Transcript - Policy

Sunny Chendi: I was exposed to do this yesterday before the Opening Ceremony, but that went quite a long time, so I skipped my presentation to allow the time for other sessions to continue.

This is just an introduction to APNIC Conferences, most of you are actually familiar with these topics, but I still would like to run through them.

To begin with, I would like to acknowledge all our sponsors and supporters and speakers, moderators and specially you being here, taking part in this Conference, since yesterday or some of you actually been here since last week, joined our workshops for the first time in APNIC Conference, we had one week workshops, thank you for being here and supporting the APNIC Conferences.

For local host, they have been helping us quite a lot. You can see at the registration desk, at the bag collection counters and everywhere, they are all MekongNet staff helping us in every way and we request them to do it.

So that's a very great help. I would like to request you all to just acknowledge that support by -- APPLAUSE Thank you.

If you notice, we are quite busy over there on the left side of this Conference room. There is a lot of APNIC staff and you can see how many more laptops been there on the table there. We are doing webcast for all the sessions that have been held in this room and this is the first time we are actually using Adobe Connect.

It's working, fingers crossed, very well so far. We have been having good response from the remote participants as well. If you would like to see how we are running it, we are most happy to show you, if you come around our desk. On Thursday, for the Policy SIG, we'll have two remote hubs, one from Kathmandu, Nepal, and one from Medan, Indonesia, joining in the policy proposals.

We also providing our traditional low bandwidth options for those who can't get access to the Adobe, they can still access individual elements of the webcast.

Just health and safety. We have been hearing some of the delegates getting sick over the last two days, who have been here since last week. Yes, the room is a little bit cold. We requested the hotel staff to reduce the control the air conditioning here, but if you are concerned about your health, there is a medical doctor on site. You can call the extension 6707. There is a hotel recommended hospital as well close by.

We don't expect any major incidents to happen in this room or in this hotel, but if they do happen, follow the exit signs. You see the exit sign on all the doors and the closest safe point is Hunsen Park, which is just offset to the hotel, when you cross the road, just opposite. You'll be safe over there.

When you make questions or comments, respect the diversity of this Conference, the delegates, the language. When you approach your microphone, please state your name clearly before you speak and you can take your organization name, company name if you wish so, but please begin with stating your name, so our captioners, as you can see in the middle of the screen here, we have a live transcript, which is webcast as well, which is live on the Internet, people are watching it, so I received some questions in the chatroom, who is the person who just asked question, what is their affiliation? We are relaying that from the desk, but it would be good if you say your name and then ask your question or comment. Keep it brief and also speak slowly, so our stenographers can capture your words and put them on the screen, as well as in the Jabber rooms.

I think you probably know all of this. But we do have a prayer room if anyone would like to go and especially those who are following the Muslim calendar.

We do have prayer rooms on 8th floor, 828 and 826. You can get the keys from our member services lounge.

I don't have to go through the mobile phone. You have been very good. I was sitting there and observing.

I didn't hear any phone ring, except the first session, one phone call. But thank you so much for understanding how this Conference works and following the procedures.

Information for speakers, if you like to see your slides up on the screen here when you walk to the stage, please upload them 90 minutes before the session start time. So we can make sure that your slides are on the website, they are on the presentation laptop. We really do not encourage swapping the laptops, because if you look -- if you come and see the set up here around the podium, there is no space to keep your laptop here and there is no cable here to connect your latch top.

Everything is done from our IT desk.

I would much appreciate if you just upload your slides 90 minutes before the session, so we can make it on the screen here when you walk up to the stage. You can see what formats we support.

That's all for my introduction part, which I was supposed to do yesterday, but you are enjoying the Conference, I can see that. Our remote participants are also enjoying Conference, with the number of participants in the room. We can definitely see that it's been going very well.

I have a few more things to mention. We found this iPhone in this room. APNIC staff.

I switched it off, because it was ringing.

Also, we found a soft toy with a paper bag written Koyoto Craft Mart. It was in meeting room 2 and 3 where the.. There was a soft toy in there. I believe it's a gift for some special person. It's in the Secretariat, if you could please go and collect it.

That's about it from now, from me and I'll hand it over to the Chairs of the Policy SIG, Andy Linton and Skeeve, and I believe Masato will be joining from the floor later in the session.

So Andy, it's all yours and thanks for giving me the time. If you permit me, I would like to do one more thing, otherwise I can do it at the end -- can I, please?

Andy Linton: Okay.

Sunny Chendi: I would like to run the early prize draw, for those who have paid the registration fee by the due date, and we have a very good prize. It's a D-Link IPv6 WiFi router, fully native, no patches, nothing. So I would like to request my colleague Art to push the button to reveal the name.

That's Laurence Benjamin, if he's in the room here, he can come and collect the prize from me, otherwise ...

One more time. That's about it.

Wen Yun Li. Okay, what we'll do is we'll keep the prize until tomorrow afternoon. If you know these people, any one of them, please let them know to come and collect the prize from the member services lounge or from the IT desk here. I'll hold the prize until tomorrow and then if no one claims the prize, I'll do this again in the Policy SIG.

Skeeve will have it, for sure.

I'll run the prize draw again tomorrow in the Policy SIG. Thank you so much APPLAUSE

Andy Linton: Thanks, Sunny, for that. Okay, so I'm going to get the slides up for the next part of the agenda, which is the election for the Co-chair position.

We're going to do that first, the Policy SIG Co-chair election first.

We got three nominations for this. I know that two of them are here. I know that Masato Yamanishi is here, so he's here. I'm going to get them to come and talk.

Kashif Adeel, is Kashif here? Apparently not.

Brajesh Jain, who is over here.

The Policy SIG guidelines are quite clear on this, that if someone is nominated but isn't actually here when the election process takes place, then they can't be considered as a candidate, so you do need to be here to be a candidate. That means that we only have two candidates for the position of Co-chair.

This was the order in which they were received. I'm going to ask them to come up and speak in that order, so I'm going to ask Masato first and then I'll ask Brajesh to come and give you a few words of why they think you should support them in their candidacy. I'm going to ask them to keep it reasonably short, but given that we now have two instead of three, then you have a little bit of leeway.

First of all, Masato will come and explain to you why he should be your -- why you should support him.

Masato Yamanishi: Hi, my name is Masato Yamanishi and my position is SoftBank IPV, which is Japanese operator.

In my opinion, the role of Co-chair is to assist reaching consensus among all interested parties and in past 12 months, I have made all possible efforts and I believe I could succeed in it.

However, my current time which is one year is a little bit too short to accomplish, how can I say, meaningful result. So it is very appreciated if you would allow me to work as Co-chair for another two years. Thank you APPLAUSE

Andy Linton: Thank you, Masato. Brajesh will come and -- don't feel compelled. You can have a little bit more time than Masato, he was quite brief.

Brajesh Jain: Good afternoon to all of you. My name is Brajesh Jain. I am working on operating Internet service provider in India over last more than a decade.

I thank APNIC and you all for giving opportunity to stand here. We all appreciate that while presentation helps in engaging stakeholders in all 52 economies in APNIC, the responsibility of SIG Chair and the process is not only to build consensus, but also build a wider consensus. When we give a chance to new members, for such role, we actually give a chance to broad base our process and this broad base process and approach brings new ways of thinking.

Today, we know that APNIC makes policies for around ... time now has come to look at bottom of the pyramid and find ways to reach the waiting next billion of Internet users. I do assure you transparent and democratic process, if given a chance, I just will take a short time of yours.

I have experience as ISOC president back home in India and I have been practising equal opportunity and participate of all stakeholders in Internet back there.

I have over a decade of Internet experience at various levels and bodies in largest democracy of the world and many occasions, I have indeed got chance to build consensus among persons having very different, widely different views.

I would work towards building consensus inclusive of all and in the most transparent manner. I do recognise and surely you will agree with that. It calls for a lot of patience and hard work.

Thank you once again.

APPLAUSE

Andy Linton: Thank you to both our candidates for those speeches.

We'll now move to the process where we actually have the election.

I have three APNIC staff who are going to help the counting, so we'll count in separate blocks.

I will ask for a show of hands. I will ask for a show of hands for the first candidate and then I'll ask for a show of hands for the second candidate and we'll see how the tally comes out.

So the first candidate that I'm going to ask you about is Masato Yamanishi. So those people who are in favour of him being the Co-chair, would you please put your hands up and can you keep them up so that the scrutineers can get a chance to count those views.

Will you do that now, please. Put your hands up high and keep them up until I tell you to stop.

Okay. Thank you. You can have a rest now.

So those who support Brajesh in his candidacy, can you please put your hands up and again let the scrutineers have a chance to count your votes.

I'll just wait until I get the actual numbers from the scrutineers.

I think I have looked at the numbers. I think people can see from the show of hands that this is quite clear cut. I declare that Masato has been elected as Co-chair. So congratulations to him.

APPLAUSE And I would also like to thank Brajesh for standing.

I think it's really important that people are willing to do that. I think it's part of our process that people are willing to put themselves forward. It's a good thing for us to do that, so thank you for doing that.

Thank you.

Masato, if you would join us here, please.

Let me just confirm with Adam. That term is for two years now?

Adam Gosling: That's correct.

Masato Yamanishi: That's for two years. That's until this time in 2014.

Policy SIG agenda, we'll get on with that now.

Skeeve has just reminded me that one of the things that hasn't actually been an issue this time around, but the last time when the three of us were elected, 12 months ago, we were actually at the position where three of us were on the stage here as a brand new slate of Chair and Co-chairs with no continuity from the previous one. That can't be helped in some ways, but one of the things we felt we discussed this and we went back to the community and talked about this a little bit. We also talked with the executive APNIC EC.

If we're in the position where there is a change, then the person who is the outgoing Co-chair or Co-chair will actually stay on for the duration of the meeting, because they have been ... they would sit on the stage, deal with that and the new person would actually take their position at the end of the meeting, even though we have the election at the start.

That's to give us the chance to give everyone a chance to have some continuity, pass over, do a proper handover, rather than be in the position where you have just got a bunch of people up here, three people or two or three people up here who are going, "So what do I do now that I have been elected?" Just to make clear that's the position.

It hasn't been an issue this time, but in future situations, that may welcome up. At some point, it will flag up, so I just flag it with you so it's on the record.

Our agenda here, we have made a slight change to the way this is going to run, so let me just run through that for you.

What we're going to do in this session is we're going to do the administrative stuff that we have, we are going to do some proposals overviews and then we are going to do the two informational presentations from session 2 this afternoon. This is different from what's on published on the website in this slide.

Then on Thursday, we'll have the NRO NC elections, we'll have discussions on proposals 101, 103 and 104 all on Thursday, spread across the two sessions 2 and 3.

We will have a further informational presentation at some point on session 3 and we'll make sure that the other agendas reflects all that stuff, so there is a bit of change to what's actually on the board here. You can see this.

I also will try and get the Secretariat update into this session as well today, so we'll try and get through that stuff fairly quickly.

We have some action items. We actually have four action items from the last meeting. We have policy 33-01, which the Secretariat implemented, were to implement prop-102 and this has been done. It was implemented on 20 August. If you remember, this was the proposal which was to document the sparse allocation process and that has now been done and there was material posted on that to the Policy SIG mailing list.

33-02, prop-101 was returned to the mailing list and it's to be presented at this meeting. So we will see that one on Thursday.

32-02, prop-099 was returned to the mailing list.

As a result of discussions on the mailing list and the result of prop-102 and the documentation that surrounded that being published, the authority seemingly decided to withdraw prop-099 because he believes it's no longer necessary and is covered by that other policy proposal.

Then the other action item we had was that global policy GPP prop-097 was implemented by APNIC. It was passed up to the ASO AC and passed to ICANN who have ratified it, who have now passed it to IANA and Elise is going to tell us a little bit about that tomorrow morning.

We will have an update on what's going on there.

That proposal is the one about return of address blocks and we'll get some more update on that. That has been passed across all the five regions, long process, took effectively two or three years for that stuff to get through, because we have to get all five RIRs to agree.

Those were our four action items.

I know many of you know this policy development process, the SIG charter and stuff inside out. You probably can come up here and do it better than I can, but it's useful just I'll go through this just briefly for those people who are newcomers, there are some here and they haven't quite -- this would be new for them, so it would be good for them to know about it.

Our job is to develop policies and procedures which relate to management and use of Internet addresses.

That's laid out. We do that through some mailing lists.

We have a SIG Policy mailing list which is a public list where people can discuss ideas and issues and we also have a SIG Policy Chair list, where the Chair and Co-chairs and some of the Secretariat discuss the work that we have to do to get things up and running and items onto the agenda and so on. That's an administrative list for us.

The steps are well laid out. We have a proposal submission. That comes in, at the latest, four weeks before this meeting. Although we have asked people to submit those earlier and there's no reason why you'd have to treat that as a time when you have to do that.

It's a guideline. It's not the time when it has to be done.

We have discussion on the list before the meeting.

Then we bring those policy proposals to this meeting where we present them, we discuss them, we endeavour to reach consensus. If we reach consensus, then we report that to the APNIC Member Meeting and, if that's the case, then we take the policy that has reached consensus back to the mailing list for confirmation period.

That's about eight weeks. At that point, we pass it to the executive committee for them to endorse our recommendations.

It can happen and it has happened that the EC have said, "Think again" and passed it back to us. So there are a number of checks and balances in the whole process that allow us to say, the community thinks, the community thinks, the EC says think again and so on. So the whole process here, so that we don't end up with policy being passed in a rush or in haste and we get time to think about it all.

There's then a period where the Secretariat actually take our recommendation, turn it into the policy document, if you like or add it to the policy documents, put that out for comment and if that's agreed, then we go to implementation.

So the process is quite a long involved process.

Consensus, for those of you who are familiar with this, this is easy -- "easy" may not be the right word -- but when we try to reach decisions, we don't do this by voting. We don't say 54 people said this was okay and 52 people said it wasn't okay. Therefore, we have reached a majority and it passes. We don't work like that. We try to get everybody on board, try to get everybody to agree to something and we try to reach a decision that we're all comfortable with or most of us are comfortable with or comfortable with sufficiently that we'll live with it.

Our job here as the Chair and Co-chairs is to try and gauge that opinion and its based on what we see on the mailing list. It's what we see in the meeting. We will ask you perhaps to put your hands up and say give us an idea what's going on here. Let us try to gauge the opinion. But it's a combination of what's on the list, what's in the meeting and our job is to try and gauge that and make sure we get that right.

It's not the easiest job. Sometimes we get that wrong. Sometimes we misjudge that. But our job in good faith is to try to reach that decision or help you reach that decision and gauge your views on it.

We take comments via Jabber. We're very happy to do that.

There's a couple of example definitions here from the Tao of the IETF chk which may help people sort of come to see how reach the decision.

We're looking for a very large majority of those who care, they have to agree. If you don't really care, then that's okay. We're not desperately worried about people who don't care about the decision. If people care about it, we would like to see the vast majority.

We'd like to see that if there are strongly held objections, we debate those. And we actually get the issues out on the table and note that we're saying most people are satisfied that the objections are wrong. So it shouldn't be a veto and some of those are quite hard for us to make the call on. If somebody says, "I don't like this and it's because," you now, we have to try and tease that stuff out.

Things to think about here. Minor objections.

Minor objections where there might be some problems for some members of the group. A major objection would be where major objections would occur for parts of the community.

So our call to you and our job here is to encourage you for us to work together to find solutions.

We are a community. Sometimes that's hard for us because we're spread across a wide areas and we have different languages and we have different cultures, so our job is to try and work together to get good decisions made here.

There are some references and if you are having trouble sleeping over the next two nights, you can go and get these URLs out and you can read them and I promise you you will be wide awake.

Any questions on any of that? I think that's all. We usually go through this, but if anybody has any questions, I would like to hear.

No. Good.

So if we can get the other slide deck up.

As I said, we did make some changes to the agenda, so we are sort of making this up as we go along.

Adam has some things he would like to talk about from the perspective of the work he does on our behalf and on your behalf from the Secretariat.

Adam Gosling: I'm the senior policy specialist at APNIC.

My job is to support the Chairs and the community in the development of policy, among some other things that I do as well.

This presentation I had hoped to present at Delhi and unfortunately we kind of ran out of time. So I have kind of kept the presentation, but I have continued the work and so we kind of missed a step.

What I was going to present at the Delhi meeting was the progression that I had begun to rework the documentation for the policy, the resource policies in particular at APNIC. Regular visitors will remember that I had split the IPv4 policy into two sections, so that one contained the definitions and the environment and the goals of APNIC policy and then the other section was the real nitty-gritty, the criteria and the rules for registration and so on.

The reason I was doing that was because those definitions and the goals repeated in every policy document and it's kind of a weight of documentation that I think is probably better suited to last decade or last century or something.

The next step to that was to go and do the IPv6 document and split the goals out of that. So while I had an extra six months, so I have been working on a single policy document and I'll start my presentation now, because that's probably going to explain why I want to do it.

We are actually talking about the editorial policy, which is a policy document like all of the other ones.

It's in a different area of the website. It's under our corporate documents and it lives alongside the PDP, the policy development process, which is a policy in itself.

The editorial policy talks about how APNIC will manage these documents. It classifies what an APNIC official document is. They always have a number and, if a change is made to them, they are put out for an editorial comment period and there's a process that we go to.

In that editorial policy, the ... is plain text and I wasn't here, but I guess it was ... I guess that's the big advantage. Plain text will always be accessible.

But it's got some disadvantages and the documentation process that's built up around these text documents, text files, is probably a little bit over the top for an organization like this and I would like to run through some of that.

These text versions are all stored on the FTP server and I doubt that most people in the room have ever seen any of those official versions of the documents.

The document layout is like doing ASCII art, it's quite a time consuming process whenever those documents have to change.

Most people kind of don't look at the FTP server, they go to the website version and those website versions, the HTML version I guess I should say versus the text version they have to be kept in sync and sometimes, human error, they get out of sync. So it's actually another of the disadvantages.

In addition to this, there's an archiving and a versioning process that backs all of these text documents up and it's extremely important for the transparency objectives of the organization. It provides an audit trail so if someone says, "When did they they stop doing this?", I can go back and look through every version of a document and say, "That was changed in August 2009." Then I can go back to the proposal archive and I can see that was in response to a change in the policy. It's very useful, it's very important.

However, doing that kind of takes a little bit of time and I guess the advantage of doing that is there's no real change log. There is an audit trail, but it is a very manual process.

And there's a lot of them. I have the links here.

If anyone wants to go and look at the FTP server, you will probably see what I mean.

A policy change as a result of this archiving and versioning process requires production of three new documents, so one tiny little policy change means three new documents, changes to two metadata documents that describe where the documents are, the change to the HTML version on the website; in short, five published copies.

This is the example of the metadata. It's like the document header and it says when a version was created and made obsolete or if it's still active.

That's all the links and if you really want to stay up late at night, go in and have a look at some of those.

There are two sets. One is the archive and one is the current set of documents and they are all numbered.

They are all named differently. It's five copies of each policy document for each policy change and quite often a policy change might require changes to three different documents. So then it's 15 copies of documents.

The different versions are both text files, but one has .txt at the end of it and one doesn't. You can see there is enough kind of detail on there for you to look at it here.

Some documents, because of the layout requirements, are in PDF format as well. So it's kind of not even consistent.

Then there are five copies of these policies, plus some of the other ones, the NIR ones, but I'm not kind of so worried about those, because they don't change very often.

Confusing? The process had to be shown to me three times by Sam Dickinson, my predecessor, before I could do it. Even today, I come across occasional error, which is not really good for a transparent archival system.

I guess it's not really transparent, because no one knows it's there. So I'm kind of looking for a way to move forward on this, chk and in recent times, APNIC staffers don't put forward policy proposals, so I kind of stuck in a hard spot here.

Looking for a review of the editorial policy, just to kind of streamline some of this, and support to re-engineer this document archive and how it's handled, support to merge all the existing policies into a single document, the work of which is just about done.

So I guess one of the reasons behind putting up a single policy document instead of the half dozen or more that we have there is better internal consistency.

At the moment, there are situations where a definition might be slightly different in one document than it is in the other, so there's some work I have been doing to bring it all into one nice big policy document.

Maybe look at different, an alternate format, either PDF, XML or even HTML, I guess. Not so much the single policy document, it's an editorial thing. I can do that without a change in policy, but to change some of the archiving process, which I think would probably be better off having a single document with a change log than the process we have got, that would have to go through the policy development process.

No final slide. Okay.

So thanks very much. If anyone would like to talk to me about that, I'm around all week. Thanks.

APPLAUSE

Andy Linton: Just quickly, is there anyone who has any problems with Adam working on this? It just seems to me a perfectly rational thing for him to be doing.

No.

Just for the record, nobody objected to that and I think that gives you a mandate to at least go ahead, Adam, and do some work on that. Thank you.

Our next item here is we have an overview of the tree proposals we are going to deal with. We are going to do them as a sort of tag team here, Masato is going to talk about prop-101, I'm going to talk a little bit about prop-103, just the overview, and Skeeve is going to look after prop-104. So we'll just sort of proceed with that part of that now.

Masato Yamanishi: This slide shows the problem which prop-101 is trying to resolve. The current policy, portable IPv6 assignment of available only in network is currently or plans to be multihomed within three months.

There are technical or commercial reasons why some will not be multihomed.

So if provider-assigned IPv6 addresses are used, then any change of ISP would require renumbering.

Then since this is a fine-tuning based on the result of previous meeting, so it made several changes.

However, -- this one is wrong, I'm afraid. Incorrect version. I added underlined for the difference from previous version, but this one doesn't have. So I need to explain by oral.

Anyway, in section 4(B) "automatically" is added as editorial changes for clarification.

Section 4(C) has I think two conditions, which shows by (a) and (b) added for other criteria for organization that have not previously received an IPv6 portable assignment and section 4(d), last item, which started by any request for addition blah, blah, blah is also added as a condition for larger assignment, larger than /48 assignment.

There is no changes in the first part of section 4(e), no changing here. Down last item in section (e) is deleted pause this item talks about the sunset clause, but there were some objections for sunset clause. So this item is deleted.

That's overview of prop-101.

Andy Linton: I'll just take you through prop-103. This proposal came through in a version 1 through to the list and promoted some discussion. It has actually been changed and the revised version has been posted. So if you haven't seen the revised version, you should check back on the mailing list and make sure that you see the revised version. The text is different.

The main points here, the APNIC community spends time and resources proposing, discussing, arguing, et cetera, about IP policies out of habit. So the author felt that that was something that we perhaps did a little bit automatically. Perhaps even more than a little bit.

It's important to maintain an open policy process, but proposals are often not relevant to the prudent and high quality operation of the Internet.

So the proposed solution was that new IPv4 proposals should be carefully examined to ensure that they're really necessary and address real needs that cannot be accomplished with existing processes.

A discussion of the problem should precede proposals for new policy and in general, we should do the same for IPv6, although it's realised that as we learn more about IPv6, we may find more policy development may be useful.

Many of you will probably consider that this certainly isn't the normal run of policy proposal.

I think the spirit in which this has been proposed is for us to think about our policy development process and perhaps when we discuss this, that's the spirit I hope we do it in. We would be very foolish to say we think our process is perfect and therefore has no need for work or change or revision or simplification or any of those things that you might want to think about.

So when we get to discuss this on Thursday, I think that's the spirit in which we should discuss it. Many of you may find that it's difficult to look and say what is it we will actually do? Let's talk about that on Thursday and see where we go.

Skeeve will take us through prop-104.

Skeeve Stevens: Briefly, proposal 104 is to bring in alignment the transfer policy, primarily between RIRs, between regions.

The current policy in the APNIC region requires recipients to demonstrate a need for the next 12 months.

What that means is all they ask is that the next 12 months of their role not be considered. The ARIN region evaluates needs on a 24 month basis. So this leads to a difference between the conditions of the transfer between the LIRs in the APNIC region and in the ARIN region.

I believe this policy is primarily to deal with any region changes, but it's using the ARIN region as an example.

It says here that 12 months is also too short for transfers within the APNIC region considering many providers plan their services and addressing requirements beyond one year. The suggestion is two years.

The proposed solution. The way I understand this proposal is that at the moment, there is interregion transfers now available. So what we have here is someone that may be able to just fry based on their two year projection, let's say an example would be a /16.

But given if only their 12 months are considered, they might only in that period be able to justify an /18 because of a ramp up and things like that.

They might have someone in another region prepared to hand that and another here is ARIN, that under that 24 months, they might be able to justify that. But given the conditions that in our region and the transfer policy, those extra 12 months are not considered. So it's a bit out of alignment.

So the change would allow for the IPv4 transfer recipients to demonstrate 24 months of need to bring that into a line. The questions are in the case of inter-RIR transfer when there is an RIR which defines a period longer than 24 months in the future, the longer period would be adopted -- the longer period of the other RIR would be adopted locally.

It says that this proposal does not intend to change the requirements for an address allocation or assignment, just the period that it is considered.

That's about it.

Andy Linton: We're not going to discuss those now, but what I would encourage you to is that if you have questions, the authors, not of all the proposals, but the authors of some of those proposals are actually here. Most of you will know Randy, who is prop-103.

David Woodgate won't be here. He will be presenting remotely and Shin is where? Over here. Putting his hand up. If you have any things you would like to discuss with either of those authors between now and Thursday, that would be a useful thing perhaps for you to do, if you think there are things you want clarification on and so on.

Of course, if you want to talk to David, he's available on the mailing list and you can mail him and raise the questions that way, but sometimes that's a little bit more difficult than the face-to-face and sometimes it's easier as well.

Our last two items are going to be the presentations -- informational presentations I think we call them -- from Tom first and then from Youngsun La.

Tom Paseka. Tom is going to do something on critical infrastructure, and Tom works for Cloud Flare. Tom, you have about 20 minutes.

Tom Paseka: Tom Paseka I just want to go through an informational session before this is put to the list, just regarding using or including PKI, private key infrastructure, as critical infrastructure.

What is public key infrastructure? It is SSL certificates. It's the biggest part that this is without these, you don't have on-line security. You don't have Internet banking. It's a very important thing to have and there's two parts to this, which are very critical, is OSCP and CRL. They are both used to validate the certificates, whether they have been revoked or they are still active or expired. They are defined in RFC 2560 and 3280 respectively.

The current problem. Current policy doesn't define PKI as critical infrastructure and resources are managed through existing allocations for these services.

It's the same in the other RIRs. But I will be submitting this policy to the other RIRs in due course as well.

The proposal is to amend two documents, it's APNIC-124 policies for the IPv4 address space management in the Asia Pacific region to include public key infrastructure and APNIC-089 to include the same for IPv6.

Benefits and disadvantages, we would have a known address space for ... it's easy to track down and the providers of the resources would then have access to allocations. Disadvantages, more allocations coming through APNIC, more member resources and more workload.

The summary is there to include it.

Are there any questions?

Andy Linton: As always, I know Dean will remember this, but please state your name and your affiliation, if you have any.

Dean Pemberton: InternetNZ ^. I think my first question about this informational thing is I would like to see PKI a little bit better defined, because at the moment, anyone who stands up, or a quill SQL ? server could use this sort of proposal to claim critical infrastructure.

There's no reason why I couldn't do that at home or on my private domain. I can see the rationale of I'm doing this for like a CA infrastructure. One of the browser trusted CAs or everyone large corporate, so I can see some benefit in having a critical infrastructure tag for some PKI stuff, but I think the definition as it's written here is far too broad.

>>: Richard Barnes, BBN. That's what I was going to say.

PKI is a technology, RFC 5280 is the right reference by the way and not 3280, just something in there ^. That's a technology and I can use that for any number of things in private networks and not every usage of X.509 or PKXPKI ? is critical at all, so I think we need a lot more specificity in here, what flavour of PKI are you talking about? >>: Tom Paseka. I think expanding that to the CA, as Dean said, is probably a good step.

>>: Mike Jager. The question that springs to my mind is my understanding of the critical infrastructure assignments that exist at the moment, because changing those assignments later is difficult in that things need to be renumbered. The two mechanisms here, as I understand it, rely on or can rely on DNS in order to find the actual server. What's the reason that we can't do that? >>: Tom Paseka I don't have an answer for that, but you are right, they are relying on DNS for the IP space.

>>: Richard Barnes. I would also like that OSPC ^ at this particular point in the evolution of PKI history is not a tremendously motivating case,chk fy even that it's use in a lot of the web PKI in particular is sort of on the way out, if you look at some of the browsers you are doing nowadays.

Randy Bush (IIJ): I think the DNS point pretty well sinks this. But just in case you wanted more weight on board, you know, with the removal of the need for multihoming with IPv6, you don't need this either. Everybody goes and gets their chunk and there's an infinite amount of it or we'll think that for about 15 years until we discover infinity is a bit small and that's going to be life.

Andy Linton: Any other comments? Okay. Tom is going to be here for the rest of the week, till Friday, so you now know who he is and if you want to discuss this with him, please find him around and have a bit more discussion with him and obviously if Tom wants to take this further, he'll take it to the mailing list and people can discuss it there. So thank you for bringing that us to and thank you all for your attention on it.

We are going now to move to our last item for today.

Thank you, Tom.

APPLAUSE

Andy Linton: Our next item is on IPv4 reallocation among RIRs. Youngsun La. Again, about 20 minutes, some time for presentation, some time for discussion.

Youngson La: Good afternoon. I'm Youngsun La from KRNIC, KISA.

Today I would like to give information on presentation about IPv4 address imbalance. Now we have much address imbalance between RIRs and countries. This might lead to a serious problems IPv4 address shortage.

I think we should recognise this situation and should start discuss it.

Let me show you some data of current situation first.

IPv4 address allocation status. At the bottom area of this slide, there are IPv4 address allocation among RIRs. APNIC has 45 and ARIN has 36, AfriNIC has 5, LACNIC is 9, RIPE NCC 35. On the right side there is a table showing population of each continent. As you can see, the number of population of APNIC area is the largest and is about 3.8 billion.

Then you see ARIN is population is about 346 million. So the number of population of APNIC region is about 10 times bigger than ARIN's.

Available IPv4 /8 ? address in each RIR.

AfriNIC has 4.91, ARIN has 3.19 ^ and RIPE NCC has 1.79.

So the APNIC has only 0.92 address in its pool, which is the least amount of IP address among the RIRs.

APNIC has already initiated the last /8 policy.

This table show you the number of IPv4 address per capita. There are columns show you countries and number of IP address and number of Internet users and populations respectively and the last column there is a number of IP address per capita.

Countries here has more than one IP address per capita. United States is at the top and they have about 5 addresses per capita.

This table has same structure with previous table, that numbers in here are quite different. Countries here has less than 0.3 address per capita. You see some of them. Let's see China. China has 0.24 IP address per capita. This means more people can use one IP address.

Indonesia, 0.079 means 79 IP address available for 1,000 people.

In India, 0.029, 29 IP address available for 1,000 people.

Then you see Laos and Myanmar is even severe.

Myanmar 0.0005 means there is only 5 IP addresses for 10,000 people.

So so far, IP address were not distributed without considerations to the population of each countries. So some countries might face prop with their shortage.

There are two factors will aggravate these problems.

There is number of Internet users and mobile device is going up steadily with their economic growth. The second, it will take many years before IPv6 is fully deployed. IPv4 shortage problem will last for pretty long time.

In conclusion, we need to open up discussion on relocating IPv4 address.

Andy Linton: Thanks for sharing those numbers and things with us.

Is there anyone here who has any comments they'd like to make or any thoughts?

Dean Pemberton: InternetNZ. Could we just roll the slide deck back to the graph showing the different RIRs and how much address space they have left? Thanks very much for coming and giving this proposal. It does give us something to think about.

I'd just like to make a couple of comments. First of all, are you aware that in order to make this fly, you would need to run this as a global policy proposal? So all of the RIRs would need to adopt this as a good idea and then we would need to pass that global process.

It occurs to me that, looking at this graph here, only one of the RIRs at the moment is going to any that this is a good idea and four of them are going to any this is a pathologically bad idea. So it doesn't really seem like it's got the legs to pass on a global basis.

My other point I would make -- and one of the Secretariat can correct me if I get these numbers wrong.

During the last month of IPv4 exhaustion, within the APNIC region, APNIC was going through upwards of a /8 per month. So if this proposal were to go through and we were to do a land grab on every other /8 above the one line there, we are looking at three from there, two from there, that's eight ^. So we are looking at eight months and that's if the growth rate were to stay stable, but given that that would be the last of the last, I would say that that would go up and even if this got through, we would be looking at less than six months difference it would make.

So in the time between two chk of these meetings, we would do this and then we would be back in exactly the same situation that we are today and you would not fix the imbalance that you see within the population slides that follow this. There isn't that many addresses left on that graph to fix the problem that you've outlined. So, yeah, thanks for coming, but I don't think that this proposal will address those things.

Thank you.

Youngson La: Yes. My idea has some problems. In fact, before coming here, my draft was reviewed by some of the APNIC communities and the main points of the comments was that we don't have any mechanism to hand over addresses from one RIR to another RIR. So they say it's very difficult and it will take very long time.

Yes, it's true. I know I'm well aware of that. So my draft is not a perfect solution. I just want to bring up this topic.

Andy Linton: Thanks for those comments, Dean. I think normally, if this was a policy proposal, I wouldn't comment on it. But I think many of us will agree that it's a difficult one here to persuade people to give something away to be altruistic or do the right thing here.

Youngson La: Let me have any feedback whenever, during this meeting or after going back to your office, you can email me, whenever and whatever. The policy should be made with consensus, with community. So if nobody wants this idea, maybe I should think again.

Thank you for your attention.

APPLAUSE

Randy Bush (IIJ): You're fighting over scraps. The game is over. V4 is gone. Take the energy of this bullshit and apply it to deploying IPv6.

Youngson La: Yes, true, V4 has gone. But, as you see, the problem is exist.

Randy Bush (IIJ): Can somebody translate "fighting over scraps" into some more appropriate language? These are little. They will go away in a year. This solves nothing. You're fighting over scraps.

Youngson La: Yes. Thank you your feedback. Some people might think it is waste of resource and time, but some people is not. So my ears are open to anyone.

Masato Yamanishi: Let me say a comment about procedure.

It is talking about redistribution among all other RIRs, so if you want to propose this one, it should be global proposal, so it means it's not enough to reach consensus in APNIC community, it also need to reach consensus all RIRs, including ARIN.

Youngson La: Yes, I know that.

Tomohiro Fujisaki: Actually, global policy is only the policy between the IANA and the RIR, so I think this is not global policy, but global coordination policy. Just a comment.

Andy Linton: Thank you.

Martin Levy: Hurricane Electric ^. I want to follow on on that global point. We know from previous experience this would take a very long time if you were to take it to all the RIRs and to go through the process. It would take a very long time, much longer than possibly most of the /8s that are left on that graph. They will go.

They will just go.

But, you made a very good point. Randy spoke and gave a pretty accurate numerical this is not worth it, this is scraps, this is tiny. But you made the point that some people believe that this is important and some people believe it absolutely wrd and that I think is a perfectly acceptable right. So I think you have every right to pass this policy or try and pass this policy within APNIC such that the available space in APNIC is available to the rest of the RIRs. But I don't think you can do anything else than that. If you take it to the other RIRs, look at the graph. Psychologically, they are going to say no.

If you go for this, I think all you're saying is that the rest of the space in APNIC is available to other RIRs.

I don't think anybody is going to vote for that.

That's the honest truth. But you have every right to say that this is something that some people believe in. That's all I want to say.

Andy Linton: Are there any other comments on this for Youngsun?

Youngson La: Thank you again.

Andy Linton: Okay. Sorry to chop you off beforehand, people.

APPLAUSE

Andy Linton: So that's the stuff that we wanted to go through today and we have actually got 10 minutes in hand, so I'm going to make you sit here for the next 10 minutes and do something -- no, you can go now.

Okay. Our session on Thursday will begin with the NRO election, so we have about half an hour at the front of that where we actually have the process for the election and then we have our three proposals. So I hope to see you all at the two sessions we have on Thursday and again I iterate if you have any questions for the authors, now is a good time to raise those issues with them, so that we can actually, if you have things that we can sort out before the meeting, then do that before we get here on Thursday.

Thank you.

APPLAUSE

Andy Linton: We'll make a start on the second session of the Policy SIG.

The first thing we have to do this this session is to conduct the election for the NRO NC and Paul Wilson is going to come up and take us through that first part of the process, so, Paul.

Paul Wilson: Good morning, everyone.

It's my responsibility, on behalf of the EC, to take us through the election process for the Number Council election, which will be happening today. We have five nominees for the election and you will have a chance to hear from them after I have finished here, but it is necessary for the sake of formality to just go through the election processes with you.

I think probably almost even here has heard these procedures before and I'm really going to try not to labour them too much, just in the interests of time for everyone here, but if there are any questions, if anything is not at all clear, then please ask a question and you'll hear from these in this presentation what else you can do in case you have any questions.

I'll be taking you through the entire process here.

The background is that the NRO's Number Council is defined by the MOU that forms the NRO. It has three members from each region, who effectively act to serve the function of the address council together on the ICANN ASO.

Each region has its own selection process for its three members, as determined by that region. On behalf of the APNIC community, the APNIC EC has determined the selection process for our three members.

That process is that one member is appointed by the EC directly each year for a one-year term and two members are elected as we will do today. One member is elected each year, but they have a two-year term, so that's a staggered term with one member being elected each year.

The election isn't governed by the by laws, but the EC felt it important to conduct it formally in the general manner of APNIC EC elections, which are documented under the by-laws, just for the sake of consistency and proper process.

Here we are electing one member to fill a vacant seat on the NRO NC. That successful candidate will serve a two-year term until 31 December 2014. It means they don't formally become seated on the NRO NC until the end of this year, but they are I think welcome to join in as an observer in NRO processes until then.

We had a call for nominations which ended at the end of July and we have both on-line and on-site voting and there are some links here that give you more to look at there if you're interested.

We have an independent election Chair for each election in this case, it is the duty generously agreed to by Prof Kanchana Kanchanasut from the AIT in Thailand, who has been asked by the APNIC EC to fulfil the responsibilities of overseeing this election process, appointing scrutineers, resolving disputes and declaring the result.

The election officers are responsible for managing the overall process, appointed by the EC from APNIC Secretariat staff, and George and Connie are doing that for us today.

There are also election tellers from the APNIC staff whose responsibility is to watch the ballot box and to count the votes, reporting the results afterwards. Anna and Frank from APNIC staff are doing that.

We have also the provision for two election scrutineers to be appointed from other RIR staff who are present. George from AfriNIC and Leslie from ARIN have agreed to do that. They have been appointed by the election Chair. They're independent completely of APNIC or of APNIC members or candidates. They observe the process and notify the Chair if there's any problem.

The on-line voting started on 13 August and ended on 28 August in accordance with the published procedures.

Each member of APNIC is entitled to one vote via MyAPNIC. That's each member organization.

The system is built to preserve the anonymity of those votes while keeping a record of the votes that have been cast.

The on-site voting is what we're doing today. We'll start shortly, as announced by Prof Kanchana. It ends at 2 pm sharp today. The voting is open to individuals in this room and every registered APNIC 34 attendee is entitled to one vote as an individual.

The logistics today: We have a desk outside. The ballot box will be there when the voting is opened. It will be supervised at all times and enquiries, again, if you have any, should be directed to those election officers, George or Connie. They will be seated outside as well.

There's a ballot paper which is used for casting your votes and they have pretty clear instructions on them, I think you can see that on the ballot papers.

Just vote carefully, to make sure that your vote counts.

This is an example of the paper that you should have.

The counting procedure is undertaken by the tellers under the scrutineers as, I said. All the ballot papers re-examined, there are tally forms that are used during that process, the voting reports from the on-line system are printed out and incorporated into the counting process.

The declaration of the result will be the responsibility of Prof Kanchana as election Chair. That will happen at 3 pm today, according to your program, according to the agenda for the Policy SIG today. So, as I mentioned, the voting will close at 2 pm sharp and the scheduled announcement of the election results is at 3 pm.

Prof Kanchana will announce, as Chair, the name and the total votes for all of the candidates and will report if there's any disputes along with the resolutions, and any communications from the scrutineers about the counting process.

If there's a dispute, then it is the responsibility of the Chair to oversee that dispute, to receive any complaint that might be made. It's important that any notice of complaint has to be lodged no later than one hour before the scheduled declaration of the election.

So I'll just make it clear that the scheduled declaration is 3 pm, so any dispute that anyone would like to raise would need to be lodged with Prof Kanchana by 2 pm, which happens to also be the close of voting.

Those notices may be lodged by members through their authorised voting representatives or by candidates.

The election Chair will resolve disputes entirely at their discretion and, as I said, will provide notice of of any disputes lodged with the declaration.

That's it for the explanation of the voting procedure. I hope that's clear enough, but please, if there's anything that's not clear, I think you have the opportunity to ask a question now. You certainly have the opportunity to approach the election officers outside at any time during the process today.

I think next will be for me to hand over to Prof Kanchana, who is going to call, I think, for the election candidates around make it possible for candidates who are here to make a speech and so on.

Kanchana Kanchanasut: Thank you very much, Paul. I am very honoured that the APNIC EC trust me and invite me to do this task. For those of you who do not know me, I'm a professor at the Asian Institute of Technology in Bangkok. I am also running Internet research in a laboratory that provides training and research at AIT as well.

For this task, I'm going to invite the nominees one by one to share their vision with you before the election and each one of them would have three minutes to speak. Then I will open the ballot box to show you that the box is empty, there's nothing inside. Then we will take the ballot box outside to the desk and you all can -- the voting can then start.

I will invite the nominees to give a three-minute speech each, and the order of the invitation will be the order of the nomination received.

So the first nominee is Kashif Adeel. Is Kashif Adeel in this room? He's not here, so we will move to the second one.

Naing Winn Oo, please come up.

Naing Winn Oo: Good morning, ladies and gentlemen. I'm Naing Winn Oo, working for Myanmar ISP as a network engineer. So basically I do monitoring, settlement and configuration of the main server, DNS servers and ...

system also, as well as spam filtering.

As you all may know, Myanmar is very new with the election and voting thing and we might have very less exposure to the international affairs and issues. I may be only go with very ... I may not have the very impressive experience about NROs and things, compared to the other nominees, I will try my best if I could be one of the selected ones.

Thank you.

APPLAUSE Kanchana Kanchanasut: Thank you very much, Naing. The next nominee is Naresh Ajwani.

Naresh Ajwani: Good morning, everybody. At the outset, I thank this great country Cambodia to show the way for many emerging economies to not only engage in APNIC process, but in other policy making processes for Internet.

Lovely people out here, how they manage and balance their way of life is an example for everybody. In this building itself, ground floor full of entertainment, 1st floor full of restaurants and the 2nd floor full of deliberations.

This balancing act of Cambodia is an example for many countries, and actually I'm advocate of balancing things in life. I went for education, didn't go so far that I become a doctor or a professor. I went for writing papers, posted on a website of Telecom Regulatory Authority and many other places, but didn't write any book.

I simply believed in a philosophy of middle path.

I am engaged from last four years with addresses, but I also believe in leveller of Internet. I personally feel experienced people have a lot of contributions to make, but it's youngsters who will show us the real path and the forward path.

Last four years, the support of community has grown in terms of number of votes, first time when I got elected, I got around 50 votes. Second time when I got elected, I got around 156 votes.

I don't want votes and I have decided this is going to be my last election, because I somehow believe that what I preach, I practise also. If we will not give the way for youngsters, then who would give the way for youngsters to come and make policies? I need love from you. I need you people to forgive me for any of the things which I did to get my NIR.

Please take it as my -- as passion for like you people have for your country, moment we got, the first thing which we all did, we brought in normalcy to our relationship. Still if there is anything left in your heart, please remove it. Give a space in your heart.

That's what I want. Thank you very much.

APPLAUSE Kanchana Kanchanasut: Thank you, Naresh.

The next nominee is Aftab Siddiqui.

Aftab Siddiqui: Good morning, everybody. As you know, this is Aftab Siddiqui. You must have seen me couple of times on this stage for delivering so many other things.

Usually, I'm the operational guy. I have been involved in operational stuff for last seven, eight years.

I have been coming to many APNIC Conferences and APRICOTs, and those who follow the Policy SIG mailing list, I'm active as well, trying to share what my operational experience can guide me to the best of things.

Little background about myself. I have been involved with IPv6 Taskforce as a core committee member for last three years, and for last two years I have been in this SANOG program committee and recently got elected as a Co-chair for program committee in SANOG.

I have been trying to put whatever best I can, with the community, and sharing things, operationally, as well as in the policy matters as well.

As Mr Naresh said, it's not about getting votes or getting like the seat, it's about the support, it's about the recognition. Like I see many faces, new faces all the time, whenever I come, I get great deal of appreciation and great deal of support. This is what I'm looking for. It's about time that we should collaborate much.

We hosted SANOG in Karachi two months back and I really thank everybody who participated from APNIC and other organizations, even though the situation was not that great in Pakistan, but I still thank everybody.

So this is me. This is my profile. I hope I will be able to deliver the way it should be, so thank you so much from my side.

APPLAUSE Kanchana Kanchanasut: Thank you very much, Aftab.

Next nominee is Ren-Hung Hwang.

Ren-Hung Hwang: Good morning, everybody. My name is Ren-Hung Hwang. I'm a distinguished professor at Chung Chen University in Taiwan. I have many academic experience. I'm the network textbook for ...

I published more than 200 journal and conference papers and I have editors for many international journals and Internet ventures for international conferences.

Currently, I'm also deputy director of the computer centre of my university and I had also had been department head for six years.

I actually participate in APNIC NIR activities.

I have been a member of the IP address and protocol committee of TWNIC since 2005. I have conducted many research projects on IPv6 deployment and IPv6 exhaustion issues for TWNIC. I actually participate in Taiwan's IPv6 initiative program.

I have been actually participated in APNIC meetings and Policy SIG since 2009 and I also thanks to Dean Pemberton for giving me the chance to be the co-author of prop-102, which has just been implemented, August 22.

Why am I running for the position? With the experience in academic and APNIC community, I'm eager to devote my capability to this community. The duty of NRO Number Council is to oversee recommendations on global IP address policy and implementation of these policies.

If I am elected, I will actively commit my time and resources to fulfil duties and with this position, I will listen to the APNIC community and convey the opinions to make the recommendation based on the best interests of the address community.

I will proactively cooperate with the other two NCs, namely Tomohiro and Andy, to make good contribution to the community.

I think APNIC and all of you for giving me the opportunity to present myself here and I hope I have the chance to serve the community.

Thank you.

APPLAUSE Kanchana Kanchanasut: So we have seen -- we have listened to all of them. I thank all the nominees for their speeches. So we will move on to the next step, which is the ballot box.

We are going to show you that it is empty. There's nothing inside. It's a little bit dark, but all of you can see that nothing inside. This is not like Thai election! So I'm going to seal it now.

Sealed and locked. Now we are going to take this outside and voting will start. Paul and I will do that.

Andy Linton: Thank you, Paul and Kanchana for that. It's great. Please all don't all rush outside to vote right now, just wait till a little bit later.

So our agenda is this morning, we have one proposal to listen to and we're going to have David Woodgate connect via Skype, I think that's what we're doing, Sunny, and David is going to talk about prop-101.

This proposal has actually been in front of the Policy SIG meeting and on the mailing list over a number of iterations and there has been recently on the list some discussion where a number of people have shown support for that, and my interpretation is that no one at this stage on the list has, as it stands, any strong objections to this proposal.

If we can have David, if he's there.

David Woodgate: I'm here.

Can you hear me? Can the Conference floor hear me? Andy Linton: Yes, we can hear you.

David Woodgate: Okay. So thank you. I would like to talk about prop-101 again. I'll try to skip through some of the material that we talked about at APNIC 33 fairly quickly, but for those who haven't read the prop-101, I will cover the material quickly.

Sunny, could we go to the next slide, please.

Thank you. I'm not seeing all of the bottom of that slide for some reason. I hope I have sent you the right version.

But basically, IPv6 policy in APNIC as it stands is that only multihomed organizations will be assigned -- granted portable assignments. The issue is that not all organizations are not multihomes, at least at different carriers, they may have multiple connections to a single carrier. Of course, the difference with IPv6 is that whereas under IPv4, you could assume that for certain types of networks, you'd probably have priority addressing within the network, a demilitarised zone of public addresses facing the Internet, and the assumption being that it would be fairly easy to change the DMZ to reflect a change in service provider with IPv6, of course, addressing throughout the entire network, that assumption may not apply any more.

You basically would have to renumber the entire network of the customer, if they would wish to change provider.

That's obviously a major drawback for an organization which is in a position where it only really wants to connect to one service provider, but it doesn't want to be locked into that service, because if it has to accept its IP addresses from the service provider, then it would be a massive effort to change the addressing if at any time they wish to change to another ISP.

Of course, not all organisations are able to use dynamic IPv6 and some organizations have -- their networks are just too large to change. Unless they want to pay for additional services and multihomed just to get a portable assignment, then this is a disincentive.

We heard yesterday that there are certain companies and organizations already who would like to see a NAT 66 model. I suggest that if there are impediments to them getting their own address space, then you will see further drivers for NAT 66 along the way.

Next slide, please.

So APNIC is the only RIR that insists on multihoming for portable assignments. All other RIRs have removed it as a mandatory requirement. For reference for our later discussion, there is also no other RIR that has a mandatory technical criteria for portable IPv6 allocations. There are requirements for descriptions of planned usage, but in terms of justifications, most of the other RIRs are just specifying reasonable technical justification. If I interpret the ARIN rules correctly, even ARIN states a series of criteria, but which includes reasonable technical justification as a last option as well.

There's no RIR that doesn't have a reasonable justification clause, including the appendix, but I won't go through each RIR in detail, but it comes up in discussion later.

Next slide, please.

So the proposal as first presented at APNIC 33, basically that they must be APNIC members, automatically eligible if the organization has previously received IPv4 portable assignment, a reasonable technical justification to demonstrate why ISP/LIR assignments are not suitable. A minimum /48 to be assigned to the routing table so as to be below the APNIC ratio threshhold or having impact only one prefix to be assigned to an organization and the Secretariat to use sparse allocation methodology to try to increase to just a change in the mask link, and the additional prefixes and also requirement to demonstrate good aggregate practices by the organization.

At APNIC 33, not 34 SIG, a major point of discussion ws massive increase of portable assignments leading to a global routing table -- sorry, are we on the next slide, the discussion at APNIC 34? Clause 4.E.(d) was added, asking the Secretariat to regularly report on IPv6 portable address assignments.

I think we actually have seen an example of that in past couple of days anyway, incidentally.

There was also a suggestion of whether a sunset clause be incorporated to capture ... that there would be an explosion of requests, guaranteed shutdown, but that was rejected by the community.

I ... that SIG as well. The sense was that there was a broad support for the initiative, but it was just this particular point which was about handling of the portable assignments which is talking about ...

resulting which caused the discussion.

... since APNIC 33, there has been support expressed for the current draft, there is no opposition expressed to the principle of the proposal raised by Terry Zhang, specifically about the reasonable technical justification rather than specific mandatory criteria.

Those who are following this would have seen that there was a suggestion ... there was an offer made to adjust the draft to make the example ... last critical civil or medical services be the mandatory, rather than examples of technical justification and perhaps including multihoming. Terence said this approach would address his concerns, but since then, the only comments on the list have expressed corrections to the draft as it is ... there is a preference for reasonable technical justification.

Removal of complexity of IPv6 addressing for some classifications and consistencies with other RIRs and the risk as discussed is that if there is an explosion of requests for portable assignments, then that would potentially have an impact on the IPv6 routing tables.

As also ... the suggestion is that ... factors ... the economics of paying member of APNIC to receive a portable assignment anyway, which would be ... plus the complexity to go through that process, plus the assumption that people will want this dynamic addressing wherever they can, which is believed to mitigate that risk anyway.

Hopefully there won't be an explosion and we have the reporting by the Secretariat to keep track and ...

if it proved that there was a real problem.

Again, this process is entirety consistent with all the other RIRs. So it would be through everyone's consensus and the only changes would be ... IPv6 other assignment policy. Portable assignments.

RIR similar policy.

That is the proposal.

Masato Yamanishi: Thank you, David. It was very good wrap-up of what was the result of the previous meeting and what was changed and what is the current discussion on the mailing list.

Is there any comment, clarifications? Can I close the microphone? Paul, is there any implementation issue on this proposal? Sorry, Dean, go ahead.

Dean Pemberton: We support the proposal. That's all.

Masato Yamanishi: Okay.

Aftab Siddiqui: Actually, we vigorously discussed this proposal on the mailing list and I just want to say that I support this proposal as is and we have no objection.

Thank you.

Randy Whitney: Verizon, I also support it as is. I think it helps to promote v6 adoption which is a good thing.

Guangliang Pan: From implementation point of view, we don't have problem to implement it, just lacks a bit on the criteria.

Valens Radia (IDNIC): We are supporting this proposal, as to support the IPv6 implementation.

Masato Yamanishi: Is there any further comment? Gaurab Raj Upadhaya: I personally support this proposal, but I think on the mailing list, we had one issue that was raised. And I don't see Terence here, so I think I'll kind of try to look at that view and I thought that as a Hostmaster, I think the Hostmaster probably want better clarification of what technical justification can mean, because they don't confirm a technical background all the time. And that is something when a senior engineer at a big telco says, this is my technical justification, the Hostmaster might not be able to say no, that is not a valid technical justification or, you know, a reason to justify it. That leaves a big vague thing to anybody to write in the name of technical justification and throw it at the Hostmaster. I think that concern is there.

So I support the proposal as is, but to bring back what Terence was bringing on the mailing list, I was thinking, like, maybe if the Secretariat could publish the list of technical justifications used on an annual basis, or, you know, there are ways in which we can know what justifications are valid, there might also be cases where people don't know what to do. They say I need IPv6 and the reply is what is your justification or I need IPv6, you know, there might be a multihoming justification or something.

So I will still suggest a list or annex with a list of possible justifications are not hardcoated into the policy itself. Give a bit of leeway to the Hostmasters.

At the same time, they should not be overwhelmed by somebody claiming technical justification in the name of technical justification. Thanks.

Masato Yamanishi: Izumi and Dean, your question related with Gary? Izumi Okutani: Since we are also Internet registry, we can understand what the comments that Terence has made and Gaurab has mentioned and we felt that under the revised version, the proposal has given some examples, not limited, but specific examples, so it would give us an idea of what kind of cases that we can improve. It would certainly help us if it's documented somewhere, not necessarily as a fixed policy, but, for example, in the guidelines document, as a set of example of the cases that APNIC could approve and if there are any cases that's not listed, we can always consult APNIC Hostmaster.

So I think we can work that way with the revised version. So JPNIC would support the revised version of the proposal.

Masato Yamanishi: So you mean it is doable with current text, right? Izumi Okutani: Yes.

Gaurab Raj Upadhaya: I think the point is I'm just thinking from the other side, why that is necessary. As I said, the policy as is I support it, there is no question about it, it's necessary.

But if the IPv6 address allocation guidelines, that is the document published by APNIC, can be updated to include a couple of examples, so that helps somebody completely new to the process to justify it, as well as a set of base guidelines. Of course, APNIC will set up a procedure saying if there is a new technical justification coming in, it could be sent over -- Masato Yamanishi: I think Guangliang has an answer to your question.

Guangliang Pan: I think I can answer his question. Yes, I think the APNIC Secretariat can review the guidelines and then make it more clear and maybe give some example in the guidelines, so the people can easily use the policy.

Gaurab Raj Upadhaya: But it should not be limited to those examples. That should be the key.

Guangliang Pan: I think this should not be limited to those examples, of course, yes.

Dean Pemberton: Yes, I think we shouldn't put more criteria on here than we do for people getting other allocations. You had some justification that might be technical when you're applying for justifications now, so technical people within large telcos could put forward a whole lot of technical justifications, so that's kind of a problem we have anyway.

So I agree that we could put in a small set of examples, but just as long as we don't mean that it's only those examples and we don't make this actually a much harder thing to do than getting any other allocations.

Gaurab Raj Upadhaya: I agree with you. To repeat, I would like some of the examples to be in the guidelines, but that should not be the only reasons for justification.

I agree. Thank you.

Masato Yamanishi: David, are you still on-line? David, do you have any comment for Gaurab's comment? David Woodgate: I just want to say that I repeat that please understand that while four wrongs may not make a right, the text in the draft is pretty much as specific as any of the other RIRs and perhaps more so than most.

The comparable one is probably ARIN. I could go through the things. They're in the appendix of the pack, in people want to go through them, but you have, for example, AfriNIC, end site must justify the need for IPv6 address space is the only statement. ARIN, as I said before, it's either being multihomed or more than 2,000 addresses or more than 64 subnets, which, to be honest, I think are fairly arbitrary sort of specifications, but still a reasonable technical justification. LACNIC, I don't think actually requires details of how the addresses are going to be used, but I don't believe has an actual technical requirement or justification requirement as such. RIPE, likewise, the actual text is very vague as to whether what qualifies.

It doesn't appear to require a specific technical justification.

I'm just advising that if you try to get too specific about this, you will be different from the other RIRs. But again, as I say, four wrongs don't necessarily make a right.

Masato Yamanishi: But David, I think Gaurab's point is a little bit different. He's saying -- I think Gaurab's intention is not changing criteria, technical justification criteria. You want to add -- okay, go ahead.

Gaurab Raj Upadhaya: I just want some examples to be available in the guideline document. As I said, I support the policy as it is written right now. I'm fine with that. But when this goes to implementers, I would like to see some of these examples in the guideline document, not in the policy document.

David, do you get that point? Masato Yamanishi: David, do you get the point of Gaurab's point? David Woodgate: I do now. I'm sort of having to read the text rather than listen to the audio.

So you're essentially suggesting the Secretariat volunteer for that, you know, beyond the proposal? Gaurab Raj Upadhaya: Yes, more or less.

Masato Yamanishi: Izumi, you have a comment? Izumi Okutani: Yes, I just basically want to add that I support Gaurab's suggestion that certain criteria in the policy, but give examples to give people idea of what specifically technical justification would be and certainly not limiting to what's being listed. I think it's a really good idea and it will certainly help JPNIC Hostmasters and also people who would apply.

So I would support Gaurab's suggestion.

Gaurab Raj Upadhaya: I think we hear back from David? Andy Linton: I'll just chip in here. I have a note from Sanjaya, saying -- I think it is, but Adam, just correct me here if I'm getting this wrong -- the Secretariat can update the IPv6 guideline document, which is -- Adam Gosling: That's from Sanjaya.

Andy Linton: -- with examples of reasonable technical justification. So that sort of I think covers -- Gaurab Raj Upadhaya: That covers what I'm talking about.

Andy Linton: Presumably, if there were new and strange corner cases, then the Secretariat would report them back to the community and say, "Here is a new justification," or we would be able to see that document anyway.

Adam Gosling: It's a numbered document.

Andy Linton: So if we see any changes to that with new and strange cases, we could deal with it then.

Gaurab Raj Upadhaya: I think that's all right. It's better if that list of things are published and it can go longer over time, but it will give an indication to Hostmasters, as well as to people who are trying to apply for it, can see and look at it and say, "Okay, I can get v6 upload this." For people who already know why they need v6, this is more important.

I wanted those examples to be here somewhere and the v6 guideline documents will be the place to be.

Thank you. I think we are all in agreement here.

Masato Yamanishi: It seems there is no further comment.

But David, you have anything you want to say? David Woodgate: No. Thank you, everybody. Does that mean that the proposal has been accepted? Masato Yamanishi: We have not yet asked the consensus.

Andy Linton: Next stage.

Masato Yamanishi: If there is no further comment, let us ask the consensus.

Andy Linton: Okay. Given that those people who want to have a say have had a say, can we just try -- for the Chairs, can we have a show of hands so we can try and gauge what the process here is now. I have an actual script that I have to follow here, so we get it right.

This policy proposal has been discussed on the APNIC Policy SIG mailing list and at today's open policy meeting the APNIC PDP requires the community to reach consensus or a general agreement on the proposal before it can proceed.

To do that, what I would like to see is a show of hands from people, by those people who support or strongly support this proposal.

Can we see who supports that.

If you are going to put your hand up, put it up all at the same time, so we get a good feeling.

Anybody who is opposed to this? We also have two remote hubs listening to us and I just wonder if we can get any feedback from them.

Adam, are you shaking your head to indicate there is no feedback? Adam Gosling: Yes, I am not getting a response.

Andy Linton: Have we any indication how many people are at the remote hubs? Just so we can take it into account.

Are you making any progress, Adam, or should we just move on? What I'm hearing is that we have support from the remote hub in Indonesia and nobody against it, so I think, if I look at my -- so I have conferred with my Co-chairs and we have noted the communities for and against the adoption of this proposal and in consideration of these inputs, we believe that the community has reached consensus on prop-101.

The next step in the policy development process is to take this proposal to the AMM and seek a continued consensus there and then the proposal will be posted back to the mailing list for an eight-week comment period.

That's that item finished.

David, I just congratulate you on getting things through and we'll move on to the next item.

David has actually passed back his things to everyone, so thank you.

Our next item is an informational discussion from Tomohiro Fujisaki. I'll leave him to introduce it, but it's related to stuff in prop-097 about delegations of IPv4 address space in the APNIC free pool.

Tomohiro Fujisaki: Good morning, everybody, and thank you Chairs for giving me the opportunity to present this presentation.

My name is Tomohiro Fujisaki and this time I'm with the hat of JP IPv4 address allocation discussion team.

The purpose of this presentation is to ask APNIC community if there are needs to revise current final /8 policy to distribute non-103/8 and to distribute non-103/8 blocks in APNIC pool ^.

This is current IPv4 allocation policy. We are now using final /8 policy to distribute IPv4 address block.

With this policy, LIRs can get maximum /22 and this is the point that same policy applies for the distribution of the IPv4 address block to APNIC and reallocations from APNIC to IANA.

We think situation changed for IPv4 address allocation changed after current policy decided. One reason is new global policy was ratified by ICANN board and will be implemented soon.

This global policy defines a rule to reallocate returned IPv4 address block to IANA to RIRs. Recently announced some RIRs, ARIN, RIPE and APNIC returned their historical IPv4 address space to IANA and it is estimated in my calculation correctly, about /10 IPv4 address could be reallocated to APNIC.

So under this situation, we think it might be better to reconsider current IPv4 policy. We understand that the essence of current final /8 policy, that is to secure final /8 block for future APNIC members.

However, we have some thought about non-103/8 address block which supplemented to APNIC pool.

This figure is some of our concerns. Currently, the final /8 policy applies to all APNIC pool. We think it's okay to keep this 103/8 address space for future use, but we can utilize newly re-allocated or returned block.

So to make sure the needs, we conducted a survey to the LIRs under the JPNIC management.

With this survey, we wanted to find out it will be helped if LIRs can get more IPv4 address block, in addition to the final /8 address and if they get it, their IPv6 deployment plans changed or not.

This is overview of the survey. The period is about two weeks, a little bit short, and the target is LIRs under JPNIC management. The response, we got 61 responses from LIRs, this is 15 per cent in all LIRs.

We asked these six questions to LIRs. The first one is could you apply in additional IPv4 address is available? Question 2 is usage of IPv4 address for such allocation. Question 3 is effect on IPv6 deployment if they have more IPv4 address. Question 4 is minimum size required to make such allocation useful. Question 5 is opinion about revising the current final /8 policy. And last, free comments.

In question 5, we introduced current discussion in the Policy SIG that is their opinion that changing the current IPv4 policy which should put ... on IPv6 deployment and taking such opinion into consideration, LIRs still feels needs to reconsider the final /8 address block.

This is a summary of the survey result. About 70 per cent of LIRs said they will apply if additional space is available. LIRs who answered will apply a higher percentage of IPv6 allocations. 84 per cent of LIRs answered they have no effect to their IPv6 deployment if they have additional IPv4 address.

71 per cent answered they were in favour of revising final /8 policy.

Detailed result is attached with my material on the web, so if you want in more detail, please refer my document.

This is final slide. From the survey, we might be some demands for revising final /8 policy in Japan.

I made this same kind of presentation at the NIR SIG last Tuesday and asked to conduct the same survey in other economies.

If possible, I want APNIC to conduct this kind of survey to the APNIC members. Also, please give us your comments, your thoughts about revising the final /8 policy.

Okay. This is end of my presentation.

Andy Linton: Thank you for that, Tomohiro.

Has anyone any comments that they would like to share on this or any questions? Kenny Huang: First of all, I would like to thank JPNIC for doing this kind of extensive research regarding this issue. And basically, I'm not sure whether you have conducted the life cycle management, for example, for the final /8 policy and based on my position, based on the research outcome from Geoff Huston he did last year, the final /8 allocation in APNIC region based on our current member growth rate was roughly around 15 years, then my gut feeling if we just add one more /10, probably a few more years.

So I think even a final /8 can last long enough, based on our current membership growth rates. So have we done this kind of life cycle estimation based on your research? Thank you.

Matsuzaka Yoshinobu: At this moment, I don't support revising the final /8 policy. This is just comment.

Tomohiro Fujisaki: Could you please teach me the reason why? Matsuzaka Yoshinobu: Because revising the policy makes chaos. I don't like to see such situation.

Tomohiro Fujisaki: Thank you.

Randy Bush (IIJ): I'm not colluding with Maz. I think if I understand, really, there's the final /8 policy which applies to 103/8 and then there was a subsequent policy passed that said, anything we discover later, we will apply the same policy to it. Is that correct? Tomohiro Fujisaki: Correct, I think so.

Randy Bush (IIJ): What you're proposing, really, is not the final /8, 103/8, you're saying leave that alone.

You're saying the other stuff, you want to rediscuss that second proposal that said the things discovered since will be treated the same as the final /8. Am I correct? In other words, you're not saying change the policy for 103/8. You're saying change the policy for anything that comes after that. Does anyone know the policy number I'm talking about? Tomohiro Fujisaki: Yes, exactly.

Randy Bush (IIJ): So we should be careful. We're not talking, really, about changing the policy for the final /8. We are talking about changing policy for things that come after that we're applying the same policy to it, but it's not the final /8.

I just wanted to clarify that and make sure we understood.

Tomohiro Fujisaki: Thank you, Randy. Yes. Thank you for your clarification.

Randy Bush (IIJ): In which case, I am ambivalent. In other words, you know, what I cared about -- I was involved in the final /8 policy -- was that there was enough space and allocations to go on for many years for new people entering the network. Especially as IPv4 becomes very scarce and it's an IPv6 world, they're going to need little bits of IPv4 to stick in front of their networks and do NATs to IPv4 or something.

That's what I cared about in the final /8 policy and I think there is enough space in 103/8 to last for a long time and accomplish that goal.

So the other stuff, I think we're discussing very small amounts that will not last very long, et cetera; as I said, fighting over scraps. But it doesn't damage the final /8 policy.

Dr Linton, you had an actual policy number, so we can have nice nouns to differentiate things.

Andy Linton: I'll just point out that the policy that Randy is talking about is proposal 88, which is the one that was titled "Distribution of IPv4 addresses once the final /8 periods starts".

I think the key sort of paragraph in this is: "When APNIC has distributed all other available IPv4 resources and has started distributing from the final /8, any IPv4 resources received after that point will be placed into the pool and this pool will occur even if the result is larger than a /8." Randy Bush (IIJ): I think just, you know, I have a policy 103 I think it is, proposed to reduce policy wonkage, and here I am being a policy wonk.

But I would just like to clarify that Fujisaki-san's proposal is asking to change 88, not I think the number was 55 or 50 or some stupid number, which is the final /8 policy.

Tomohiro Fujisaki: Yes.

Brajesh Jain: SpectraNet India. I support this additional /9 IP addresses which are becoming available either returned or reallocation from IANA and this will surely help some of the ISPs who are old and they are sort of growing. So the /9 contribution in whichever manner it deems fit, it should be distributed while 103/8 continues at /22 basis. Thank you.

Tomohiro Fujisaki: Thank you.

Aftab Siddiqui: I seriously don't support this thing, because we have just passed prop-101, which will ease the IPv6 availability for everybody. As Randy said, we are in the IPv6 era, we shouldn't be asking for more IPv4. 103/8 will last for 10 or 15 years, because only /22 will be provided.

If let's say, as per your estimates, APNIC gets a /10, we all will be asking for /18, /17, /16, of course. It will run out in six months.

So you are asking for a proposal which will only last for six months or maybe a year. Right? Tomohiro Fujisaki: No. Just now, I personally think we distribute this additional space as same distribution policy in final /8, for example, applying this to 22 or 21 to LIRs, not to distribute the final /8 policy.

Aftab Siddiqui: Again, let me just clarify. For the last /8, it will remain like this, right? Tomohiro Fujisaki: Yes.

Aftab Siddiqui: There is no change in the last /8, but for all the new recovered space or the space we will get from the IANA, you are proposing or you may propose later on, to enhance the limit from /22 to /21 or /20 something? Tomohiro Fujisaki: Actually, current LIR can get /22 from the 103/8 block and one more /22 or /21 from the additional block, if they want.

Aftab Siddiqui: Okay.

Andy Linton: I mean, my interpretation of your question and from what is that's exactly what -- rather than it being a 21, it would be to 22s, is that right? Tomohiro Fujisaki: Yes.

Aftab Siddiqui: Okay. Let me put it that way, just to understand. The last /8 policy will be applied to 103/8 as well as the new recovered address space. Am I correct? So I'll get a /22 from 103/8 and a /22 from the new recovered address space? Tomohiro Fujisaki: Yes. Actually, just now, I don't -- I didn't propose distribution method, just ask for the needs to revise.

Aftab Siddiqui: But do you have any -- Tomohiro Fujisaki: Just now, I personally think, just you say we apply the final /8 policy to the existing final /8 block and the new block, yes. But I think the size is maybe it's okay to distribute 21 or 20, depending on the number of LIRs in APNIC or something like that.

Aftab Siddiqui: Because this is not a proposal, so we cannot say we don't like this proposal or hate this proposal, it's just a study. So I would say it's a good study, but the end result, it may lead to -- I don't support it. Thank you.

Tomohiro Fujisaki: Thank you.

Ren-Hung Hwang: Ren-Hung Hwang from TWNIC. I think it's a good start point to think about this question, but there's another side to look at it. There's a question of whether current allocation of /22 is not enough or is not good enough or for some company to start up or I think the final /8 policy, is there any problem for allocating /22? If not, then I think it's not good time to discuss the rest of this returned or recovered address. This is so complicated to think about how much to allocate from this returned or recovered address space.

Thank you.

Tomohiro Fujisaki: Actually, the result in Japan LIR says they want additional block to the final /8 block if they got 22 from the final /8 and if possible, more block they want.

Ren-Hung Hwang: My question is why they need more? Do you have some explanation or any evidence that they do need more? Tomohiro Fujisaki: Actually, in our survey results, they can use additional block to their equipment, something like that, IPv6 transitional ... existence. So they can use even the /22, they can use that block in many purpose, yes.

Randy Bush (IIJ): First of all, proposition whatever-it-was, the final /8 policy, does not say "/22", it says "the then current maximum allocation".

So if the then current maximum allocation becomes a /34, it will be a /34.

No. 2 is, yes, if Fujisaki-san's proposal or idea goes forward and you can allocate -- get space out of that /9, it will not be enough. Nothing you can do will make it enough. V4 is gone. Bye! Dean Pemberton: I agree with Randy that we are talking about scraps. But I think it's important for people to know what would happen to address space that they could decide to return, because I think that's going to impact their decision to return space or not.

So I think the questions are worth asking, I think the conversation is worth having, as long as we are just cognisant that this isn't going to be a sea change in fixing people's current IPv4 exhaustion problems.

It is deckchair shuffling, but if it's going to help people to make a decision to return the address space or not, then asking the questions and making a better policy about what happens to them, yes, I think that's worthwhile.

I'm more than happy to go back to the New Zealand community and ask a similar six questions to the ones that were displayed here and then report back to JPNIC what the results are.

Andy Linton: Thanks for that.

Gaurab Raj Upadhaya: I agree with Randy. We have gone into the last game, the end game, and we are trying to change the goalposts. It's not really worthwhile doing that. We just create too much chaos. And bringing /9 addresses back, APNIC actually has /10 worth of pre-approvals for people who need address space. So that address space will be gone in, like, one day. Just consider the pre-approvals and everything.

So, you know, I will say let's not change the goalposts. We have gone into the game. It's finished, it's not worth playing around changing things at this point. Thank you.

Izumi Okutani: I have been observing what's been happening, like discussions in Japan, so I just thought that I just add a little bit of the background and I totally understand a lot of the points that people are making, that it doesn't solve the long-term problem and it's just it might save people for additional one year, but then it doesn't really lead to solving any problems.

People in Japan were very much aware of that and then the need for v6 deployment, but even being aware of that and being prepared for v6 deployment in parallel, they felt that they also still need v4 in parallel.

So I think the reason why Fujisaki-san presented was that was simply a case in Japan and we simply wanted to find out what the APNIC region as a whole would feel about it and if everybody feels exactly the same way that others who spoke on the floor like, oh, it's not worth discussing, it's just scraps, then of course we wouldn't move ahead. But if there are substantial number of people who feel that it might be worth reconsidering the current prop-088 based policy, then we thought it might be worth considering, and that's Fujisaki-san's motivation. I just wanted to add.

Andy Linton: Thank you for that clarification.

Randy Bush (IIJ): Just to be clear about my opinion, I think the final /8 meets my goals. What you do with the rest, I don't think this does harm. I don't think what you are proposing or asking is going to hurt anybody, so I'd rather not spend a whole lot of time on it. I think the worst thing it does is make the APNIC web pages even more unusable. But that's neither here nor there.

Andy Linton: Sunny has a question from David Woodgate, which he's going to read for David.

Sunny Chendi: There is a comment in the Jabber chatroom from David Woodgate: would the APNIC Secretariat like to comment on the complexity of reopening the IPv4 queues especially in terms of assuring a fair and transparent process of allocations for questions where it can be assumed that requests will greatly exceed the supply of addresses that will be available? Paul Wilson: I think implementation would be difficult, but I think the more difficult thing would be for this community to actually agree on any set of policies that would make it feasible for us, that would result in a direction to the APNIC Secretariat to implement anything, frankly. I think that's a pretty difficult ask.

I mean, by all means, open a discussion about how it would work, but until we knew the policy side of how we would actually manage such a thing, I don't think we could really comment on how the implementation would be done.

Andy Linton: Paul, while you're there, one of the other sort of questions that was asked which was similar to David's was are we talking about effectively saying a /22 out of the current 103/8 and another /22 out of the other parts of the pool? Would that be more straightforward or would that be difficult to implement as well? Paul Wilson: Not difficult at all, no. Our systems will track multiple pools of addresses and allow those things to be done.

Andy Linton: You would effectively be boosting the allocation size to /21 or two /22s until the other stuff ran out and then we would go back to /22? Paul Wilson: That's right. My reading of the proposal is that it wouldn't alter the final -- the trajectory of the consumption of the 103/8 at all, because it would tend to be a second allocation to people who had already received the first and make no difference to the consumption of the last /8 block itself.

While I'm here, I think it's worth just reviewing the numbers here. There are 16K, 16000-something /22s in that last /8 block. We have consumed less than a tenth, but I think 8 per cent of that block so far.

I would point out, though, that at this time, we are seeing -- and you'll see this tomorrow -- we're seeing a pretty rapid acceleration in new memberships to APNIC.

In rough figures, I would say our new membership, our rate of new membership has gone from 500 a year to well over 1,000 a year and we're sort of in the stage at the moment of breaking monthly records each month. If we were to keep on that rate of growth then we would have really a lot of new members and we would be consuming the last /8 pretty rapidly.

I remind you, RIPE NCC has over 8,000 members and if we had that then we would have consumed half of the last /8 pretty much very quickly.

That needs to be borne in mind when we consider why the last /8 policy was made in the first place and my recollection of that is that we wanted there to be a rationed supply of IPv4 address space that would last through the entire v6 transition. That is, it would last as long as there was going to be any future need for IPv4.

That could still -- that, I think, has got to be many years left, so you put the rate of consumption and the likely ongoing demand together, I think we're not in the same position as we were a few years ago, I think, where the last /8 policy was argued as something very overly generous and it was potentially wasting IPv4 space.

Anyway, that's sort of a set of personal opinions.

Thanks.

Andy Linton: Thanks for that. We're actually approaching the point of where we're coming to the end of the session, it's almost 12.30.

I would like to thank Tomohiro for raising this issue and I'm sure it isn't sort of quite finished. I'm sure there will be some more discussion on it. But this isn't a formal proposal, so it's not a question that we have to sort of try to gauge consensus. It's really just for information or seeking information in both directions.

I think if some more information appears on this, and people would like to communicate with you.

Tomohiro Fujisaki: Thank you so much.

APPLAUSE Andy Linton: So we'll now break for lunch. Sorry, Prof Kanchana has a brief message for you about the election process.

Kanchana Kanchanasut: Just like to remind all of you that, you know, please cast your vote at the desk outside this room. Don't forget to do that. We'll close the voting at 2 pm sharp. Thank you.

Andy Linton: Okay. That coincides nicely. You can cast your vote and then come back to the afternoon session where we have two more proposals to discuss and we'll see you back then at 2 o'clock.

Kanchana Kanchanasut: It is now 2 o'clock, there are no more votes in. We will start counting. If there is any complaint, please come and see me. Thank you.

Andy Linton: Thank you for that. We will start into our final session. We've got two proposals on the table. We will deal with prop-104 first, Shin Shirahata. There are a couple of co-authors as well, but he will tell you all about the proposal. I remind you that we will do this proposal, then we will go on with the final proposal we are going to deal with, prop-103, and during that process we will have the results of the election at 3.00.

Shin Shirahata: Good afternoon, everyone, my name is Shin Shirahata from Clara Online. We are a web hosting company in Japan. Today I am going to talk about the new policy proposal on IPv4 address transfer.

I would like to explain about the background of the proposal. The background of the proposal is insufficient period for demonstrated needs from IP before address transfer. Since prop-96 has been adopted, demonstrated needs is a requirement to recipients of IPv4 address transfers. The period of demonstrated needs under the current operational practice is 12 months.

In this proposal, I would like to clarify the requirements on a period of approval for the transferred resource to recipients of IPv4 transfers based on the demonstrated needs and define its period as 24 months. In this slide, I would like to explain the current problem. One of the biggest problems is that 12 months is too short for network planning. Many ISPs and many Datacentres plan their network under their addressing plan beyond 12 months, but the current operational practice allows for only 12 months' needs. There is no guarantee for IPv4 address availability after 12 months, so service providers need to have a longer period for their justification. The address transfer process involves a lot of paperwork or legal advice and it means a shorter period increases the hidden cost of the IPv4 address transfers.

Also, the current requirement is designed for a new allocation from APNIC IPv4 address pool. So I think there is the implication that the recipient could obtain continuous allocation but, as you know, the free IPv4 address pool has already run out in the APNIC region. Here is another problem of the current policy. At this moment, APNIC and ARIN are the RIRs that adopt the inter-RIR address transfer policy, but now ARIN extended their period for demonstrated needs from 12 months to 24 months. If the address transfer happens from APNIC to ARIN region, up to 12 months' demonstrated needs will be justified.

On the other hand, transfer from ARIN to APNIC region, only up to 12 months will be justified. The difference is creating unequal conditions to APNIC members. I would like to share another issue on the current policy. This is an unnecessary route fragmentation. Here is an example. If recipient B needs /22 within one year and /21 within two years, in this case transfer of /22 is allowed from source A to recipient B. At that time, source A still has a /22, so source A needs to find another recipient for IPv4 address transfer, because of transfer size mismatch. After one year, the recipient needs to find another source for IPv4 address transfer because they need another /22 block. But the source may have a large block, and if the new source B has a larger block than /22, /21 will be divided into /22. As a result, it causes unnecessary route fragmentation.

Next, I would like to explain the situation in other RIRs. For AfriNIC, AfriNIC doesn't have an IPv4 address transfer policy at this moment. For ARIN, they have an IPv4 address transfer policy and the period for the justification is about 24 months. They also have inter-RIR transfer policies. The justification period was extended from 12 months to 24 months this February. For LACNIC, they also have an IPv4 address transfer policy, but they don't have an inter-RIR policy at this moment. For RIPE NCC, they have an IPv4 address transfer policy but the period for justification is limited to three months at this moment. There is a policy proposal under discussion which proposes to extend the period of the demonstrated needs, in case of IPv4 transfers. Also, the policy proposal for inter-RIR transfer is under discussion.

I would like to explain the outline of my proposal. This proposal clarifies the requirement on a period approved for the transferred resource to recipients of IPv4 transfers based on the demonstrated needs and defines its period as 24 months. This change affects only IPv4 address transfers. Twelve months' requirement remain for last /8 block. The reason for the proposal is demonstrated needs for 12 months is not sufficient for IPv4 address transfers in post-IPv4 exhaustion era. It is also important to ensure APNIC region does not have disadvantaged condition in inter-RIR transfers.

In this slide, I would like to explain why a 24-month period is important. According to the LIR address space management in APNIC-124, address estimates should be made for immediately, within one year, within two years. Also, APNIC-89 requires a two-year projection for address assignment for IPv6. In addition to that, 24-month projection is the same period of ARIN region. I would like to explain the benefits of this proposal. APNIC members can apply for address transfer as a recipient under the same condition of demonstrated needs as other RIRs in case of inter-RIR transfers. So if APNIC member obtains the IP address from ARIN members, the same conditions will be applied. Here is another benefit for extended period for justification. The extended period will allow the larger block size to match the longer-term needs of the requester. In the last example, source A could transfer to recipient B for /21, needs for two years. So source A doesn't need to find an extra recipient, and recipient B also has no need to find an extra source. So it also reduces the risk of unavailability of the IP address resource after one year later. As a result, it will help to reduce IPv4 address block fragmentation by transfer.

It also has another benefit, it reduces the risk of underground IPv4 address transfers which do not get registered in APNIC database, because there is a possibility that the recipients could not obtain justification for enough IPv4 address by the current period of demonstrated needs. A 24 months' justification period decreases uncertainty of the address transfer. It decreases incentives of underground transfer. Disadvantages. We do not see any disadvantages to this proposal. Implementation. I think it's not difficult to implement this proposal. I expect that it can be implemented within three months from the consensus, if consensus is reached. Impact on NIRs. It is up to the NIR's choice.

As for this proposal, I would like to summarize what I have been talking about. We propose a new criteria for demonstrated needs up to 24 months for the transferred resource to recipients of IPv4 transfers. It is in line with other RIR policies, in case of inter-RIR transfer. Also, if we adopt this policy, the recipient could obtain a larger address block on one transaction. It will have help to route aggregation on transfer. Also, it will help to reduce transaction costs on IPv4 address transfer. Also, it will help to reduce the risk of underground IPv4 address transfers. Thank you.

Skeeve Stevens: Thank you very much, Shin-san. I guess we will open it to the floor for any questions.

Dean Pemberton (InternetNZ): I think, having a look at the New Zealand community, I don't know that a lot of our members particularly would take advantage of this, but I can certainly see, through your presentation, that there are certainly advantages for other members in being able to reduce the compliance costs, for going out year on year to try to get transferred addresses, and certainly given that it would appear that APNIC policies, even if we ignore what's happening in ARIN, other APNIC policies centre around this two-year block of allocation, it certainly seems that if we were to extend it out to two years, we would bring it into line with the rest of our own policies, plus we would reduce the compliance costs for APNIC members that are going to go out and find the addresses. We wouldn't force them to go out and get it year on year on year and then get potentially non-aggregatable blocks. So we would support the proposal.

Shin Shirahata: Thank you very much for your comment.

Skeeve Stevens: Any other comments about the proposal, for or against? Can I ask for a comment from APNIC, from Guangliang about the presentation's comments regarding the implementation?

Guangliang Pan (APNIC): I think the Secretariat has no problem to implement this policy, but the timeframe, I'm not sure yet because we have to change some software forms. What I want to clear up a little bit is that the two years means that you definitely will use that amount, right? If you just have an allocation of /16, it means that you are going to use the /16 in two years' time.

It's not like, if you look at the IPv4 policy, there's an initial allocation criteria and subsequent allocation criteria. So I assume this transfer is like a subsequent allocation, is like apply to the 12 months and now changed to 24 months for this transfer case. What I mean is, in the initial allocation you say you only use 25 per cent immediately and 50 per cent in a year's time, that is 12 months, but for subsequent allocations you normally need just the amount you are going to use in the next 12 months.

I'm not saying this applies to any IPv4 request, like initial request or subsequent allocation, but all I'm saying here is your meaning of this two-year period, it is like exactly the amount if you apply for that two years, it's not like 50 per cent and then you can say -- for example, in two years' time you have used 50 per cent of the amount you applied for. For example, if you request for approval for up to /16 and then you tell me you only use /17 in two years' time, which is like half, something like that. I just want to clarify this point.

Shin Shirahata: Thank you for your comment. As my understanding, the original current proposal is designed for allocation from APNIC, not for IPv4 address transfer. So if IP address allocation is made from APNIC, the requester could expect for continuous allocation from APNIC, I think. But for IPv4 address transfers, the requester has more uncertainty about the IPv4 address availability, so 24 months' justification window is important, is now becoming more important than previous time.

Guangliang Pan (APNIC): Okay, I think I probably understand, exactly this 24 months, then we can approve up to 24 months for transfer request, if that is the policy you want. Okay. I think I just want to say, we have no problem to implement this policy.

Shin Shirahata: Okay, thank you.

Randy Bush (IIJ): I think you are completely correct: if you are asking for a /16 from APNIC, you have a definite clear expectation, the answer is going to be "no". It's not there. We're gone. IPv4 is gone. Get over it. Okay. If you are asking on a transfer, you are completely correct, it's not sure, because you might find somebody who is willing to sell it to you. Okay? Then we are going to hit a little problem, and somebody in the Secretariat better do the math. I'm new, coming to the Internet, a new member; what do I have to do to qualify for that /16? I may have a perfect justification for I will use this /16 in the next three months or two years or whatever. But if it's incremental justification, how do I get there? Am I being clear? For subsequent allocations, I have to have used a /22 to justify getting a -- I think Guangliang has the answer.

Guangliang Pan (APNIC): Randy, thank you very much to raise the very good question to the Hostmaster. How the ostmaster evaluates the request; right?

Randy Bush (IIJ): No, say the request is very clear, I have all those devices, they are sitting there, I can send you pictures of them. I need the address space. Do I have to sneak up on it? Because we do not currently have policy that gives somebody a /16 on day zero, do we?

Guangliang Pan (APNIC): You mean a new member?

Randy Bush (IIJ): Right, cannot get a /16.

Guangliang Pan (APNIC): You get pre-approved for /16. We do have exemptions for initial allocations larger than the minimal allocation size in the past, as you can demonstrate, as you say, clearly a pair.

Randy Bush (IIJ): Fine, not a problem.

Guangliang Pan (APNIC): Thank you.

Aftab Siddiqui (Cybernet Pakistan): This proposal is based on the transfer policy of ARIN; am I right?

Shin Shirahata: Yes.

Aftab Siddiqui: Of course, we got /18 from APNIC a long time ago, and we used it up in six months. We had a huge broadband project, we used it. I don't see any reason that 12 months' justification won't be enough. Of course, if you are going for /16 or something bigger than that, you must have a very clear justification. Let's say if RIPE, like RIPE opened up the transfer policy and somebody wants to transfer it over there, they have three months' justification. Right? So people are happy over there with three months' justification as of now, if I'm not mistaken, RIPE NCC can clear that up. Maybe this proposal will solve a couple of corner cases, but in a generalised format I don't think so, it's not a help anyway. Thank you. I don't support the proposal.

Shin Shirahata: Thank you.

Skeeve Stevens: Do we have any other comments before we close off? I guess we should move towards the process for consensus.

Andy Linton: This policy proposal has been discussed on the APNIC Policy SIG mailing list, and at today's open policy meeting. The APNIC PDP requires the community to reach consensus or a general agreement on the proposal before it can proceed. At this point I would like to ask people in the room to indicate if they support the proposal, and can you do that by putting your hands up, and keep them up, so that we can see what your view on this one is. Those people who support this proposal, will you please put your hands up, so we can clearly see them from here.

Skeeve Stevens: Please clearly put your hands up. There are a few people that are quite low.

Andy Linton: Those people that don't support the proposal, can you put your hands up. I'm seeing a couple of people. I would like to ask the people at this point who don't support, is this something that you strongly disagree with or you just think it doesn't help? Could you explain a little bit more about why you think?

Aftab Siddiqui: I just shared my concern. I don't know, it's surprising for me so many people supporting it. I didn't see that much support during the Policy SIG mailing list discussion. As I said, it can solve a few corner cases but in a generalized way, as a policy proposal, it's not helping the community. Again, it's the v4 space we are fighting for. I would really support any v6 policy. Anything like that with v4, sorry, I cannot. This is my concern.

Andy Linton: Can I ask you a question: if you think it doesn't particularly help, does it do any harm?

Aftab Siddiqui: No, it doesn't harm anything, but it is not adding any value to the community or to the services.

Andy Linton: Izumi, did you want to say something?

Izumi Okutani (JPNIC): No.

Andy Linton: The other person who is here who wasn't supportive, is that your feeling as well?

Chanki Park: Same opinion.

Andy Linton: Is it possible for us to get a feeling back from the remote hubs? There are some people in Indonesia against this, and Nepal has gone to lunch. So it is hard for us to get an opinion from them. Is there any indication from the Jabber chat about the strength of feeling in the Indonesian remote hub? If we can get some feedback from that direction, it would be useful to us. The indication I'm getting from Indonesia, it's not a strong objection. In this situation, I think we're there. What I'll do is use the formal words.

I have conferred with my Co-Chairs and we have noted that the community's comments for and against the adoption of this proposal, so in consideration of the inputs, we believe the community has reached consensus on prop-104. The next step in the policy development process is to take the proposal to the AMM and seek a continued consensus there. The proposal will be then posted back to the mailing list for an eight-week comment period. That's our process for that. Thank you for presenting, and congratulations.

Shin Shirahata: Thank you for giving me the opportunity to present my proposal and thank you for your support and understanding.

APPLAUSE

Andy Linton: We move on to the final proposal, prop-103, which is the final IP address policy proposal which Randy Bush proposed. It started off as version 1 on the list, then there was a subsequent version submitted to the mailing list on Tuesday, I think is the right date. As I pointed out at the start of this, we do have a point just before 3 o'clock when we will have a commercial break for the election results, and if we are still going at that point then we will have that break, and we will come back to it immediately afterwards. Randy, please go ahead.

Randy Bush (IIJ): There was an amended proposal on the Tuesday and then there were some very good comments by Dean. Dean, are you here? There you go, because I'm going to stick you with speaking to them.

Dean Pemberton (InternetNZ): Awesome.

Randy Bush: Essentially, this region is less guilty of it than some, but we create a lot of silly policies, or small policies, or we don't create them but we discuss them too much and spend a lot of time on them. So I proposed to stop doing that. Now, I will admit to having two reasons. One is I actually hope to raise people's consciousness about this as a problem, so maybe in the future we will think a little more before we propose policies. Secondly, because I've kind of been in the policy community for a while and a lot of it really isn't -- this week is not the example, but we have spent considerable time on policies which we shouldn't have spent time on. Fujisaki-san pointed out something very good, which is that it's important that we maintain an open bottom-up member-driven process for coordination and cooperation. I specifically do not use the two words "regulation" or "governance". I believe once you start using those words you have bought the mental model of the enemy. So this is the Internet and we coordinate and we cooperate. I think Fujisaki-san was right, that we do need to maintain a mechanism and a community for continuing to coordinate and cooperate. I modified that proposal on Tuesday to essentially say that. Then Dean said, "Those are all very nice words but what's the operational meaning of it?" What do we do? That's philosophy; what do we actually do? Then he replied to his own post with a suggestion of what we actually do. Instead of putting words in Dean's mouth, I would like to invite Dean to put words in Dean's mouth. Sorry to put you on the spot, Dean, but you're a big boy.

Dean Pemberton (InternetNZ): No worries.

Andy Linton: Dean, would you like your words up on the screen?

Dean Pemberton: I've got it here, but up on the screen would be good. I'm calm either way.

Andy Linton: Sunny, is it possible to find Dean's email on the list and put up the block of text, his proposed friendly amendment? Just so people can see it. You can all refer to the mailing list and find this. This is the response to the posting of version 2 of this. I'm taking it that --

Randy Bush: My apologies for not coordinating this better.

Andy Linton: I take it Randy is accepting that as a friendly amendment?

Randy Bush: I don't care about the process.

Andy Linton: But I have to.

Randy Bush: Dean had a good idea.

Dean Pemberton: I won't do it again. Sorry. Found it. Obviously I post too much; there were too many emails to look through. I think one of the things that tends to happen with the policy development process is we require people to turn up to the process with something that's fully formed, and then we kind of batter it around and sometimes we agree and sometimes we don't agree. What we actually miss out on a lot of the time is the problem that people are trying to solve, and other possible solutions to it. Rather than just expecting people to come here with a proposal and then judging it purely on its merits, and not listening to what other people might consider a solution might be, or some people not even understanding the problem, what I thought we could do is maybe change a little bit of how the process runs and put a couple of steps at the start. Rather than saying that someone should just turn up with a proposal for us to talk about, authors could come to the Policy SIG mailing list with a problem statement. What is the problem you are currently seeing? All the proposals that we deal with here, on some level or another, deal with an actual problem. So start off by stating what their problem is. It might be a problem just for you, it might be a problem for the community, but this is the stage to work out succinctly what the problem is. The proposer and/or the Policy SIG Co-chairs could then lead a conversation on the Policy SIG list to develop possible solutions to that problem statement. Now we're not just restricting ourselves to what the author's idea of a solution might look like; we are actually doing that member-driven community interaction and giving everyone a chance to have a look at the problem, say whether it affects them and then get a better idea about what the community thinks that a solution might be.

All up until this stage, we haven't rolled the big armature that is the policy development process, we have had just a discussion on the mailing list, but already we are getting more community involvement and therefore more consensus about what would pop out of this process. It doesn't just come out one day and it's the first time people have seen it and the first time they get to have their say on it. I would also like to put in here as well, at this stage, I would like the APNIC Secretariat to comment on whether they see the problem statement could be covered within the current policies. In the lead-up to the meeting, we saw one of the proposals that was going to be presented here actually could be covered off by the current policies, and that was brought out on the mailing list and the author has since removed the proposal. A little bit more work upfront means we are not constantly coming to these meetings with a whole lot of proposals and trying to develop that stuff here. When the proposer feels that they have a solution to their problem statement -- and I'm not saying that they need the whole community to agree that they have a solution; my idea was to still leave that up to the proposer -- when they feel that they have a solution to their problem statement, then they draft a policy and submit it in a similar fashion to what happens today. What I think we would get out of making these changes is policies that people have already seen, they have already been socialised, people understand the problems that led to this policy, and hopefully we get policies presented in the policy development process at these meetings that people already feel they have had involvement in, and consensus should be a lot easier to get. Stuff would never make it to this stage if there were polar opposite views, I'm hoping. Those are the changes that I thought would help us, meaning that we weren't constantly seeing proposals brought through and not gaining consensus and really just, as Randy said, serving to run this process kind of for the sake of it. Thank you. Back to you.

Randy Bush: I think Dean's last point, the one that sold me, is that we are supposed to be still close to our engineering heritage, and finding ourselves looking at a proposal and trying to figure out what problem it addresses, what the scope of the problem is, what different flavours of the problem they are, et cetera, et cetera, is backwards. We start with the problem and then try to engineer towards a solution. So I think Dean's proposal is really good. If people like the direction we're going in general, then we'll work with the Policy SIG Chairs to try to work in Dean's text, and co-authors or not, and all that nonsense. We first need to know: do people think the problem this is trying to solve exists and that this approach is generally a reasonable approach? Thank you.

Andy Linton: Questions or observations, whatever.

Izumi Okutani (JPNIC): Just a comment.

Randy Bush: Sorry, I forgot. I was anxious to leave.

Izumi Okutani (JPNIC): I totally support the motivation behind the proposal and the modifications that Dean has suggested, it really helps. I think it's a really good idea to share with the community whether we think certain things are a problem and need a fix, rather than just directly into a solution, because sometimes people don't agree with the solution, but agree that there are issues that need to be solved. So I think it's a really good way of getting the process. Actually, I do have one clarification question. We do have our own policy process in Japan and I don't see any reasons that people would be against this idea, but we would also like to have some time and decisions of our own, whether we would implement the same process. So could we understand that it's up to us whether to implement exactly the same proposal in the JPNIC community or any other communities?

Randy Bush: I believe in this community the Japanese could decide to play roulette and it's fine.

Izumi Okutani (JPNIC): Okay. Thanks.

Randy Bush: I don't think we tell other people what to do.

Izumi Okutani (JPNIC): Great. Thank you.

Aftab Siddiqui: I totally support this proposal after the elaboration from Dean. I just want to add something, a comment from my end, that the policy proposal is a drafted document, it's a complete process, how you submit it to the Secretariat and how it works out, and the complete document comes out. The problem statement should be in the same manner. It shouldn't be that I get up in the morning and say, "I'm getting this problem, let's discuss." That's my point. It should be in a proper way, a simple draft which the APNIC Secretariat or whoever can put in a simple way, this is how you put your problem, explain it in a better way, and explain the issues you are facing with that in a proper way. That's my point. I hope this can be added to some extent.

Randy Bush: Actually, the current form/process/protocol/ whatever has upfront a problem statement, so we have the protocol for dealing with this. What we are really trying to do is to separate the two procedurally.

Tomohiro Fujisaki: Thank you very much for devising the proposal and for Dean's amendment. I totally support this proposal, but I have one question on behalf of the ASO AC, under this proposal we need to use this for all proposals? What I want to see is there are global policies from the other regions; maybe in that case we cannot use this process. Just I want to see that point.

Randy Bush: Let me see if I understand by rephrasing. In the global policy negotiations, sometimes proposals come on missiles from other continents, and should we be asking them to adapt to our modified process? What do we do now with a -- quote -- "global policy proposal" and our pro forma, fill out this form policy proposal requirement? How do we make those two work together?

Andy Linton: I think, from my understanding -- and I haven't been in this position when we initiated a global policy proposal, but I have been there at the end of the process -- it still needs to be submitted as a normal proposal in each region under their normal processes and passed under their normal processes and then taken back up again. I wouldn't normally comment on proposals because I think it's not the job of the Chair to do that, but in this case I think it's appropriate to make a couple of comments and observations because we are going to have to implement this. The people who are going to have to work with this are the Chairs and part of the Secretariat. I would like clarification from here, that none of this would preclude us submitting a policy proposal in the current fashion, if it was absolutely necessary. So if some unforeseen emergency required us to put in a proposal and go through the process really quickly, I wouldn't want to see this preclude that, but I wouldn't want to see that as abnormal behaviour. We would normally go through the problem statement, but we could telescope that and go quickly through it. I would like clarification on that sort of thing before we throw out the baby with the bathwater. I think that is important. The other process is that we have Randy's proposal and an amendment to clause 4 but we don't have a substantive text. We would need to have the substantive text in front of us before we could say -- that's not a difficult thing for us to do but I think we should have that. We would need to have the group here actually understand what the actual proposal was all the way through. The other thing I would like to do is ask -- normally it would be the Hostmasters and Guangliang would comment about this, but in this case I'm going to ask Adam because this is very much to do with the policy development process and what implications that has for the policy development process. I think that's appropriate to do. Adam, I would like you to comment.

Adam Gosling (APNIC): I think you kind of answered your question with the second part of your statement. The proposal as written isn't what we see on the screen. We are on version 2 of the text. If version 3 included this text, then I think there would be a problem with someone going directly to a proposal and not having the discussion thing. So if that's something the community wanted to build in, then we could tweak this text before it becomes version 3. I think, as written, no, it's pretty much -- it says "should be carefully examined". I guess there's a little bit of wriggle room in that, "should be carefully examined to ensure", it doesn't say "must be". But what change are we making if we do that? Does that make sense? My point is you need to be very clear about what it is you want and then put that in version 3 text.

Andy Linton: I think our dilemma here may be that we if want a version 3 text here, we have some time to do that, but it's not clear whether we can do that before we finish here. I just would want to make sure we get it right because I want the proposal to be quite clear what I'm asking people to reach consensus on.

Randy Bush: Dean, start editing.

Aftab Siddiqui: Andy, I need to clarify one thing. You are saying that even this proposal, there will be room for emergency cases, where anybody can propose and directly go to the process; right? Is that possible?

Andy Linton: I'm not saying that's what happens. I want you to tell me if that's what should happen. I want the community to make that decision, not me. Certainly, if we have to implement this and we have to run the process, we need clear guidelines from the community about this. So we would need to get it clear that this process is one that, if you like, doesn't completely replace -- it replaces what we have now, but it doesn't preclude --

Randy Bush: May I interrupt? Version 0 or 1 or however people count here did have an escape clause.

Andy Linton: Okay.

Randy Bush: So we can put one back, sure.

Andy Linton: I certainly want to be in the position that the community doesn't feel we are changing this and we can implement something here. I don't want us to build a process and then go, oh, wait a minute.

Aftab Siddiqui: I had a disagreement with that escape clause initially. I believe every problem statement should go through one process and that one process should be problem statement and solution and comment from APNIC Secretariat, then proposal. There shouldn't be any room for any escape clause or anything. If it is a do or die situation, which I mentioned some time back, there shouldn't be anything like that. It's not going to break anything. So just put a problem statement, discuss it. If it's a very urgent issue, everybody will jump in, of course.

Randy Bush: I think what Andy maybe has in mind is some sort of policy emergency. I have a hard time thinking of a policy emergency, but I'm sure someone can create one. I think what I'm afraid of here is not that we have an escape clause for policy emergencies but that we spend two months trying to define what is an emergency and what isn't.

Andy Linton: Yes.

Randy Bush: I don't know how we will weasel word around that -- pardon me for the American idiom -- how we will word this carefully to not get wrapped too tightly around the problem.

Aftab Siddiqui: Is there any possibility to get a quick consensus if people want to have an escape clause in this proposal or not? Is this possible?

Randy Bush: I don't think the Chair was listening.

Aftab Siddiqui: He was not. Sorry, Andy. I was asking, is it possible to get a quick consensus from the community sitting here, whether they are willing to have an escape clause in this proposal or not? Because I'm opposing it.

Randy Bush: He wants to find out if the feeling of the room is whether an escape clause is needed.

Andy Linton: Yes. I think our dilemma here is that we don't actually have a formal proposal. Our rules are that there is a proposal on the table that we have to say, this is the proposal. I'm going to suggest a compromise to us here, that if we're in agreement that this is what we should do, we may not be able to actually reach consensus on this right now because we don't have the formal process but there is nothing to stop us adopting this process over the next --

Randy Bush: He's not asking that. He's just asking how people feel about the escape clause.

Andy Linton: Yes, and I'm going to ask about the whole process of it and come to that. But I'm suggesting that we don't have a proposal, that this group here, the three Chairs, can actually sit down and ask you about consensus because we've got the proposal, we've got an amendment, we've got an amendment to the amendment and I can't ask this group right now to reach consensus on that.

Randy Bush: Nobody is asking you to.

Andy Linton: Not in any formal sense. I can ask for an opinion from the room.

Sunny Chendi: I have two comments.

Andy Linton: Just a minute, Dean. We have two comments from the Jabber room and they have been waiting for a moment. I don't think we can reach this as a consensus, but we can agree to agree about it and bring it back next time, using this process to do it and we can get it sorted out then. There's nothing that stops us taking Dean's bullet points and going with those, if the community is happy with that, we can experiment with that and then the next time somebody proposes something, it's been to the list, it's been discussed, we've had a problem statement and then we have a formal proposal, and this one can be the first example of it. I can ask the group to reach consensus, but it will on Randy's proposal as was submitted, which is version 2.

Randy Bush: That's not what he was asking.

Andy Linton: I know, but I'm going to ask that question afterwards.

Randy Bush: But if we had a feeling from the room on that point, that would affect the proposal that we would write.

Andy Linton: Yes. Our new process will be quite capable of capturing that, so I'm going to go for that feeling from the room after this. As it stands, the proposal that is on the table is what I am going to ask people about for consensus, which is Randy's proposal as written. Dean's amendment is fine but it's not what's on the table. I know that's being pedantic about process, but I'm going to be pedantic about it.

Randy Bush: I'm going to be pedantic. I withdraw the proposal.

Masato Yamanishi: We should interrupt this session at 3 o'clock and we only have two minutes. Sunny, can we get David Woodgate's comments from Jabber.

Sunny Chendi: Sure. David Woodgate has two comments on the Jabber chat. The first one is: I support the principle of Dean's proposal, perhaps with a caveat that the spirit remains informal rather than overly bureaucratic. Example, I believe I did something like this, discussing the issue of prop-101 on the list, before proceeding to create a proposal. However, I will note that it can be difficult to tease out sentiments from the SIG policy list, except for a small group of dedicated posters and sometimes the real sentiment of the community doesn't emerge until the SIG policy sessions at the meetings. The second one is: as Andy is saying, perhaps it's an issue of doing the best practice of behaviours, rather than specifying it as a very formal process. Thank you.

Andy Linton: We need to do the election result. We will do that and get back to this shortly. While Kanchana is coming up, Randy tells me he wants to withdraw this. I am very happy -- we have half an hour here and if I can see a revised text of this with the various bits and pieces in it and we have time, we will try to reach consensus on it. I'm very happy to do that, but unless we see that revised text, we will have to go with what's on the table.

Dean Pemberton I am going to leave it -- as the author, I'll leave the final decision about what Randy would like to do with the proposal up to Randy.

Masato Yamanishi: We should interrupt here. We can hear your comment after the result.

Kanchana Kanchanasut: It's 3 o'clock, so I'm happy to say that the results are here in the envelope. There is no dispute so far. I mean it should have been informed to me by 2 o'clock, so since there's nothing, then it's very smooth. I would like to invite one of the scrutineers to explain how the voting was counted. Leslie or George.

George Nyabuga: Good afternoon, everyone. I think we had a very good opportunity to scrutinize the counting exercise and I should say that I vouch for the process. There were both manual votes that were cast outside today and then there were online votes that were actually cast using the electronic facility. Both of the facilities that were available to the people I think were utilized to the maximum and efficiently, and I think it made the voting and electioneering very smooth, as she said. I think there shouldn't be any, maybe, basis for objection, unless, of course, maybe someone feels very strongly about the whole process. But I should say, as scrutineers, that we were very happy with the process. Thank you.

Kanchana Kanchanasut: Thank you very much, George, and of course, Leslie. So I will open the envelope and read the results. In 2012 NRO Number Council election result, total ballots counted 95, total valid ballots 95, so invalid ballots zero.

Candidates: Naresh Ajwani, 185; Ren-Hung Hwang, 63; Aftab Siddiqui, 57; Naing Win Oo, 2; Kashief Adiel, he's not here, zero. So the nominee who is successfully elected is Naresh Ajwani.

APPLAUSE

Andy Linton: A brief word, Naresh.

Kanchana Kanchanasut: Then I will hand over to the EC Chair.

(Documents handed to EC Chair)

Maemura Akinori: Thank you very much.

APPLAUSE

Andy Linton: Thank you for that, and congratulations to Naresh. Commiserations to the others. I'm on the NRO and you may be glad you didn't get the job. I understand from Randy that he has withdrawn this. Is that the case, Randy?

Randy Bush: The procedural discussion of this proposal that just occurred is a wonderful example of what this proposal was designed to prevent. I withdraw the proposal.

Andy Linton: Okay. That means that we can't sensibly discuss this any more because it's been withdrawn. But if this has sparked interest from people, then this is clearly something that the rest of the community can do something about if they want to. We will take our direction from there. I will only say that we have a procedure and I am required as the Chair, and the other Co-chairs are required to follow the procedure. If we want to change the procedure, the first thing we have to do is change the procedure using the current procedure. I have no option with that. I'm sorry, Randy.

Randy Bush: I have sat in your seat. We had no trouble asking people's opinions.

Andy Linton: I'm not going to go into an exchange and a whole bunch of stuff on this. We had an opportunity to ask people's opinions and I can no longer do that because the proposal has been withdrawn. So this will have to go back to the mailing list and all the discussion elsewhere. So I'm sorry. Thank you all for your attention. That's the last item we have on our agenda and I declare the Policy SIG session closed.

APPLAUSE

Naresh Ajwani: Chair, it seems inadvertently the only announcement in terms of total votes cast was for on-site. The total votes casted were not announced.

Sunny Chendi: You would like the number, the online plus the on-site, total number, for each candidate?

Naresh Ajwani: No, total votes cast, that's all.

Sunny Chendi: Do we have a figure on this? It's online plus on-site total for all the candidates.

Naresh Ajwani: Meanwhile, while they do the counting, I have a second request. As a convention and an opportunity, I want to thank the community, thank the community for their support, for their love and enhanced love, and I would like to mention that this kind of participation I think is growing now, and that is a classic example of when we really get engaged into such kind of processes. I want to thank especially my young friend Ajay from NIXI and many other young friends from Bangladesh, from Sri Lanka, from Indonesia, from Butan, and definitely my friend Skeeve and many other people. Thank you very much, I will try to do my best to come up to your expectations. Thank you.

Sunny Chendi: Thanks, Naresh.

APPLAUSE

The community is here and the community is watching what you said. I personally would like to congratulate you on your hat trick, I think it's a big milestone when you play cricket, it's not easy to have a hat trick in cricket. Congratulations and I hope the best for you.

Naresh Ajwani: Sunny, I must clarify, this hat trick is for NCO, because this is what I have been preaching, that people should not go for the fourth term and should let the youngsters come. I may contest other elections but I will not contest Numbers Council again. So hat trick stops here. Thank you very much.

Sunny Chendi: Thank you. You made a better decision than Sachin Tendulkar. Philip Smith isn't here. Would any other APOPS Chair like to start the lightning talks? We will have a small break, a 5-minute or 10-minute break and then we will start the lightning talks. Would else would like to Chair? I don't see Philip here anywhere. He will probably come at the right time.

Today's social dinner is supported by PHCOLO, our good friends, they have been supporting the Conference for a long time. When you say PHCOLO, Judith comes into your mind. She is not here this time but she sent her son who is sitting next to Blandine. David, Judith's son, is representing PHCOLO at the Conference. Thank you, David, for your support. Let's take a 10-minute break and I'll try to find Philip Smith so we can start the lightning talks. Thank you.