Welcome to the United Nations | Department of Economic and Social Affairs

FINSIHED FILE

EIGHTH INTERNET GOVERNANCE FORUM

BALI

BUILDING BRIDGES‑ ENHANCING MULTI‑STAKEHOLDER COOPERATION FOR GROWTH AND SUSTAINABLE DEVELOPMENT

OCTOBER 24, 2013

9:00 A.M.

WORKSHOP 144

MARKETS AND LEGACY SPACE

* * * * * * * *

This text is being provided in a rough draft format.

Communication Access Realtime Translation (CART) is provided

in order to facilitate communication accessibility and may

not be a totally verbatim record of the proceedings.

* * * * * * * *

>> We are going to get started three or four minutes. I know we are running late. It's one of the problems of the first session of the day and I'm waiting for one more panelist. Thank you.

>> MODERATOR: I'm going to start. I will have one extra panelist that will join us in a few minutes. So first off, let me introduce the session make sure you are all in the right place and then, of course, our panelists for today. In no particular order I have Izumi Okutani from Japan. This is the tech oak cal community. And Andre as from lack NIC lack NIC covers the Caribbean nation. Joining us via web ex is Peter Tinish from address ex it is private sector. He is in the United States at the moment. And Rachel from AFRINIC. AFRINIC covers the Africa region. We have nick Nick Hilliard here he is from the internet exchange in Dublin. And myself ‑‑ I meant to be last, hold on. Paul Wilson from APNIC based in Australia. He covers Asia Pacific region.

We are waiting for Musab Abdulla from the Bahrain tings regulatory authority ‑‑ telecommunication regulatory authority. I'm Martin Levy from Hurricane Electric based in the United States. I would be technical community. So I'm going to give a couple of really short opening slides to just to put this in place. As I said, you are in workshop 144. We are going to be discussing V4 marks, legacy space. So I will give you a few opening comments and then we will hear from the various panelists and speakers here.

So first of all, absolute facts, APNIC and RIPE ran out of space on dates in 2011 and 2012. However, they are not out of IPv 4 space. There is still this reserve set. We will hear about this further as the speakers discuss this. Aaron, for example, went through stage one, stage two, and then a stage three process which adjusted and controlled how the members could allocate additional space making the requirements a little tighter but simply ‑‑ this is all highly documented.

The actual numbers for true exhaustion dates in the future are estimates. You can calculate them as best you can, but they are still based upon how the community, how the members of the RIRs do allocations, this is today's listing from Jeff Wilson's ‑‑ Jeff Houston's site. And it shows projected dates for the future for the RIRs.

What are we really talking about here? Potentially legacy space, potentially the value. This was quite amazing press from about a year ago from the U.K., one billion pounds worth of address space available. We will hear about that maybe. Finally, and I show this occasionally to realize this, if you go way back, you will see something well before there was any policy about selling, about moving, about transferring, about any of these here, this was an eBay bid that got removed, but it simply started the conversation about transfer policy that all hear about today.

So these are just a few things for everything to start the conversation. I will now leave the stage and moderate from the floor, and I will hand it over to Izumi Okutani for the first presentation.

>> ANNE RACHEL INNE: And also, well, there has been quite a lot of interest and discussions about IPv4 transfers and how we can distribute IPv space, but if we think about IPv 6 which is another version of IPv ‑‑ deploying this or having the networks deploy this IP version 6 would be ‑‑ this is something that needs much larger collaboration by stakeholder.

We first need to address policy Forums, there are five policy Forums operated by Regional Internet Registries ‑‑ the presentations are posted on websites so there is equal opportunity to visit. And these discussions are addressing the issue of various regions in RIRs.

Today we have four out of five registries adopted. The idea is if you have people with ready IP address and have face the transfer policy allows you to distribute transfer space to those who need it. So the regions that are marked the same colors you transfer the space within these regions. Somebody in Latin America will pass on space that they can distribute. They can distribute space to another organisation on the internet.

The issue we want to consider once the address space has exhausted is that depending on the region, address space available is different from ‑‑ so it's considered the North America area has the most space available and can be distributed. So to have more opportunity for organizations to receive this space, IPv 4 address space it is in RIR transfer policy and this allows transfers to happen in registries.

So at this stage it's partially those regions, but it's not. The other point I want to make is address space is finite in nature. So this is not meant to solve the long term solution. It's a policy to distribute space. So this is where IPv 6 comes in. And the current issue today is that we are still able to receive additional address space which is another version and we don't have to worry about exhaustion. So we can still continue receiving 6 from the internet registry but all IPv 6 networks are not transferable.

(No audio) so these are the things that are happening globally, but we need much wider stakeholder involvement the contexts and websites available in IPv 6 just in the same way IPv 4 users are able to use. That's another area we need to work on and that's starting to be, have a large proportion of internet connectivity and I think that's something that I would like to discuss about the effective ways of getting all of the stakeholders involved to help with IPv 6 deployment. Thanks. That's all for me.

>> MODERATOR: Thank you, Izumi Okutani. You brought up, of course, IPv 6 as one of the issues to discuss here, and the compatibility. Andre yays from LACNIC. I gave Izumi extra time and I will make sure we get through the full list here to an yays.

>> Thank you. We have a different situation compared to ‑‑ee (Cannot hear audio). He mentions that 42 million have been introduced as part of this ‑‑ (Cannot hear audio). Regarding other issues that for us are important to consider not avoid deploying this, to avoid taking these policies of transfer but to consider policies are effectively discussing policy quorums adopted by the community is that their location of IPv operations or IPv operations directly by the RIR to the members, to its members has an advantage, a much more important advantage. For example, security and they had unfortunate issues and the transfer policy we do that we have to be very careful in order to ‑‑ for example ‑‑ database could provide for regions so it's not only the availability of addresses that's the one single dimension for us we face in this issues, so this perspective from there may sound a little bit different and maybe more ‑‑ maybe a little naive could be a good sound also, I hope it doesn't sound this way because in our head we are confident that IPv 6 can make this some then the market could just, the policy could be just not using more.

So for us this is our current take on the problem. I hope we can develop things we can do. Thank you very much.

>> MODERATOR: Thank you. It's interesting to note that you talk about just one or two operators can make the difference in V6 traffic inside the region. Again, that's very pleasing to hear. Anne Rachel Inne from AFRINIC will be next and and we will continue with this discussion. Thank you. And actually, if I may make a comment, the transcribers would like stronger audio, I believe that's okay to ask of you, Ann. Thank you.

>> ANNE RACHEL INNE: Good morning, everybody, my name is Anne Rachel Inne and I'm from AFRINIC. So as you see from ‑‑ you saw from one of the slides that Martin showed at the beginning, AFRINIC is the registry that right now has the biggest of IP addresses left. So there is a little bit ‑‑ we are starting to feel a little bit of pressure from exhaustion of the same space in other regions. I will talk more about that.

And before that ‑‑ what we are doing to alleviate that fact in our region. We don't ‑‑ we are the youngest registry. We have about 1200 members. We have in fact very good, you know, pretty good news because in our region right now, the one pool of people who are really interested in IPv 6 are the Governments, which is great because if they do that, we are really hoping and seeing in other places where Governments have started being interested in asking us to help them, for example, design V6 plans, transition from V4 to V6 to their own networks. Others are starting to feel oh, okay, if they are doing it, maybe we should start thinking about it.

And AFRINIC has been doing for the past few years quite a lot of training both in V4 and in V6. As Izumi said, one of the things that we see is sort of pick up of and hopefully doing away with maps that were something that are still something that are plaguing the region really networks in the region, and that is also because being the youngest RIR, you know, because the region had been served by other entities for the longest time, people thought we could not have IP addresses, so they went away and basically used NAT allot. So we are trying to help them a lot.

We do a lot of hand holding during the trainings, after the trainings, and our registration service people are really having a lot of work, you know, in helping people do away with knots so that's something we are doing. Registration service is seeing a lot of requests now legacy space holders who are mostly universities in certain located in South Africa mostly who are now requesting that AFRINIC, for example, and some individual holders, we do have individual legacy space holders in the region who are asking AFRINIC to help them in transfers.

So we are talking to them right now, and we are trying to put a process together so we can help them do that because I think it would be in the interest of everybody that that space is, you know, brought back to the pool that is seen by everybody. So some other ‑‑ one other thing that has not gotten traction in our region is the transfer policy. There is one, but it's being discussed and it has been discussed for the past year.

But basically it kind of comes and goes. There are moments when we are thinking, well, yes, why not, let's do it, and oftentimes we end up not advancing on the policy because a lot of people do not want a transfer policy. There is another policy that two people from the region have put forward that is about B4 for academic networks and this is all in the spirit of making sure that we do away with the IPv 4 space region and use IPv 6 finally on the networks.

Another thing that we are seeing is now that some other regions have, you know, exhausted their pool or their biggest part of the V4 space we see people coming in our region, for example, establishing business so we can get V4 or 6. And just to make you aware, I am going to read you something that I received yesterday that says we are contacting you to participate in this for the lease or sale of your 196.1.0.0 IPv 4 space. As you know, IPv 4 is in the transitional period and we will soon ‑‑ and will soon be replaced by IPv 6. We are introduced in leasing or acquiring your IPv 4 for cash dollars before it becomes obsolete.

This is one of the things we receive a lot in the AFRINIC region now. So I am going to stop here. I gave you a little bit of the space where we sit mostly, so here we are really debating what the markets are in legacy space and how we are trying to alleviate that is really by pushing for IPv 6 as much as we can in the region. And ours being a young region, people are, you know, willing to take it up, but at the same time, the fact that IPv 6 hasn't been picking up internationally too much is not helping us either. So dual stacking is for the moment the way to go, and we are not too sure where this whole discussion in the community is going to lead us given that some are saying give up the space, others are saying keep it. So policies are being tabled, you know, our next meeting is going to be November and we know that the policy on V4 for academic net work, for example, is one that's going to be debated and the transfer run is coming back on the table. So one more that we will talk about. Thank you.

>> MODERATOR: Thank you very much. Quite a different view from the AFRINIC region because of those dates. Next participant and speaker is Peter Tinnish from address ex he is in the United States. He will be joining us remotely. Peter, the audio should be perfect so please go ahead and we will try and keep everybody nice and short and I think we are doing good for time. Thank you.

>> Thank you Martin. Good morning, my name is Peter Tinnish and it's an honor to participate in this workshop. I represent a marketplace for network operators to procure the right to IP space. We focus only on registration of IP space and legacy numbers issues 1983 to 1997. Many countries around the world were not used and fall low because they no longer need them. These companies no longer need to maintain their own network infrastructure.

Returning to business IPs ‑‑ use the space given them. This is typical about 1995. It was lost through the years and so many of these companies don't know they have these blocks. What kind of companies are these, hospitals, grocery store chains and the like. None are aware of the registry. These blocks are frozen in time. This is the supply that awaits.

Due to every growing demand for new services is a myriad of devices internet service companies must have additional space to combroa and due to dual stacks they must have IPv 4 and IPv 6 for the foreseeable future. So we invite network operators to become marketplace and to provide access credentials for sail and uphold the rights. They are not brokers per se. Once an operator comes to marketplace ‑‑ none of the RIRs call them and ask do you need additional space today and neither do we.

We invested millions of U.S. dollars identifying listing early registration spaces gone unused. The total size of unused space by early registrants is vast and significant. We currently have over hundreds of network operators and internet companies that the majority are right NTTIRs.

Surprisingly to many ‑‑ Europe, which is serviced by RIPE. In reverse order AFRINIC. The same with LACtld NIC we estimate they had exhaust 2015. We think more like 2017. While Arran retains over 29 million numbers in this inventory. It should last to 2015. That means two regional registries have decided to only ‑‑ due to new or existing members ‑‑ APNIC is unable as of April 14th, 2011 to contract additional space to members. That does not mean the network operators are out of additional space.

The run to exhaustion at APNIC was by huge requests by network operators. Many hold large stockpiles of unused number blocks in corporate inventory. Many Asia Pac officers have reflected information on how the market place works. They are still holding back and are not participating. The RIPE NTC region is the sole region is that has a vibrant and easy to navigate market today. RIPE has the same policies in place as APNIC. It's how they choose to apply policies that is unique. RIPE only applies its policies to contracted space. It does not believe that it has the right to enforce this policy on early registration holders.

Instead RIPE holds the rights holder of the registration for legacy space. We have managed a great number of transfers in early registration phase maintained in the right database as of today. So how large is the early registration holding for RIPE NTC. Currently there is over 197 million. We spent a lot of money to determine this, there is about 991 ‑‑ that are unused and fall low. That is 64.9 million dollars.

We currently maintain listing of 90 plus ‑‑ to transfer to network operators. All of these holdings are maintained in the right space and are not subject to the proposed RIR transfer policies. I said earlier our goal is short supply for the years to come. It's a Petri dish for how early registration should be treated by RIR since none of the other regions will develop a true market until next we're, we propose they take a deep breath and see what happens in the regions. Speculators and hoarders and other such words are bandied around suddenly stash vast holdings then there is a proof of the nay sayers, but if not, we have been doing this for over a year now and none of that behavior has been in site.

And we can remove the artificial barriers placed on other regions. That's why I believe it's important that if we participate in discussion like this, because this discussion is about markets and transfers of IPv 4 states. And while we can all talk about IPv 6 which is a wonderful idea, the concept is if there was not a market, what would be the answer? Because today version 4, originally known as the ZOD protocol standard is still the primary source for people to get on line with.

So I propose to this whole group workshop we need to stop talking about IPv 6 so much and figure out how to make them ‑‑ thank you.

>> MODERATOR: Thank you very much, Peter. And you actually just, you touched on one subject that should be just rementioned. You talked about accuracy of the registries when you talked about one of the focuses within RIPE. That's a focus inside all of the RIRs, but that accuracy of registry is something that is very important as we see IP addresses move around. I'm going to call on Nick Hilliard out of INEX in Dublin to speak next. And nick, the floor is yours.

>> NICK HILLIARD: Thank you very much, Martin, good morning, everybody, my name is nick hill yard and I'm CTO of INEX which is the internet exchange in Dublin. It might seem a little bit strange to have an IXP representative up here. I'm heavily involved in RIPE addressing policy. So I have been active in that area for several years.

I actually want to pick up where the last speaker finished off in terms of talking about internet address registry space and integrity because I think this is actually critical to dealing with the entire issue of legacy space. Now, several years ago in the RIPE region we introduced a policy to clean up the registry of certain types of address spaces. I have been assigned to end users and there was approximately about 16 or 17 years' worth of assignments which haven't been tracked terribly carefully, probably about 20,000 in all.

We are still dealing with that problem or at least when I say we, the RIPE agency is still dealing with this problem. We have two‑thirds or at least three quarters at this stage. And at this stage, they are dealing with a very small number of people who work very hard to contact simply because the majority of these people don't know they hold the space. Or they no longer exist or whatever.

And this provides a very interesting data point in terms of dealing with the legacy address space issue. There is an awful lot of legacy address space registered which is a couple of percent of the entire address included. There is actually no policy to help with this address space. RIPE NTC feels that it does not have any policy basis to make announcements about IP address space, however, they are a registry. And it takes time, resources and money to run a registry.

This is a very large block of users were using this registry and it makes sense for RIPE NTC to put in place policy to handle the address blocks. The primary intent of the policy is to ‑‑ and in particular if we don't implement a practice, if we don't actually know who the address space is registered to, then this makes that issue how to do legacy address space transfers incredibly difficult because of these and there is no due diligence done on behalf of the holders, and there is no guarantee that the transfer will be done correctly and there is no guarantee that ‑‑ have the right to use the address space. You can see that this is quite a serious problem.

In addition o to this, the RIPE agency has been offering ‑‑ to the legacy address holders. If the policy goes through, there will be various services. So the difficulty face the in the RIPE service area is to insure that the address transfers that have happened in the future are handled in such a way that the registry (Cannot hear audio).

In terms of the scale of this problem, as Peter mentioned 35% of address space assigned users. There are quite a few single entities who hold address space. For example, the Department of Defense ‑‑ about 25% is assigned to the region.

Very briefly I want to talk about current RIPE policies. They have a new policy proposal for regular RIPE addresses. This policy removes requirement for assignment to justify the use of space they are requesting. It's a pretty dramatic move away from all policies which require to state a need for the address space. And it's essentially a deregulation of the market in RIPE service region. The reason that this has come into play is because the RIPE community has realized that the address market is the reality.

Previously we did assign them on the basis of stated need, but because we don't anymore address space ‑‑ so we need to wake up to the reality that this market based transfer is going to. This also, of course, feeds into the legacy address situation.

So that's pretty much all I wanted to say.

>> MODERATOR: Thank you very much. Yes, this change of last policy is unique within the RIR world. Our final comments from the stage here will be from Paul Wilson from the APNIC region, and then I will after that open it up for any additional questions from the stage and obviously to everybody here in the room and anybody remotely. Paul, if you would, please.

>> Paul Wilson: Paul Wilson from APNIC the IP address registry for Asia Pacific. I don't want to get in between the panelists that have spoken and the question time. There are a lot of issues here that could be discussed but I think I would be very interested to hear what questions exist in the audience, because we could spend our time here talking about the transfers, our experience with transfers, our experience with exhaustion, what's the current status of activity levels in transfers which are going on today, and there are a few controversial issues as well or questions that you will have different answers for, market prices and what brokers are doing in the environment now. There is an issue being raised about leasing of addresses as opposed to a once and for all transfer. There are questions about transfers and transition to IPv 6 and viceversa the effect of IPv 6 on the transfer market. We have heard about legacy addresses that have different views where you are coming from.

We have heard about the issue of demonstrated need being part of a transfer policy, which is also an important one at the moment. I do think we all need to have a meaningful discussion and we also need to base it on facts and I think there are a few misunderstandings about what exactly the facts are behind many of these questions, but having said that, I have got many things in my mind so I don't want to spend time talking. I'm very interested to hear questions and discussions. Thanks.

>> MODERATOR: That's perfectly okay. That gives us some extra time and I appreciate that. Thank you very much. Okay. So there were some interesting little nuggets that were mentioned, and one of the key things that you would have noticed if you were listening was the difference between the different regions, the different mind sets. So I'm going to hit quickly some questions based upon what was said.

Anne Rachel Inne I will start with you out the AFRINIC and the Africa region, you are both the newest registry, but you are also from the various projections if I use Jeff Houston's, the registry that will have the longest amount of time before running out of space. But you said there is a discussion internally to say should we give some up or keep it. Keeping it, we can understand. Giving it up, where does it go? When you say give up some IP space, where would your region transfer or send back those IP addresses? What would happen if that was your members' request. Could you a little bit about this?

>> ANNE RACHEL INNE: So I suspect that given the fact that the community has not been very keen on having or putting together needs based transfer policy together, there have been some ramblings about given that we are seeing people from other regions coming to actually ask for big space, you know, from AFRINIC, maybe AFRINIC should put together some special pricing for those people to do away with the space.

So these are talks have been seriously, I mean, comments that have come from the floor, comment that's have been made on lists, and I'm not too sure if it's tongue in cheek or not or if, you know, some people really thinking that this is the way we should do away with the space, but the other thing that has gotten quite some traction is the one policy whereby people are, want to give the space basically to education networks.

That is the one thing that has somewhat, let's say, with the help of Afrin, African research and education network community who are all interested because they have as members all of the universities and higher education institutions in the region as members, they were very much in favor of having that policy pass and there was a very heated discussion at the latest meeting. Even though the Chairs of the PDP Working Group decided there was no consensus and they put the policy back on discussion, but this is one that is, I think, having the most traction.

So basically allow education institutions to ask for the space forgiven that we give them lesser pricing already, let's say preferential rates so that everybody, each one of them can get space and do away with, of course, the maps. You have to understand that in the region because universities were traditionally where internet started, you know, this is where we have the most networks also.

So, of course, if they were to come together and ask for, you know, big chunks of space, that could actually help let's say depletion date come really much earlier than, for example, Jeff predicted.

>> MODERATOR: Understand, yes, projections are only as good as we know today and things can change. The acronym you used PDP, poll policy development process a member based environment that all of the RIRs use from their members to decide how to move forward. I will hit the opposite of this. Izumi, in Japan as in the rest of Asia Pacific, you are dealing with the opposite, the exhaustion exists both in your case at the national level, but also within the region, within the APNIC region, you are in a whole different space.

So is the operator community ‑‑ how is the operator community reacting now knowing full well that there is no more IP space? Are they going to the market? Are they truly, are they implementing ‑‑ are they whole heartedly moving forward to IPv 6 and what is their sort of exact position today?

>> IZUMI OKUTANI: I think they are trying to address it in multiple ways. So they accept the reality that today IPv 4 connectivity is the majority, so it's not realistic for everybody to fully just only support IPv 6. So there are some transfer activities happening within Japan, so operators with these additional address space or expansions for customers, they actually try to get more space through transfer mechanism, but at the same time, they feel that it's just not enough to, as I mentioned, it's just not going to be a long term solution to the problem.

So most of the major operators make sure that they support IPv 6 in addition to V4, and so that's something that's happening. And so I think according to the ratio of number of network, I think, over 70% of the ISPs support IPv 6 in addition to IPv 3 IPv 4. We don't think IPv 6 will happen in the near future, but making sure we are prepared for the situation. And then in parallel, trying to get ahold of IPv 4.

And the big issue that we are facing is to get more content providers to support providing IPv 6 and the mobile providers start providing IPv 6 that's an area that we are struggling with in this effort.

>> MODERATOR: Thank you. I will ask one quick final question and then open it up to the floor. Nick, could you give us a quick review of as the RIPE region has just finished its meeting not a week ago on the interRIR transfer and the difference between where RIPE's members are thinking and the rest of the other RIRs around the world? I will ask you to be brief, but to the point.

>> NICK HILLIARD: This is an interesting point. There are policies to transfer address spaces between RIRs which have compatible address assignment policies, but because the RIPE NCC is branching to this new area where we are dropping requirement for stated needs requirement. So NCC is going to find itself on its own. This will be an interesting situation as to whether the rest of the world is going to change their policies to match RIPE NCC or whether the RIPE NCC is going to stay on this island.

>> MODERATOR: Thank you. Paul, you wanted to mention something?

>> Paul:yes, going over history, IP NIC implemented a transfer policy many years ago, some years before the exhaustion of our IPv 4 space with the, on the assumption, I think, within the community that we had the opportunity to make transfers available to try them out, and to get used to them even before we ran out of IPv 4, and to some extent that worked although the policy to be specific, when it was introduced, it was timed so that it would drop the needs based requirement at the time that app NIC ran out of IPv 4. So we went through a period, we ran out of IPv 4. We then had this transfer policy in action that was not needs base the. It was, it you like, free market transfer policy.

Now, that was fine. It works either way, to be honest, but what we then found was that in trying to establish an interregional transfer policy we were unable to do that within ARRAN because of the community decision that they would not release address space to transfer into a policy system which was not compatible and needs based. We had a lot of debate about that, but what it came down to in the community was a pragmatic decision that said ARRAN is firm on this policy. We are sitting here in expectation of being able to receive transfers from the ARRAN region where there are a lot of legacy addresses that may be liberated by transfer policy and reimposing the needs based component to our policy was not a show stopper, if you like. It's not that that creates any impediment that our community is not used to. It's the same needs based policy that we are used to. So that's a long way of saying we had a free market, if you like, transfer policy. We have reimposed the needs base policy in order to allow transfers.

What worries me about the RIPE communities' policy is the potential to reverse again the discussion back to a dispute with ARRAN, which is the source of most of the legacy addresses that could be transferred. Again, declining to enter into interregional transfers, the RIPE region may be having compatibility issues about multiple transfers around the world from one policy system to the next, but it does concern me that we are entering into more complexity and not more convergence in the overall global system.

>> MODERATOR: Actually, keep your Microphone on because I will just ask you a quick follow‑up. There are five Regional Internet Registries that operate and support their members around the world, and they aren't all equal, both in size, in policy and the like. Is it okay that they are different, they have different members? That's how I see it from the outside, but is it it okay that they have different mindsets?

>> Absolutely. And the RIR community take a close eye on each other to look at how the policy systems move along in parallel with policy changes often happening here, and then later here or vice versa. And there is a lot of exchange of experience and observation that this is a point of departure which would be okay. I mean, it has to be okay. It's a bottom up process.

But what concerns me is that the interregional transfer is the point at which level of expattability is required ‑‑ compatibility is required at least by one party who we are kind of relying on for pragmatic reasons. We can argue about the economic models an the effects of one versus the other, but we have got an internet to keep running in the meantime so I think pragmatic response is the one in the APNIC case and we would like to see a shared approach to be honest and frank.

>> MODERATOR: I understand. And that's great comments both for the audience and for discussion. I know that I don't really have to ask this question without knowing, but transfers do occur within the APNIC region today, within the region and they do transfer successfully. RIPE and other RIRs within each region have got that capability.

The floor is open for questions, and I'm seeing, yes, I will go from the back first, then Elvis, then Woody? Anybody else in the front. Do you have a microphone in the back? And, of course, name and affiliation, please.

>> Hi, good morning, my name is I'm a professor from NDIT at the universitych ‑‑ foundation called media for change which works with women and youth in India. The next billion will come on line from countries such as India, and emerging economies, developing countries, both IPv 4 and 6 are crucial to our integration. We also understand the internet as a finite critical resource.

Paul, my question is to you. This is a region which is as diverse as you can imagine in terms of a classic case study, multilingualism, multiple ethnicities, languages and Governments which are just trying to come to terms with the power that the internet holds and the possibilities that it possesses. I do want to understand as an organisation that is putting together such diverse viewpoints in terms of obstacles there are many, what is it that we can do as a registry and as groups that represent different points of view to do more advocacy when it comes to deployment because we have seen this empower and change stories.

As a matter of fact, we have the next stetion right here artingality 11:00 which is a Government of end gentleman session that is going to try to talk about problems and solutions that we are posing to the world but I do want to understand, what is it that we can do as a region, because we can make a difference at this point in history.

>> Paul Houston:  that's a big question even if we started talking about advocacy for IPv, that would be a very big question. But I will pull it back to the subject of the meeting, which is no matter how much advocacy we do for IPv 6 in this region, we will always have a regional and unconnected IPv 6 network if the rest of the world doesn't move with us. And I have to say one of the eventual effects of an IPv 4 transfer market is to redistribute the motivation towards IPv 6, which only comes when addresses, frankly are in short supply and to encourage the rest of the world which currently has more IPv 6 than needed in some cases to help to redistribute that motivation so that we can all move towards IPv 6 together. So I would argue that there is a strong link between an effective and equal transfer market and in fact the brokers will love to hear this, getting to the stage where P4 addresses are worth zero because we have all moved in a co‑hairnt parallel way to IPv 6.

>> MODERATOR: Thank you, Paul. Starting the first question with the phrase the next billion is a non‑trivial question to ask. Peter, you are remote, and I hope you have been following. Could you maybe comment on this, because this truly is, truly is in a way your space. There is a billion people coming on line. What do you have to say?

>> Peter Tinni shvment:  It's a great question but we don't look at numbers nib we look at cyber pace, when you had cyber 16, you used to think it's 16,536 number, but to add to sharing tools and techniques which make some people cringe, efficiencies are gained. The worst thing on average that most of the networks operators that use a site 16 are supporting a half a million to a mile onmonthly accounts. The issue you have to start saying to yourself is that you have been misled to think that there is, well, 4.3 billion numbers that can only support 4.3 billion people.

That's not how the internet works and when people start looking at numbers as a one to one ratio, they are absolutely not understanding how the internet is designed. I have many children in my house, and we have 31 internet connected devices, and we have a single number that we get from an address system and those numbers are very fundable.

So when you look at the next billion people, you can add the next billion people quite easily if you redesign how you do things which is much cheaper than swapping out everything. So we are proponents of a system that works. We believe that the next billion people can be brought on line without having to force the people that have large number blocks and have no interest at all in upgrading, you have to realize if I have a pool of numbers and all of my testers are happy, I have no business reason to change, none.

And that is per huge swaths of the world. So there is no business case to upgrade into a different system. And that's probably what has to be realized is that this is a business case. This is a technical and policy arena. Each network operator has to make a business case and that's probably the answer you don't want to hear. This is simply business. And that, unfortunately, is the reality of how we work with things. We look purely from a market. We do not make subjective discussions. We don't say this player abettor than that player. We let them determine based on economics and business settings.

>> MODERATOR: Peter, thank you. That was both transcribed and heard well. My next question from the audience was in the middle. Elvis? You have the mic if you can give your affiliation, et cetera.

>> Sure, thank you very much. Elvis Vella, I'm working for one of the brokers. I'm working for V4 escrow, we are a newly created broker and a broker that tries to recycle IPv 4 and promote IPv 6 as much as possible. This discussion has been really interesting and I do have a few comments. First of all, while we all agree that IPv 6 is the way forward, we do need to realize that only 2% of the world is going to be using IPv 6 and connected over IPv 6.

And there is still enough room to grow the IPv 4 internet while IPv 6 starts up because we have seen a lot of development over IPv 6 in the past year or two years. But still a huge amount of users of the internet are not connected and cannot use IPv 6. For the past three or four years I have been asking my internet service provider at home, when will you be able to offer IPv 6 for my only connection. And every single year his answer has been next year, and next year.

And while most of the content providers have managed to offer the content over IPv 6, I think the biggest step that needs to be made is that ISPs can also start connecting end users over IPv 6. I would like to pose a question to the panel and although I would like to ask the five RIRs what their opinion is on this, I think we should limit it to two types of RIRs, the ones that have finished their available free pool and the ones that still have it, and the question would be right now we to have interRIR transfers between ARRAN and APNIC. And these do work. They are not too many but they do exist and word. While the RIPE agency is coming with the new approach to the transfer market, would the RIRs that still have plenty of addresses and time talking about LIC and app NIC, would these see global policy as a way forward in a lobe market actually, and ‑‑ global market and the same question to the RIRs that no longer have a free pool would a global policy proposal be able to make this market work?

>> MODERATOR: Andre as, do you want to talk from the point of view of having space.

>> Yes, I was about to respond to your question, but part of the comment regarding the previous speaker, his comment on the nature of this, he said this was business and that is what it is, and that brought me to your region question about if there was something to consider about the differences on the ‑‑ or different perspective from RIR from the lack NIC perspective, we are internet registry, but more than that, our mission, our core values are we are an organisation that promotes development of secure and stable internet to our region and we deal with our workers, RIR in order to promote the goal of factoring the development of the network in the region.

So with this perspective, we, for example ‑‑ take into consideration not only the market needs but the development needs, and not only the next billion users that in our region would be like two years, we will have 80 million users, but also the new ‑‑ that will be connected and the needs of new type of interactions, new type of ways of dealing with emergency and new demands from the market, but also the developing needs. For example, in our region, the current consensus right now, political with Governments it's that every country, every single Government should have and does have, they do have, even the less developed Governments have, the central American Government, they have a programme plan and they have measures that they want to develop and LACNIC is working together with Governments in order to deploy programme plans. And that is one part of the strategy we are working with the ITU where we can ‑‑ OES, we are working with other governmental sub region and governmental organizations and we are working with our members in order to promote development and in order to grow to make the market grow not satisfy just business needs.

So this ‑‑  but also to in response of the question, the region does not demand at the moment a policy, there are discussions but there are not a focus on the policy of interRIR transfer, but there is a lot of demand for especially in the parts of the members that are active in our community, and there is demand from them in order to deploy IPv 6 to final users.

So, yes, the percentage is 2% not of the world, but percent of the ‑‑ which doesn't mean csh that is 2% of the world because many of them, more than 75% in our region, we have ‑‑ our region is more near to 1. We are behind in IPv 6 but 66% or more of the members have IPv 6 are located, we have a lot of efforts and training, the operators abilities to deploy IPv 6 so this is more there are many steps that have been already add done. So maybe this could be done in the future. We are confident we can achieve that and this is much more our focus. I wanted to emphasize that. Our focus is promoting development and promoting IPv 6 than resolving market needs that could be in the future. The community should decide which policies should apply, but this is not the focus I mentioned before. Thank you very much.

>> MODERATOR: The transcript missed one key word you said, you talked about philosophical difference between the business that was talked about by Peter and your RIR, and that's completely understandable, many different players come to the table with inside this environment. I had Bill. I'm running out of time here, guys, but make it quick. I have got one at the front and then John will be the final question from the floor.

>> Bill wit coch, PCH. This is a question directed to both Anne Rachel Inne and Peter, it's a pretty simple question. Isn't it a bit late in the course of history for Europe to be mining Africa for resources?

>> MODERATOR: A quick response, Anne.

>> ANNE RACHEL INNE: No comment.

>> MODERATOR: Peter, if you heard the question, I will give you a quick time to respond as well?

>> Peter:  did we don't look at that that's an insulting discussion point. We have European companies that have large holdings of IP numbers that sell to European operators. We have never looked at Africa once and I think that's kind of pejorative way of looking at things and kind of insulting.

>> MODERATOR: Next question at the front and name and affiliation.

>> Thank you. Martin from Morocco, I'm computer science professor in university, and I'm participating in IGF as isock Ambassador. My question is to Anne Rachel so from your perspective, the African countries that are most implicated in IPv 6 actually? And what could be the best approach for quick IPv 6 taking into account that can be a multi‑ stakeholder approach and as IPv 6 trainer, I have experienced that even if we make lots of IPv 6 courses, workshops, there is still some take time to get pressure to the operators and the operators actually don't follow as quick as possible. So what can, what more can be done? And my last question is why the universities are requesting more IPv 4 space address or from AFRINIC rather than trying to move forward? Thank you.

>> ANNE RACHEL INNE: Thanks very much, and it's great to see you here. So some of the things that we have been doing to as Andre us says is look at the development perspective so a lot of trainings. One of the things that we are doing right now that we are considering is, for example, starting tracks in universities on both IPv 4 and IPv 6 because we realize that it's definitely, you know, we have space where networks are growing. We are lacking the human resource to basically manage those networks.

We are going to need those, so going into universities and having charts that teach that are, is something we are doing. We also have on line labs right now that people can use. We do trainings, as I said, this year, for example, we are going into 17 African countries. We have been to right now 49 countries in the region so far and we have trained, I think, close to 3,000.

So training, training, training is really our model to get people to realize that, okay, IPv 4 is here, V4 might be lasting, but to grow the networks and for others who will come, we need to also do IPv 6. That's what we are doing. And one of the things we have seen lately is that Governments are the ones who are coming forward asking for hope to actually go dual stack, you know, and start using IPv 6 so we are hoping that also the fact that they are coming in there and showing the example, you know, in a region where basically Governments are pervasive and all over the place, but that can help us in terms of the uptake of IPv 6.

And for the question about the universities wanting more space, it's really a bunch of people from the university world who came with this proposal. When you look at it, the ones that are the universities that are in the southern region, so mostly, you know, South Africa, Uganda and others are some of the ones that are the biggest ones who have the most space, but others in the region don't really have that space even though they did espouse internet early, most of them are behind in that in general.

So instead of ‑‑ this discussion really came in parallel with the other one where others are saying why don't we just give, you know, sell the space? And others said instead of selling it why don't we just put together a policy that would say, you know, give the space to universities. Meantime though, you know, the issue is as we just said here, 98% of the networks are still under V4, and looking at the development perspectives, some of the discussions that we are having with the communities simply that, look, in any case, anybody coming now on the networks need, you know, V4 to start business or there is no business case.

This is what they all say. So this is where the community ‑‑ this is for the community to decide. So universities or some people have been proponents that university take the space instead of giving it away, and others are saying, no, we need to keep it and just, you know, go the regular route of giving it out and all of that. So it's really up and please join the discussion.

>> MODERATOR: Thank you. The final question, and then we literally we will do a tiny summary and then we have a coffee break and have to get out of the room. Name and affiliation.

>> John occurren, President and CEO of Arran. I was going to respond to something you said earlier, but if you are short on time, I will defer.

>> MODERATOR: Okay.

>> Very quickly, you asked the question is it okay that we have five registries, you know, should we have one registry? And I just want to ‑‑ right, I want to point out, we actually have one registry. It's one pool of numbers that has to be unique regardless of whether any given address block is in one RIR or the other, collectively, it has to only be in one. It's at an RIR or at the ‑‑ and there is coordination to make that happen. The other thing I want to point out is actually there are some strengths in the current system completely independent of operations. We are here at the IGF and talking about empowerment and talking about being able to have involvement in policy development.

The global policy mechanism of the RIRs mean that policies are discussed around the globe in local meetings with organizations that people know and trust. And that those policies actually have to converge in five regions to become global policy. So I want to point out the strength of our system is we actually have a system that has a lot of engagement and a lot of discussion. And, yes, that does mean it's hard to get consensus sometimes, but it means not that one party dominates the conversation over another. It means it gets discussed again, and party and individual groups move from side to side and it converges slowly based on what is really multi‑stakeholder empowerment. So I don't want to lose the fact that the policy development process of the RIR system is probably one of the few things we know is working exceptionally well on this model.

>> MODERATOR: Yes, and if anybody wants to join in any of the regions the whole process, the policy development process is normally email based plus meetings and as somebody who has attended many of them, lively events. Now, it is essentially coffee time. I know that other people have got places to go, but you were very diligent with your hand up, so you have 30 seconds for a question and maybe less for an answer.

>> I'm Victor from the operators association in India and we have 900 subscribers. A quick we is chat leverage V6 from a customer demand point of view while the large enterprises are okay, but the SMEs, the small and medium enterprises, their willingness or eagerness to ask for V6, any experience?

>> MODERATOR: What wants to take the 30 second answer?

>> There is a huge long tail with IPv 6 deployment. Some networks who switched on IPv 6 have seen immediately 50% IPv 6 usage on their networks overnight. This is misleading because it's from the top ten providers. Once you go past that, the deployment to IPv 6 is ‑‑ I don't know how it affects this other than huge amounts of ‑‑

>> MODERATOR: Okay. I want to thank the panel and I think this has been a good lively debate, lots of information. I apologize for it running late, but no one actually came into the room to kick us out, so that was good. And a round of applause and thank everybody, thank you.

(Applause).

>> Could I point out one thing which is off the agenda but pertinent to the last question, which is that many of you will be using IPv 6 right now it's on the network. It's what is running so user demand may not be there but it works.

>> MODERATOR: Thank you, Paul. Thank you, everybody.

* * * * * * * *

This text is being provided in a rough draft format.

Communication Access Realtime Translation (CART) is provided

in order to facilitate communication accessibility and may

not be a totally verbatim record of the proceedings.

* * * * * * * *