuss what is the next step? Go ahead.

JOAO DAMAS (IFC): First regarding the paper, now that you have asked for comment on it. There is one thing that I did miss when reading the paper. There is a lot of background, there is a lot of conclusions but I missed the model that you used to simulate the impact of different scenarios. I assume you had the model and that model had some degrees of freedom that you could vary and the values you input into that model and generate the predictions, and therefore the conclusions, of the paper had some documented assumptions. I missed all that background in the paper and it's very hard to be able to weigh the correctness of the conclusions, if in the absence of the description of the model and the parameters used to simulate the different scenarios.

Also there's a typical question, being an academic there's a simple question I was a scientist in a previous life, I had a few papers published in quantum mechanics. Was your paper peer reviewed?

SURESWARAN RAMADASS: OK, the answer to the first question, please email us on it and I'll get the guys that worked on it to give you a reply. And the answer to the second question, yes, it was peer reviewed by CISCO Internet engineers.

MASATO YAMANISHI: Thank you very much. Time is limited. Time is already over. So even though Mr Sures said this background problem is published almost six months but it seems communities have several views and several concerning points. So I think we need to input our views and our comment to the next ITU IPv6 meeting. So does anybody have some suggestion to next step? OK.

JAMES SPENCELEY (EC, APNIC): It occurred to me, as we were discussing this issue, that this is an incredibly long transcript to read over. I thought what might be beneficial is that I type up a brief summary, so I've sort of drafted what I think is a positive statement and I'd like to ask the chair after I read it to see if the community supports this view and we present it to the ITU so the introduction is that:

IP address management is fundamental to the ongoing Internet stability. Over the past decade the Internet has become fundamental to the world's economy. The Internet is truly global. What happens in one part of the world affects the rest of the world. Changes in address management could affect billions of devices globally, irrespective of the country where they are located.

The importance of an open environment - the Internet has become what it is today because of open transparent bottom-up processes. This has been used in protocols and in management policies. Everyone is encouraged to participate. RIR decision-making has no barriers to participation. Anyone, including Governments, can have their say. This has made transparent by our public archives, the decision-making process, mailing lists, video, and meeting transcripts. The operational stability, security, and efficiency of the Internet relies on a single consistent address management framework. The introduction of competing address management systems is not desired by us, the network operators. It carries the strong risk of fragmenting address management policies, of fragmenting the Internet itself, and of compromising the Internet security and its stability.

The equitable distribution of addresses is already in place in the current IPv6 management system and addresses are being deployed actively and effectively throughout the world. Each RIR already has the same size block to distribute within their regions.

So I think what we'd like to propose is some actions that parallel address management system involves significant risks and requires a clear problem statement. Complete explanation of its details and a thorough risk analysis of the consequences. Nav6 paper satisfies none of those requirements, therefore the Nav6 proposal, the paper cannot be considered as a substantial basis for discussion at the ITU IPv6 Working Group. Since concern about potential IPv6 exhaustion appears to be one of the fundamental concerns about the ITU's studies, we suggest that the ITU conducts a survey on this.

We also ask the IPv6 Group at the ITU to follow the example of the Internet community and the IGF process and make its documents and records available publicly, so that all Internet stakeholders can participate in deliberations, which can have global ramifications. We ask ITU Member States and sector members to recall the Tunis Agenda, called for a multi-stakeholder approach and call on the ITU to call upon the current multi-stakeholder system of address management. Chair, would you be interested in presenting that to the floor?


MASATO YAMANISHI: Thank you very much. Is there any additional point, Sharil?

SHARIL TARMIZI: Thank you. James, was it? Hi, Sharil. You sort of jumped the gun. I was going to try with my co-chair here too, try and sum up the sense of what we got around the room. The lady standing at the microphone, were you going to object violently if I say what I need to say first?


SHARIL TARMIZI: It's the elephant in the room thing and Axel knows this very well coming from me. I think there's to be some recognition, and I'm trying to sum up the sense of the discussion around the room. I think we recognize that everyone has a role to play here - governments, the RIRs, the ISPs, the ITU the question is what? In order to do that, I think there's also some consensus around having clarity on whether there is really a problem? I think there was some very good input from the gentleman from Hurricane and Bill on the other side. Is there really a problem? If there is a problem, what is the problem? The two themes that keep coming up seem to be one, perhaps something we can deal with immediately, is on the issue of capacity-building. Especially for developing countries and what is it can we do to help? How can we engage governments better at regional and national levels to improve understanding?

That's on the one side, and the sense I get around the room. And the other on the issue of scarcity of IP addresses do we really have a scarcity? Perhaps we should have a further study on that.

Saying what I also see a sense around the room just now, was perhaps instead of spending all this resource, all these brains around the room, all these brilliant people talking about whether it's RIR, CIR, LIR, NIR, I have no idea. We should focus what it is that we can do in a concrete manner to help developing countries who feel disenfranchised, how they can be helped to move from a v4 situation to a v6 and maybe one of the things the technical community can do to help is to look at having sort of a migration guidelines, a how-to guide, Idiot Books of some sort? I'm sharing the sense of what I get around the room, and I think you shared some of that in the statement. But I'd sort of try to see was that a fair assessment of sense.

RANDY BUSH: Why do you think we organized APRICOT? Why do you think these people are idiots? Why do you think the engineers of the developing countries are idiots?


RANDY BUSH: You used that word, 'Idiot Book'.

SHARIL TARMIZI: Sorry, I was just using a term called 'Idiot's Guide..' I'm sorry it offended you, Randy.

MASATO YAMANISHI: Use the microphone, Bill. If you want to say a comment, but again time is very limited, it's 40 minutes exceeded now.

BILL WOODCOCK: Is there such a developing country, is there a de that feels disenfranchised, to use your words?

SHARIL TARMIZI: I don't know. Based on the information that was available from the gentleman from Hurricane, I think it was, there doesn't seem to be.

RANDY BUSH: Can someone please identify? We have a room full of people here to help.

MASATO YAMANISHI: Discussion time is already over. So if you have some comments, for James's suggestion, please say now. Please do not step back the discussion.

JUDITH DUAVIT VAZQUEZ (PHCOLO): I thought ladies come first. Now, the root of the issue is that there are Haves and Have-nots in any developing nation. I'm a Gold Sponsor, I'm one of the Haves, and I'm happy to pay APNIC my membership fee. But there are ISPs in the Philippines that just cannot afford to be a Member of APNIC, and I think it is this 'Have-not' group that's the concern of ITU today.

So the question is how can an ITU Member Country help? A parallel structure is not the answer. It is not. But how a Member Country can help is truly to be that Member pay the fees and it's not really much for a government and subsidize the Have-nots, ISPs in its country. And I'm sharing this with you because it is my plan, with my allocation to actually provide address blocks to the poor ISPs in the Philippines, and I'm doing this as a Filipino. Thank you.


MASATO YAMANISHI: James, you have some comments?

JAMES SPENCELEY: I didn't want to particularly get involved in areas outside of what we've discussed today, and there are valid issues on providing subsidies to ISPs, training, and etc, but what I was trying to was actually sort of summarize the discussions on topic today via that statement. And then really just see if that represents a view that we can present in a concise form to the ITU. That's what I was hoping to achieve, Mr Chair.

MASATO YAMANISHI: Sorry, Xiaoya, you can say.

XIAOYA YANG: It's the first time I heard this statement and here you see the governmental manner of doing things. I'd like to ask this meeting to understand, I'd like to restrain from any station to this statement, because it's the first time I heard it and I might agree with most of it, I still might need time to consider it further. So if it would be submitted to ITU as the conclusion of this meeting it should be clarified ITU has no point here. Thank you.

MASATO YAMANISHI: But Xiaoya, we're discussing what should be the input for ITU IPv6 Group meeting. We're not discussing the conclusion of the ITU IPv6 meeting. OK.

JOHN CURRAN: I want to comment briefly on James's statement and the statement of the co-chair that followed. I think there are a lot of things that could be done between the RIRs and ITU in the area of capacity building, improving v6 deployment, there's a wide range. You noted some of the things that might be helpful, but I think a statement back, I just don't know whether that would be responsive since the ITU has given us a set of terms of reference that's very specific on one very specific set of questions. And I would hope whatever comes out of this group, even though there's a lot more that might be done between the RIRs and the ITU, since that wasn't asked but these terms of reference was it might be good to be fairly specific in the response of the findings of this group so it's applicable to the people who have to go to that meeting, like myself.

MASATO YAMANISHI: I'm afraid we're going back to Sharil's comment. I think we need to create some input to ITU meeting, then in that meeting we can list a different view if community has different views. I understand it's so what James did, I said. Let's step back to James's statement.

SUMON SHABIR AHMED: I'm Sumon, I'm from a developing country, Bangladesh. We are getting IP resolution for APNIC and so far we didn't finding any difficulties getting v4, v6 addresses. But I think ITU can really play a role here, that is that regulators in the developing countries, especially Bangladesh and Pakistan, they're coming from the incumbent telcos and they have a very good relationship with ITU definitely. And if any guy from ITU comes, who you all identified, but if we find APNIC people or even ICANN people in Bangladesh, we couldn't hold the regulators to bring them here. Two years back, two, three years back, the regulators have been talking about APNIC, they say what is it? But everyone knows ITU. So ITU definitely has a huge impact on the regulators.

MASATO YAMANISHI: Sorry, it's a comment for all panellists? Please say your comment for James's suggestion?

SUMON SHABIR AHMED: I couldn't follow that actually. But ITU, I think, next meeting, I think you can really help out to bring up the regulators to know our constituencies and create an awareness around regulators. ISPs know it very well and, OK. They're using it happily, but--

MASATO YAMANISHI: If you have more comment, please send to email address so we can catch up your comment.

SUMON SHABIR AHMED: Yes, thank you.

MASATO YAMANISHI: So we can catch up your comment to the statement. Okay, Randy.

RANDY BUSH: It's not an initiative, it's Japan's oldest ISP. I believe James stated the consensus of the room. I second it. And I wish the chair would call for consensus.


DAVE CROCKER: Part of the challenge in forming a response to the ITU is that the ITU posed the question. We need to be careful that the response is a useful response; a simple response to the question might not be the most productive. If the premise of the question is itself a problem, I believe that the sense of the room is that the premise of the question is a problem. I think James's draft following on Sharil's comments suggest we need a response which is a bit more broad than, than we have been asked. And that the draft and the additional comments form a very good basis for making this point.

MASATO YAMANISHI: OK. The deadline of contribution for ITU meeting was this Tuesday. We agreed this will be postponed until this Friday. Still, we need to try and capture all comments as much as possible.

DAVE CROCKER: I'm only suggesting an editing pass. I think some of the added comments enhanced if they're folded in, and I don't have immediate suggestions, but it felt to me that the added comments enhanced the draft a little bit. I'm only suggesting that there be some private effort to revise the draft which is a quite good draft but there be an editing pass to perhaps flesh it out just a little bit. I'm not suggesting that take place during this meeting.

JAMES SPENCELEY: One of the issues is that have limited time to present this. So I think the part of this draft that wasn't covered was potential other options for assisting ISPs with funding. Perhaps I could, you know, also additionally add some part to that?

MASATO YAMANISHI: Lorenzo, use microphone.

LORENZO COLITTI: Please project what you're trying to get consensus on.

JAMES SPENCELEY: I'd like to get consensus - on the screen? Oh, I don't know how to do that.

MASATO YAMANISHI: We just discussed at this session, I'm afraid the drafted statement is all ready but you are not. So if everybody agrees to take consensus in this session, I can. Everybody agree? OK.

PAUL GERMANO (Google): I'm in agreement with Randy's previous comments, and I did want it's sort of an academic question for the co-chair. I'm just curious, there's been a lot of mentioning of disenfranchised countries. Do we have a list of those? Because it's not clear to me.

SHARIL TARMIZI: Maybe let me clarify. Disenfranchised, because I started telling the stories about layers of governments, they're government ministries and agencies that don't talk to each other and they go forum shopping. The Idiot's Guide I was talking about is actually for governments, it's how to move to its v6. Very often the technical community in some countries are completely different from those who go to the ITU So they end up doing forum shopping. If you ask me for a list, I'm sorry I don't have one, but I can point to you a few that I know personally, who have a, you know, there's two camps, essentially, nationally, domestically, who have different views on how things should be done. I don't know whether that's helpful in any way to clarify your question?

BHADRIKA: I have a comment from Jabber, McTim, who says that the notion some ISPs can't afford the APNIC fees, in the first place developing countries get a 50% discount on APNIC fees, and secondly IP address distribution should be done on a cost-recovery basis so that ISPs or LIRs should charge their customers a fee that covers their costs. He says the cost issue is a red herring in my opinion, as many ISPs use IP addresses as an actual profit center.

MASATO YAMANISHI: In, yeah, even though your company or organization is not paying APNIC fees, you can attend a discussion of APNIC region, at least. Anyway, are you ready?

RANDY BUSH: Just one thing, can we not wordsmith on this? Can we not wordsmith this and start trying to rewrite it? It either is the consensus of the room or it isn't. Because that's a long rat hole.

MASATO YAMANISHI: Thank you very much. Yeah. How many pages do you have, James?


MASATO YAMANISHI: That's fine. OK, so as Randy suggests, our director of word smith - so I'd, may I ask strong objection for each part of it? Is it a good way?

LORENZO COLITTI: I'd like to be able to read the second page. There's likelihood we'll be able to agree on it in general principles, but not until we've read it all.

JAMES SPENCELEY: Would everyone tell me when they've finished the first page?

MASATO YAMANISHI: Let's go to the end of this document. Actions. Are you reading? Is somebody still reading the first part? Is somebody still reading Risks paragraph? OK, let's go to Actions. We have a fourth item in Actions, is that it?


MASATO YAMANISHI: It seems many people are still reading. May I ask consensus for Risk section and Action, I mean generally proposed text. Why are you raising hand? OK. OK. Does somebody, does somebody difficult to raise your hand if no, you're raising your hands, Randy. Anyway, let me ask, we can reach consensus. Is there any strong objection for this proposal proposed text? Seems not. OK, so OK. Still we need to do some editorial task for this text but basically we'll propose a contribution for next ITU IPv6 meeting based on this proposed text. Sorry for extending almost two hours. But thank you very much for your cooperation and very useful comment. This session is adjourned. Thank you very much.