Caribbean Sector Meeting Draft Transcript - 17 September 2008

The transcript of the meeting mentioned below may contain errors due to errors in transcription or in formatting it for posting. Therefore, the material is presented only to assist you, and is not an authoritative representation of discussion at the meeting. An audio archive of the meeting proceedings will be posted shortly.

Table of Contents

Meeting Called to Order

MR. PLZAK: Good morning. Glad to see everyone made it. Hope you had a good time last night, you had plenty to eat and drink.

My name's Ray Plzak. I'm the President and CEO of ARIN. And welcome to an ARIN Caribbean Sector meeting.

First of all, let me say something about this meeting itself. This is an official ARIN policy meeting. What's going to be discussed here, the shows of hands that'll be taken here, will all be recorded, and that information will be provided to the Advisory Council for their use as they go through and determine the consensus of the community on the policies that we're going to discuss. Every policy we are discussing today will be discussed in about one month in Los Angeles at the ARIN meeting there.

There will be a written transcript of this meeting. It's being recorded. Therefore, I would ask you that when you wish to say something, and I would assume that you all will want to say something, please go to the mic and identify yourself. We plan to then do as we do for any other ARIN meeting, put the recording up on the website so that people in the future can go and listen to this as well as read the transcripts. There will be an Executive Summary of this meeting, and it will be posted to the PPML for the community to see what kind of discussions took place here.

So I know that the rest of the community is very interested in what you have to say about these policies. This is a very critical time in the development of the Internet as a whole, and of the Caribbean in particular, so we're not here to do trivial business.

Meeting Introductions

So having said that, let's take a quick look at who's here. From the ARIN Board, myself, and we have Bill Manning, who is hiding over there in the actually subdued shirt today. Yesterday he had a little bit brighter plumage on.

MR. MANNING: Elbow, elbow, wrist, wrist.

MR. PLZAK: Right.

MR. MANNING: Thank you, ARIN.

MR. PLZAK: From the Advisory Council, in back, we have Leo Bicknell, the chair, who's hiding.

If he could stand, please. And we have down here in front Stacy Hughes. There she is. Lea Roberts is sitting -- or standing right there. And finally, R.S., Rob Seastrom right there.

For those of you who are not aware of the role of the Advisory Council, it is the body that goes through and makes the consensus determinations.

They assign particular members to be the shepherds of policy proposals, they author policy proposals. They are the persons that you should engage for two reasons -- well, actually three. They're nice people to talk to is number one. Number two, they want to get your input. They want to know what you feel, what you think. And the third thing is if you have questions about the way the policy process works or you've got an idea for a policy proposal, they'd be more than willing to sit down and help you.

They'll actually help you draft it, show you the things that needs to be done. So they're here to help you. And like everyone else, they're volunteers, and they are putting in a little extra effort into the system, so please take advantage of them from that regard.

From the Number Council of the NRO, the Number Resource Organization, we have Martin Hannigan, who was here all day yesterday and I don't know where he went to. The Number Council functions as the advisory -- as the Address Council of the Address Supporting Organization of ICANN. I'm not going to go into the organizational structure of ICANN. I would prefer to let you read the factsheets that we have on it and go from there.

So from the staff, you've got me. We have Cathy Handley and Richard and Mark Kosters back there. You saw him yesterday up here. Leslie and Einar.

Out front, the meeting planner, De Harvey. And sitting next to her yesterday was Hollis KARA. Sitting next to her today is Megan Kruse. You saw the two Matts: Matt Rowley and Matt RYANCZAK. There's RYANCZAK there and Rowley right in front of him. So you can figure out which one you want to call Matt and which one you want to call something else. We do that all the time at the office.

We've got three Cathys, so we can -- we've got two Matts and three Cathys we can play with in that regard, too. Actually, we've got two Rays. There's another guy named Ray besides me, but that's all right.

From the CTU, I'd like to welcome Bernadette Lewis -- of course, you saw her yesterday afternoon -- and Nigel Cassimire.

And so, just a general reminder of some of the rules. Please be courteous and take all your cell phones and silence them or at least put them on stun. Same thing with your pagers. If your computer's got those little wonderful things that say "You've got mail," or "Hello," or play charming music and everything else or beeps when you get something, please turn that off.

Also, if you've got your Gameboy or other portable game device with you, turn off the sound effects. We don't need to see those cruiser blasters going off. Hit a nerve, didn't I, R.S.?

Anyway, we have a lot on the agenda today. We've got seven policy proposals to be discussed.

And of those seven, every one of them has to deal with the current situation in the Internet as a whole, and two of them, as I noted yesterday, are of specific interest to the Caribbean, and that's the 2008-4, the minimum allocation for the Caribbean region, and 2008-3, community networks IPv6 allocation. But the rest of them have an equal amount of weight in the Caribbean, it's just that these two are really aimed towards situations in the Caribbean.

So we will additionally have some of the regular reports we have at a meeting. In addition, we will have a couple of things that normally would happen at a meeting -- at the big meeting on the day before the start of the policy meeting, and that's the introduction to the policy resource evaluation process, and descriptions of existing numbers of resources and so forth.So we have a couple things we're going to discuss, exchange points, Internet activities in the Caribbean. So without further ado, before I do that, I want to point your attention to a couple of things.

First of all, all the presentations will be online, and you were given one of these this morning. So if you want to download it to a convenient place, you can put it right here and you'll have it, a little flash drive.

And you also have a folder. And in the folder, you'll find the agenda, and you will find a whole bunch of other goodies. You'll find a couple things about ARIN, about the community participation. We have upcoming elections. You're a member, vote, and it'll tell you how to do that. You'll see some things from the Board advising about the transition to IPv6.

And you'll see a whole series of these things, one-page factsheets, about a whole bunch of things. You've got a question? It's probably already answered there. If it's not answered on one of these things, please get to a member of the staff or the AC and, if necessary, we'll produce a sheet to do that.

Oh, there's Hollis. She was sitting down there. I didn't notice that before. Sorry.

MS. KARA: That's okay.

MR. PLZAK: I thought you were out doing something else.

And these are also available online. And in addition to being available online in English, they're also available in French and also in Spanish. So again, you can use that flash drive and download these things. They're in PDF file format and you'll have them there, too.

Then on the other side, you'll see a description of this meeting. You'll see something about some things that were used yesterday for the wireless. Then you'll see a discussion guide. And behind it, you'll see a couple of other documents I want to call your attention to.

The discussion guide is a text of every single policy proposal we're going to discuss today, and it has the information about it. And you'll see the full rationale from the author of the proposal.

And so if you've got some questions, you might be able to answer them just by reading there.

And then you will find, also, a copy of the Internet policy evaluation process in there. This gives you sort of here's what we're doing kind of a tutorial that you can read. That second document is available online.

One of the policy proposals, the transfer policy proposal, the Advisory Council conducted a survey to gain some more specific information from the community to assist them as they moved this policy through the process. And the results of that survey are provided for you on this sheet. You can see each question that was asked and what the answers were.

And then there's lastly the ARIN Number Resource Policy Manual itself. And in this, you will find every policy that is currently in force in the ARIN region. So if you have questions about something or if you see that a proposal's going to modify certain paragraphs of the manual, you can go right there and take a look at what it is.

So now I can go on without further ado, because I just did the ado. And that is to ask Leslie Nobile to come up here and provide us with the Internet Number Status Report.

So I'm going to drive this from up here?

MR. JIMMERSON: Yes.

MR. PLZAK: Where are we? Okay. I have to now decipher the -- ah, here we go. You're joint stats, right?

Internet Number Resource Status Report

MS. NOBILE: Okay. Good morning, everyone. This is the Internet Number Resource Status Report. Basically, it is a look at number resources in all five the regions. It's IPv4, IPv6, and Autonomous System numbers, and it's in a side-by-side format so you can compare the growth trends over the years in each of the regions. We update this quarterly, so the data's fresh as of June 30th.

So the first slide looks at the total IPv4 address space and the status of the 256 /8s that's contained within the v4 pool. If you look at the slide on the left, the big pie, we'll start with the central registry. There's 91 /8s that were issued to what we term the central registry. That was all of the space issued prior to the establishment of the RIRs.

This was generally done under U.S. Government contract and that is where most of the Legacy Class A and B space resides.

If you look to the right, Not Available, there's 35 /8s that are reserved by the IANA for special purpose. I've broken it up further to the right. You can see multicast, private use, loopback, local ID, and experimental space. So those cannot be issued by IANA to the RIRs for customers.

If you look down into the gray slice of the pie, that's probably the most interesting part of the slide for some people. That is what is left in the IANA reserves. There's currently 39 /8s left in the IANA reserves. That is the end of the IPv4 pool.

Moving left to the RIR space in blue, the IANA has issued 91 /8s to the five RIRs collectively. It's broken up farther below, just down below. You can see the totals that each one of the RIRs has from IANA. And you can see that the three larger RIRs -- ARIN, APNIC, and RIPE NCC -- have pretty similar numbers, all right around 26, 28, 29 /8s from the IANA.

So this is the first slide that looks at the total IPv4 address space issued by each of the RIRs to their customers. We've color-coded them. Just kind of -- you know, some growth up and down, nothing too interesting until you start to get to more recent years. In the past year or two, the growth trends have changed drastically.

We're seeing much of the IPv4 space being issued in the RIPE NCC region, and in the APNIC region in particular -- last year, APNIC issued over four /8s to their community. That is the most -- the largest number of /8s ever issued by any single RIR in a year. And it looks like they're on track to do the same thing this year.

This is only a half a year's data in the 2008 bars, so we'll see what happens, but that is where the growth is. It's not -- it's no longer here in the U.S., although we are increasing. We're giving more and more address space in recent years. The real growth is in the Asia-Pacific region and Europe and Middle East and Russia, ex-Soviet Union.

So this is just another way of looking at the slide. It's the cumulative total that we've all issued, each of us has issued, since 1999. We've only gone back as far as 1999 because basically that's all we could fit on the slide. At some point, we'd like to put together slides that look at the '80s into the '90s. And we probably will be doing that soon, but for right now, we go back as far as 1999.

And you can see that RIPE NCC and APNIC have both issued more IPv4 space to their customers over the last 9-1/2 years than ARIN has, so that the trend has changed.

Autonomous system numbers. ARIN has traditionally issued more ASNs than any of the other RIRs. That has changed in recent years. The RIPE NCC region, again, we're seeing quite a bit of growth there. They're issuing more ASNs than any of the five RIRs at this point in time.

And here's the cumulative look at that. ARIN, as I said, traditionally has issued more ASNs to their customers, so we still have the largest chunk of the pie there. But within a year or two, it looks like RIPE will have issued more ASNs than any of the five RIRs.

So IPv6 address space, this is a kind of complicated slide, so I've got to walk you through it. I'm going to start at the top in the gray, that little gray pie that says /0. That is the total IPv6 address pool. It's /0. The IETF has reserved a /3 of that /0 to be used for global unicast space.

That consists of 506 /12s. So the IANA holds that space right now in reserve, and they issue from that to the RIRs.In October of 2006, there was a global policy passed. That's policy that's passed in all five of the regions. And it said that IANA would issue a single /12 to each one of the RIRs. So up in the right corner there, you can see that each of us have a /12 to use for our customers.

Prior to October of 2006, we were getting bits and pieces from the IANA. Generally, the size was a /23. So that square sort of -- that pie below is another single /12 that was being used prior to October. You can see that RIPE got most of the /23s and -- you know, followed by APNIC, AFRINIC, ARIN. There was also a special purpose, a couple reserved for RFC space. So that was all done from one single /12.

This looks at the growth. It's like the v4 slide. It's IPv6 allocations to LIRs or ISPs. Local Internet Registry is the term used in some of the regions. Here in ARIN, we use ISP. So this looks at the growth trends.

Again, this is kind of sporadic. We had some steady growth early on, and then it dropped off in many of the regions, went back up, it went down.

No real trend, except again in the RIPE NCC region, in the green. That is the growth. That is where we're seeing v6 being deployed and v6 being utilized and issued. So who knows what will happen, but --

MR. HUGHES: Are those all /32s?

MS. NOBILE: Those are not all /32s. I have further slides I'll show you. This is just total number of allocations. It doesn't tell you the size. So they're making more allocations than any of the other RIRs.

But this tells you the sizes. So if you're looking at the left, cumulative total, RIPE has made almost 1,000 allocations: AFRINIC, 51; LACNIC, 143; APNIC, 384; and ARIN 433. We're actually the second at this point, so we're seeing some growth in our region. But in terms of /32, /32 is the minimum allocation size in IPv6.

So when you look over to the right, you can see that RIPE NCC has issued over 33,000 /32s. What that tells us is they're making very large IPv6 allocations. We've seen many /19s, 20s, 21s, 22s in the RIPE region. And also, looking in the yellow, in APNIC, they are also making very large allocations, so that accounts for it. Those are just single /32s given out -- you know, as large -- one single, large allocation.

The interesting thing is, ARIN is now at over 14,000. Prior to that, we were somewhere around -- you know, 300-and-some-odd. But recently, we made a very large allocation to one of our customers in the U.S., and that accounts for the increase in total numbers of /32s. And LACNIC, 262, they're growing quite rapidly; and AFRINIC, 46. And those are, again, single /32s generally in those regions.This is just the links to the RIR stats. As Ray said, the presentation will be up online if you want to look at statistics. All of the RIRs keep stats on their web pages, so we all have them on our page. The NRO.net has the stats. And then IANA, the IANA page has the really interesting sort of historical data, traces it all the way back to the beginning in the early '80s when it was issued, so that's a good page to look at.

And that's all that I have. Are there any questions?

MR. MANNING: It's on now. So back up a couple of slides, please.

MR. PLZAK: Again?

MR. MANNING: Oh, I'm still Bill Manning.

MR. PLZAK: Yes, we know.

MR. MANNING: Okay.

MS. NOBILE: Where are we going?

MR. MANNING: Keep going back, back.

MS. NOBILE: We going to before?

MR. MANNING: Stop right there.

MS. NOBILE: Okay.

MR. MANNING: What's that "Reserved for RIPE NCC" and "Special Purpose?"

MS. NOBILE: What I believe that was was RIPE had a large allocation to a customer and they asked the IANA -- this was before the /12 days, Bill, before we all had /12s. IANA had a large allocation to one particular customer, and they knew that the customer was going to be coming back for additional space, and they asked IANA to hold a reserve.

And you know, I'm going to throw this out, I'm not sure it's positive. It may have been they gave a 20, they asked for the second 20 to be reserved. But I believe that's what that was for and they held it, and I don't believe it's ever changed and I don't believe it's been issued either, so it's still sitting there.

MR. MANNING: Get it back.

MS. NOBILE: Yes.

MR. PLZAK: I'd like to point out on that slide that that /3 represents one-eighth of all the address space. And then that 512 /12s is about, what, two-tenths of a percent of the amount of that /3. So we're talking about a lot of space here on the surface of it. But then when you start doing some very quick math, like you're only allocating out of 64 bits, because half of the bits is reserved or set aside from the standpoint that it's a boundary, it's a technical boundary, so the policy boundary can only go up to 64 bits. So in the end, it's brought us a large amount of address space, but it's still not as large as a lot of people would think it is.

MR. HUGHES: Hi. I'm Aaron Hughes. I'm really quite astonished at the number of v6 allocations. And I'm wondering if anybody is doing any tracking of dates between the time they're allocated and the time they're announced in the table, given that there are a whopping 1310 prefixes in --

MS. NOBILE: You know, some people are, but ARIN has not been.

MR. PLZAK: Yes. We -- ARIN is not in the business of doing that. We haven't really been told to do that by our community. It's something we could do, but there are other people that do that, Geoff Huston being one, where people that do routing table analysis, and there are regular reports that are published all the time that are available, so the information is available. And of course, from the standpoint of routing, as you well know, depending upon where you are and what you're looking at, you may or may not see it all as well.

MS. NOBILE: And I can add to that. Gert Doering also does that. I think you might know that he tracks that. And I do that that in the LACNIC region, they have a policy that if they issue a /32, they actually go back and check routing tables. If it is not routed within -- I believe it's a one year period -- I believe; I could be wrong -- they will pull back the allocation, because they expect you to be using it. So I thought that's kind of interesting. So they're definitely paying attention.

Okay. Thank you. (Applause)

MR. PLZAK: Thanks, Leslie. And I believe now you're going to be followed by Einar. So actually, I'll let him drive this.

Introduction to ARIN's Internet Resource Policy Evaluation Process (IRPEP)

MR. BOHLIN: There we go. My name's Einar Bohlin. I'm a policy analyst at ARIN.

I work in the Member Services Department, and this is about the ARIN Internet resource policy evaluation process, otherwise known as the policy development process, or "the process" for short.

The policy process is open, transparent, and bottom up. It's open in that anyone can participate. All you need is an e-mail address or come to a meeting such as this one, or the one, for example, we're going to have in Los Angeles next month. The process is transparent. The process itself is documented on the website, and it's also, as you heard, it's in your packets today, so it's available. In addition to the process being available, the policies themselves are available on the website and published and put into your handouts.

Procedures such as what the AC does and the Board. The AC is the ARIN Advisory Council, and the Board. They both have roles in the process. Their meetings are minuted, and those minutes are published on the website.

And the process itself is bottom up. It's the community that proposes proposals, that discusses them, and ultimately approves whether or not a proposal becomes policy.

So at a high level, proposals come about because someone perceives that there's a problem, and that's what the need is. They feel that they need to fix something, either by creating a new policy, removing something that already exists or modifying something that exists. So they suggest a proposal. It's posted online. It's discussed online and at meetings. And a consensus evaluation takes place.

If the consensus is to approve the policy -- or the proposal to become a policy, ARIN staff implements the policy. And after it's implemented, it's evaluated for perhaps unintended consequences, or to make sure that it does what the author intended, the community wanted it to do, et cetera. And if that's not the case, if something else is needed, then it generates more need and turns into another proposal.

So it's a pretty simple process. However, there are specific requirements or milestones. That's what this slide is. A proposal has to be submitted at least 60 days before a meeting in order for it to be discussed in that meeting cycle. It can be submitted anytime by anyone. And if you've missed a deadline, which we publish on the website and post to PPML, which is the Public Policy Mailing List, if you missed the deadline, it's okay. It's just discussed further and at the next meeting.

The first thing that happens is the proposal's forwarded to the AC. It's also put on the website. And the AC makes an initial review. They get to decide early whether it's a good idea or a bad idea, whether it should be formally accepted into the process for discussion.

And this is a petition point. I'll talk more about these later. There's going to be a couple petition points. I'll get back to those.

So if a proposal's accepted for formal discussion it's numbered and posted. And there's a requirement that it be discussed at least 30 days online before a meeting.

Then there are meetings that take place. It's important to note that when the AC looks to do their consensus evaluation, that they consider the online discussion and the meeting discussion to be at least equally weighted.

After the public policy meeting, the ARIN Advisory Council meets to look at those discussions, online and the meeting, and they decide what to do with the proposal. They can accept it or ask that it be revised or they can abandon it.

If they move it forward to be accepted or adopted, it goes to Last Call for at least 10 days back on the list. And then the AC will meet again, look at those Last Call comments, and decide ultimately what to do with the proposal.

If they recommend that the proposal be adopted, they send it to the Board of Trustees. The Board of Trustees has -- well, inherently they have a fiduciary responsibility to make sure that what they do conforms with law, and they're concerned about liability. But specifically within the policy process, they're tasked with making sure that the proposal has followed all of these process steps.

And if everything is good to go, they adopt it and ARIN staff implements the policy. And in practice, what we do is when we find out from the Board that something's been adopted, we publish or we post to the list and say it's been adopted, and we will implement this proposal no later than a date that we establish out into the future, normally approximately 60 days.

And as part of the implementation, we take the policy, we put it in the policy manual, and we publish a new version of the policy manual.

And the last step is community and staff evaluation of the results. Did the policy do what the author and the community intended it to do, et cetera? If it didn't, it generates perhaps a new need.

Okay, the history of the process itself. Before 2001, it wasn't such a formal process. In 2001, the Board decided to actually document and write out exactly what the process was, so folks could know exactly what they needed to do to participate. And we started numbering proposals so we could track them.

And the first proposal, 2001-1, was ultimately abandoned. It came out shortly after that first documented version of the process. And the second one was adopted. And I'll show -- this is kind of interesting in the next slide, it starts a trend.

In January 2004, the process was updated with the part about the AC doing an initial review.

Prior to this, we'd had several meetings where a lot of proposals came about the same topic and all of them were put on the agenda. And this allows the AC to whittle it down a little bit, maybe instead of having four or five proposals about the same topic, there can be one or two. It turns out to work much better. And we published a template and instructions how to submit the template.

Another update, this was a process update before this. The AC had to meet at a certain time point after receiving a proposal. This just allows them to batch the proposals and do them at their regularly scheduled meetings.

All right. Total number of proposals that have been numbered and discussed is 90. Forty-three of them have been adopted, 40 abandoned, and we have 7 today on the agenda. Okay, petitions. I told you about how there were petition points on that milestone slide.

If the AC doesn't move something forward, the author has the ability to initiate a petition and ask the community to help move around the AC to keep their proposal going forward. And it's been used three times: Twice in the early, once in the later stage. In the early stage, if the AC -- if you send a proposal in and the AC says we don't accept this as a formal proposal, it's a really low threshold to move it around the AC. It's just support from four people from four organizations. And as you go through the process, the threshold goes up.

A policy manual. We have one single document that contains all of ARIN's Internet number policy. We're on the 11th version. It came out August 5th. The first version was back in 2004. And prior to the policy manual, we had policy sections on the website in different places. So we brought it all together and made one nice document.

It has a change log, so you can see what happens from one version to the next. The reason we publish a new manual is because of implementation of policy.

And I have some references, the process itself, the proposal archive. All 90 of those proposals has its own page. You can see the status, what happened to it, what meetings it was discussed at, et cetera.

And a site where every proposal template has come in is also on the website, and the policy manual itself.

Thank you. (Applause)

MS. HUGHES: Before you go, lest we forget, I would like to point out -- Stacy Hughes, TW Telecom, ARIN Advisory Council -- that it was you, Einar, that assembled the NRPM in the first place. And I would applaud you for that. (Applause)

MR. BOHLIN: Thank you.

MR. PLZAK: While Einar's setting it up for the next presentation, a couple things. We're going to have another drawing this morning at the break. And instead of trying to come up with a birthday month and so forth, if you've got a business card, there's a little glass jar out -- it looks like a tip jar in front of where De is sitting. Just drop your -- or Megan's, between the two of them sitting. Just drop it in there. If you don't have a business card, just take a piece of paper and write your name on it and drop it in there.

And we'll be doing a Plinko this morning, and we'll be doing another one this afternoon. We've got a couple more prizes for you there.

One thing I want to point about the policy proposal archive, Einar said that you could see the history of the proposal and it's got links to meetings. It actually takes you to the minutes of the meeting. You can actually go to the webcast of the meeting. You can actually see what the discussion was. It's a very, very extensive archive. And so -- and it's cross-referenced, so you can see what happened.

And the last thing, as you're aware, we're in the process of modifying the policy development process to implement it. There's some housekeeping things that we have to take care of to implement it.

One of the things that's going on is -- you know, as we move to have regularly scheduled meetings in the Caribbean, we need to work that schedule of meetings into the policy process as well. So there'll be some tweaks to that policy process.

The role of the Advisory Council is changing, and so there's things we have to do to make sure that that role is well understood. So look for it to be adopted around the first of the year.

And with that, I'd like to ask Leslie to come up and provide a description of the existing policies.

So Leslie?

Description of Existing ARIN Number Resource Policy

MS. NOBILE: Okay. Good morning. I actually won't describe all the policies because that would take the next three days, but I will go over the more relevant ones for you all.

We're going to talk about IPv4 policies, IPv6 policies, and ASN policies, and basically how to get space from ARIN, how to qualify for these resources.

Before I start, can I ask you guys, how many of you have direct IPv4 address space from ARIN, have dealt with ARIN? Okay, not many of you from the Caribbean. Okay, interesting. So this -- actually, I'll spend some time on this so that you can understand how you can qualify for space.

First, there's some terms to go over. We use these frequently within the policies and they're very relevant.

To allocate. So allocation means to issue number resources to an ISP. For internal -- they're internal networks, and then for further subdelegations to customers.

Assigning is we assign number resources to end-users for internal networks only. So there'd be no further reassignment of that space. It stays right within -- you know, your own organization. And that's small businesses, corporations, universities, things like that, typically would be an end-user.

And then multi-homed. It's described in our policy manual as connectivity to more than one ISP, or peer, and the same prefix being announced by at least two of them. So there is a specific multi-homed policy. That's why I've gone over this term.

We're going to start with IPv4 policies to ISPs, and this would be initial allocation policies.

There's two main policies, two main ways to qualify for address space from ARIN. There's a standard policy. It says that you have to have a total of a /20 of IPv4 space reassigned to you from an upstream and it has to be efficiently utilized. So it doesn't have to be a full /20. It could be a discontiguous space. It could be a number of /23s or 24s. The minimum allocation size under this policy is a /20, so you basically have to have a 20 worth of space that you're using currently to get a 20 from ARIN. And we look at a three-month need, so we look at your three-month projections when we're issuing this space.

There's a multi-homed policy. It's a little more complex. You have to have a /23 worth of space in total reassigned from an upstream. It has to be efficiently utilized. You have to be multi-homed, or have an intention to multi-home within the next 30 days. And we generally ask for signed customer contracts or peering agreements to ensure that you will be multi-homed. You have to agree to renumber out of your upstream space before coming back to ARIN the next time. So the space you get from ARIN, you not only have to use for growth, you have to use it for renumbering. The minimum allocation size under this policy is a /22. And again, it's based on that three-month need.

So additional allocations to ISPs is one basic policy. You have to have efficiently utilized your most recent allocation -- 80 percent of your most recent allocation from ARIN and 100 percent of all previous allocations. That's what the policy spells out. If you qualified under the multi-homed policy, you would have to have returned the space to your upstream as agreed to in your initial allocation. If you didn't qualify under multi-home, that doesn't come into play at all.

You have to show your reassignment information in ARIN's WHOIS database or in your own RWHOIS database. I list SWIP, RWHOIS, or a spreadsheet. SWIP is just a reassignment template.

You have to use that to show your reassignments to your customers and then they'll show up in ARIN's WHOIS database; or you can set up your own RWHOIS server and put your customer reassignments in there and ARIN comes to check those out and verify; or you can send in a spreadsheet for the /29 and smaller allocations.

This is based on a three-month need for most new customers. However, if you've been a subscriber member of ARIN, if you've gotten space from ARIN for over a year, then we will look at a 12-month need and issue you enough space for your 12-month growth. So that's somewhat established ISPs who've come to ARIN previously.

This is the initial assignment policy for end-users. There's the standard policy criteria. You have to show 25 percent utilization of what you're asking for right now. So if you're asking for a /20, you have to show that you're going to be able to use a /22 right now and that you plan on using a /21 within one year.

So it's actually a simpler policy to qualify under, but you have to be an end-user. It's basically based on a 12-month need. I don't really like to say that, but -- I mean, we are looking at the fact that you're going to be using 50 percent of what you're asking for within 1 year, so, to me, that's a 12-month need. And the minimum allocation under the standard policy is a /20.

So if you're multi-homed, obviously you have to be multi-homed or intend to multi-home, it's the same criteria: Show the 25 percent utilization now and a plan for 50 percent within one year. We're looking at the one-year need basically. And the minimum allocation size is the same as the multi-homed policy for ISPs, which is a /22.

So there is now an additional assignment policy for end-users. There used to not be. I think the assumption was end-users come once and then that's the end of it. But what is -- the fact is that end-users do come back for space as they grow their organization. So there is a policy now.

It says you have to demonstrate 80 percent efficient utilization of all previous assignments. And again, it's the same criteria: It's the 25 percent right now, immediate, and 50 percent within one year of what you're requesting.

So we have this other policy that not many people know about, but I think it's important to talk about. Because when you really have no other options, we have a policy called Immediate Need. Not many people use it because it is somewhat difficult. You know, you've got to provide a lot of information. But there is an alternate way of qualifying for address space from ARIN if you don't have upstream space.

So the policy basically says provide justification to show address space will be used within 30 days. For us, an immediate need is about a 30-day window. So we don't see many startups these days; we used to 10 years ago, but not so much anymore. What we mostly see is companies who are establishing new products or services, so they have an immediate need for this and they're going to set up a separate account.

And we ask for signed customer contracts, network topology plan, network deployment schedule, equipment receipts, and connectivity contracts. We usually ask for five out of those six, sometimes all six of them, depending.

The policy says a /16 is the largest block that can be issued under this policy.Another one that may be relevant to this region, because I know there's a lot of ccTLD providers down here, is the micro-allocation policy.

Not many people qualify under that, because it's a very narrow scope of people who actually can qualify. But it says you have to be a critical Internet infrastructure provider. And that is defined in the policy as public exchange points, core DNS service providers. That would basically mean -- is that me popping, sorry -- ICANN-sanctioned route servers, gTLDs, and ccTLD operators. The RIRs fall under that and ICANN falls under that policy as well.

The exchange point operators have some further justification to provide and that is: The connection policy, their location, and the ASN and contact information for at least two other participants. The allocation size is no larger than a /24 in v4, /48 in v6. This applies for v4 and v6, the micro-allocation policy.

IPv6 policy. So this is the initial allocation to ISPs. It's just this set of criteria.

This is the one policy. You have to be an ISP or an LIR. As I said, that's the term they use in other regions. You can't be an end site. You have to plan to provide v6 connectivity to organizations and route the aggregate prefix. You have to plan to make at least 200 assignments to customers within five years, so we're looking at a five-year time frame here, or be a well-known ISP in the ARIN region. That isn't really spelled out, what is a well-known ISP, so we spelled it out.

And we said you have to be a 12-month -- at least 12-month subscriber member in v4's address space to be considered a well-known ISP. You've had to come to ARIN for at least one year and have received address space, v4 address space.The minimum allocation is a 32. Larger allocations can be justified. And generally there's a lot of information requested, but it's based on the number of existing v4 users in the organization structure and the organization's infrastructure. So we're looking at the v4 space and how many customers you have currently in v4 if you want larger than a 32.

There's a subsequent allocation policy for ISPs. And it uses something called the HD Ratio. The HD Ratio is defined in RFC 3194. I cannot describe it to you. It's a bit complex, but it's basically -- it says that you have to have efficiently utilized your space and that we're going to look in terms and measure it in terms of /56 assignments, how many -- the number of /56 assignments you've made to your customers. So if you want more address space from ARIN, you come in and we look at this HD Ratio chart.And I have an example here. So a /32 has over 16 million /56s. If you look at this HD Ratio at .94, a little over 6 million /56s would have to be assigned in order to request more IP address space from ARIN.

This is -- the subsequent allocation policy is based on a two-year need, so we're going to look at just your two-year need. Whereas the initial was five years, the subsequent will be two.

There's an IPv6 assignment policy for end-users in the ARIN region. A couple of the other regions have the policy as well adopted, and I think it's being discussed in all of the regions at this point. The criteria says you cannot be an IPv6 LIR or ISP.

And you have to qualify under some other ARIN policy for an IPv4 assignment or allocation, some IPv4 policy currently in effect. So if you can get -- if you can qualify for a v4 from ARIN in some way, then you can get a /48 in v6. And as I said, initial assignment size is 48. You can qualify for more with further justification.

ASN policy, really simple: You have to be multi-homed or intend to multi-home or have a unique routing policy. That is it.

If you need assistance, the Registration Services staff, that would be my department, are the people that you deal with to get your resources. We have an online sort of Help Desk, hostmaster@arin.net. You can send any questions, any templates, anything you need to "hostmaster." You can call our Help Desk. It's open from 7:00 a.m. To 7:00 p.m., so they're always ready to help you.

Einar mentioned the policy manual. We call it "NRPM" for short; a really odd acronym. But there's the link that contains all of the policies that are currently in effect at ARIN.My staff put together something called a Quick Guide to Requesting Resources. I like to call it the cheat sheet, but it is a really comprehensive document that tells you exactly what policies are what and how you can qualify, and it's just very simple. I think we may even have a copy in the handout package, which would be great, but anyway, there is the link if you need to see it.

And that's all I have. Are there any questions? No? Okay. Thank you. (Applause)

MR. SEASTROM: Is this on now? Okay, good. Okay. Rob Seastrom. I'm on the ARIN AC.

And one thing, if Leslie can flip back to the immediate need policy, I think that this is sort of one of these things that's really underutilized and might be useful to a lot of people in this region. I've used it myself and been quite successful with it. The trick to using the immediate need policy successfully is to actually have all of your documentation on hand and be able to provide it to ARIN.

Now, none of this stuff here is stuff that you would not have if you were actually doing what -- you know, what you're asking for space for. In the case of the first immediate need request I put in, we were -- actually it was the second immediate need request I put in, we were insourcing, meaning we were bringing in-house after having a contractor running it, a DSL contract with about 3,500 customers on it.

And in that case, we bought equipment. We had signed transit contracts and everything. I had a year's worth of bills from the company that we were outsourcing it all to that showed that we had 3,000-plus customers. I scanned this all in and submitted it as documentation to ARIN. It was about 85 or 90 pages long when all was said and done. And the next communication I got from ARIN was it's time to set up your account so that we can bill you properly for this; you're approved.

So just to make it clear, this is not a really difficult thing to do. It's just something that you need to bring a lot of documentation to the table on.

MS. NOBILE: Thanks, Rob.

MR. PLZAK: Thanks, Rob. Yes, Leslie was correct, inside your packet you will find the Quick Guide to Requesting Resources. So her cheat sheet is there available for use besides being available online.

2008-5: Dedicated IPv4 block to facilitate IPv6 Deployment

MR. PLZAK: So how we're going to begin the first of our policy proposal discussions. And you guys are going to be entertained all day long by me messing with this machine.

And this is Policy Proposal 2008-5: A dedicated IPv4 block to facilitate IPv6 deployment.

This was designated a formal proposal in June, and the current version is the version that was the initial version. It's under discussion. It hasn't been discussed at any meeting other than it's only been discussed on the list.

If you look at the other RIRs, you'll see that neither RIPE NCC or AFRINIC have a similar type policy. But you'll see that in APNIC such a policy has gone through and it's at the Last Call. And you will see that LACNIC has adopted a similar policy proposal to this one.

So our understanding of the proposal is that it designates a /10 of IPv4 address space from the last /8 that we received from the IANA to be used to facilitate the immediate IPv6 deployment, things such as dual stack servers, NAT-PT, NAT646, and so forth. The prefix size would be a /28 through a /24. And the allocations and assignments can be one prefix every six months.

The shepherds are Owen DeLong and Matt Pounsett, neither of whom are here.

The balance of the proposal will be presented by Stacy Hughes. So Stacy, if you would like to come up.

MS. HUGHES: Good morning, everyone. My name is Stacy Hughes. I'm on the ARIN Advisory Council and I work for TW Telecom.

This policy proposal is to facilitate IPv6 deployment by using IPv4 resources, because there are some technical exigencies that require IPv4 to implement IPv6.

Boy, that's an eye chart, isn't it?

MR. PLZAK: It's in the packet.

MS. HUGHES: Yes, it is in your packet. Ray described the policy. Does anybody have any comment on it?

Hi, Bill.

MR. MANNING: This is Bill Manning. When do we expect ARIN to receive its last /8?

MS. HUGHES: According to whom? (Laughter)

MS. HUGHES: There -- doomsday is a highly debated date.

MR. MANNING: I'm not sure I would call it "doomsday." I would call it the last /8. Because whatever we decide -- I mean, we can decide this now, and there's some thinking that this is an impending thing, but it might actually be a couple of years out. And if we do this now, we might forget, right, or the rationale behind it might change. So is this actually something that we expect to see the last /8 appear in the next 24 months?

MR. PLZAK: Stacy, can I?

MS. HUGHES: Yes.

MR. PLZAK: The proposal -- of course there's a global proposal that's wended its way -- it's wending its way through the process right now, which would say that when IANA gets down to its last five /8s, it'll give them all out at once, one to each RIR.

MR. MANNING: I've heard that.

MR. PLZAK: Okay. Which would probably put that to occur sometime late 2009, maybe 2010, somewhere in there. I'm not sure when. More likely towards the end of 2009, maybe in 2010.

But I would have another question to pose. It'd be which /8 should we use? I'm not certain that the last /8 that we receive from IANA would be a pristine /8, because they have been doing some recovery work. So one of those /8s may be a little bit more tainted than others. They've handed out tainted /8s in the past, particularly to APNIC.

So I would think that another discussion point would be which /8 should we use?

MR. MANNING: That's not part of the policy proposal as --

MR. PLZAK: I know it's not, but it's a consideration.

MR. MANNING: As stated. So I guess then to answer my question, within the next 18 months, there is a perfectly reasonable expectation that this policy would go active?

MS. HUGHES: Should the policy go through the process, yes.

MR. PLZAK: Leslie?

MS. NOBILE: Just quickly. There's 39 /8s left in the pool. The RIRs collectively, all five of us, have been -- this past year used up about 12 /8s. It looks like growth is increasing in some of the other regions. So if you anticipate more than 12 /8s to be issued over the next year or two, that sort of does the math. Right, Bill, 39, if you're using 12 to 13 /8s per year? That's a couple of years out, right? Two? Three?

MR. MANNING: Yes.

MS. HUGHES: Rob?

MR. SEASTROM: Yes. I was just going to point out that the -- we have seen this effect in the past, and the 2011 is probably a reasonable inner bound on when it's going to happen. But if events of the last few days in New York are any indication, we may be in for one of those flattening outs again, like we saw after 2001. So it could be five years out instead of three years out. We don't know. It makes sense to have this sort of thing figured out sooner rather than later.

Now, with regard to the tainted /8 question, I'm not sure, but I don't think that there's going to be any /8 that's sufficiently tainted that we can't get a contiguous /10 out of it. Do you know any reasonable scenario that would be appreciably worse than the 223 /8 that APNIC threw back?

MR. HUGHES: Can I make just one comment on that point?

MR. SEASTROM: Yes.

MR. HUGHES: I think the intent of the policy is that -- to have space available for things like the 832 bits for --

MR. SEASTROM: For router IDs and things like that.

MR. HUGHES: Who cares if it's dirty?MS. HUGHES: Right.

MR. HUGHES: Because it's not really going to be used as publicly (inaudible) requirements for internal use.

MR. SEASTROM: Well, who cares if it's contiguous /10 either?

MR. HUGHES: Right, exactly.

MR. SEASTROM: So maybe what we need to do is push a revision to this and say the equivalent of or a sufficiently clean whatever.

MR. PLZAK: To answer your question, I don't think that there's any existing /8 in the IANA reserves that you couldn't find a continuous /10. I think that this wouldn't be a problem.

MR. SEASTROM: Okay. I don't think there's any particular reason that the verbiage about it being contiguous /10 has to be in there. I mean, it could be --

SPEAKER: It's for masking.

MR. SEASTROM: Yes, yes, but I understand that it's for masking, but there's no reason that you couldn't have four /12s so that people could mask it almost as easily. That's my point.

MS. HUGHES: I agree with you. Because I'm standing here presenting this proposal doesn't mean that I agree with it, which is part of being on the Advisory Council. I am actually against the contiguous /10 portion of that. I think there is a space in the swap that could be used for this purpose, filtering or no filtering. Leo?

MR. BICKNELL: Leo Bicknell, ARIN Advisory Council, ISC. I actually have a question, probably for Ray is my guess. The proposal to -- for the last five /8s to be given out all at once, what is the status of that in the other RIRs at this point?

MR. PLZAK: I might need help on this here, but I believe that it's been adopted in every region but the APNIC region.

MR. BOHLIN: Right, Last Call there.

MR. PLZAK: It's in Last Call there. So once it gets through the Last Call in the APNIC region, it'll be bundled up and sent off to the ICANN Board. I would strongly suspect that following the APNIC policy process, that it should pop out there sometime before the end of the year, which would put it on the ICANN Board's docket early next year. So I would guess that somewhere around February or March of next year at the earliest that that policy would be adopted.

MR. BICKNELL: I asked because my impression is it seems like that policy is highly likely to be adopted at this point. But in the unlikely event that it is not adopted, ARIN may well not know what their last /8 is, because they may get a /8 and start allocating out of it, thinking they're going to ask for more later. The other four RIRs may go and ask for all the remaining space from IANA. And then ARIN goes and says we need more and gets told, oops. At which point it may be too late for this to be able to happen because ARIN didn't know it was their last /8.

So this policy seems to require for implementation purposes that that other policy be implemented to make this actually happen. So there's a dependency there.

MR. PLZAK: So I hear two developing questions here. One, the requirement for the /10 to be contiguous, and the second is for the /10 worth of space to come out of the last /8. It could be, as Stacy said, come out of swap space. It could come out of an existing /8 that was already in our -- that we already have.

So those are two things that I hear right now in this proposal.

Any discussion on those points?

MS. HUGHES: Because the policy as written states the last one.

MR. BICKNELL: There's probably a minor semantic thing there in that. I think at least once in the past, ARIN has gotten two /8s simultaneously from IANA. So --

MR. PLZAK: Yes. In fact, Leslie, you want to talk about how we get those? How we get them two at a time and how we kind of use them?

MR. BICKNELL: So when they get the last one, it might come out of the other if they got two at once.

MS. HUGHES: The penultimate one.

MR. PLZAK: Right.

MS. NOBILE: Actually, there's a policy in effect that says that there's a formula where you qualify for a particular number of /8s. And some of the RIRs have been able to qualify for up to four /8s at a time. ARIN has gotten three at a time previously. But the agreement now, the gentlemen's agreement amongst the RIRs and IANA, is that none of us will get more than two at a time because of the eventual depletion of v4. So right now, two is the maximum we will be getting.

MR. PLZAK: Anyone else want to comment on this?

MS. HUGHES: Keep those cards and letters coming.

MR. MANNING: Bill Manning again. I look at this particular policy and I see this, in some sense, as a variant of one of the things that Leslie has already presented, which is the critical infrastructure pieces. And I'm not seeing a significant amount of difference between a slightly modified critical infrastructure component that said here's this -- you need a small amount of addresses for some special thing and we just added a special thing.

The second has to do with -- and we have heard from lots of the existent ARIN community, for you guys that live in this region, do you think you need some policy like this?

MR. PLZAK: Let me ask for a show of hands. For those of you that are looking to deploy IPv6 and what you learned yesterday, how many of you think that this is a good idea? Just raise your hands. Or you think this is a bad idea.

MR. MANNING: Nobody's awake.

MR. PLZAK: Nobody's awake.

MS. HUGHES: Go ahead, Aaron.

MR. HUGHES: I'm not sure that it's been made clear that there are some requirements for this, such as the requirement -- requires a 32-bit unique ID to operate in routers. And there isn't an IETF draft or RC on the table to replace that anytime in the future, so your routers will need it. It's not a question of is it needed in some region. It's a fact. It's rather a question of how we provide this space, what we use, et cetera, not if it's needed, in my humble opinion.

MR. PLZAK: Right, okay. Given the fact that in order to deploy IPv6 in an Internet that is predominated by IPv4, you need to have an IPv4 address. In other words, you can't deploy IPv6 without an IPv4 address, okay? How many of us think -- how many people in this room think that this is a good idea? (2 hands raised)

MS. HUGHES: Because you need, right? You have to have it.

MR. PLZAK: Okay. If you can't deploy IPv6 without IPv4, how many people think this is a good idea? How many think it's a bad idea? (2 hands raised)

MR. PLZAK: How many people really don't care? And how many people would like to have more information and more time to study or think about this? Okay.

MR. SEASTROM: Ray, since we were talking about -- you know, can't deploy v6 without v4, to put this in perspective for the room, if you want to have a network that runs OSPF today, you cannot use IPv6 addresses as router ideas, as loopbacks on the router. You have to use IPv4 addresses. And if you use RFC1918 space, which is a choice that you could make, you lose the global uniqueness that you would have for those, and you run into the whole mergers and acquisitions nightmare problem if you want to sell your ISP or buy another ISP.

So it's very important to have unique 32-bit integers in the form of IPv4 addresses available even if we are not using them for standard -- you know, host addresses and things like that. So in that context, can we ask the question again, who's in favor of this?

MS. HUGHES: Yes.

MR. PLZAK: We're going to count hands. So if you could raise your hand if you're in favor, please.

MS. HUGHES: In favor of a policy like this one, right?

MR. SEASTROM: In favor of policy like this, not necessarily this policy.

MS. HUGHES: For this reason, yes. (16 hands raised)

MR. JIMMERSON: Okay, I've got the hands.

MR. PLZAK: Okay. How many don't want to do this? (0 hands raised)

MS. HUGHES: Leo?

MR. BICKNELL: I would like to encourage another avenue as well. There are a number of people working on the problem that OSPF and PGPE and probably some other protocols have this issue in the IETF and other forums to solve the problem that they can actually use an IPv6 address.

And I would encourage everyone here to participate in those forums and push for that, because eventually, you'll want to run IPv6 with no traces of IPv4 -- you know, maybe 200 years from now, but eventually. So it would be really good if we actually fixed the problem. I take no position on the policy whether it's needed or not because I just don't know that.

MR. PLZAK: Okay. R.S.?

MR. SEASTROM: I take a position on that that's similar to Leo's, which is that we're going to want to do this with all v6 eventually. But given the IETF's track record on delivering new protocols in a timely fashion and the vendors' track record in actually implementing those protocols once they're defined, it is almost certain that we will be out of IPv4 space by a couple of years before these start being something that you can order from your Cisco representative.

MR. PLZAK: Okay. Now, given the fact that -- actually two subsequent questions I'd like to ask here. First of all, so the general consensus right now is 16 to 0 says that it's a good idea to do something like this. Now, to give the AC a little bit more information, I would like to ask two more questions.

The first question is how many think that this address space should be reserved out of the last /8 that ARIN receives? So if you think that we should take this address space out of the very last /8 that ARIN receives, I'd like to see how many think that we should do that. Or should we instead take it out of some existing address space we already have? In other words, take a set of addresses and just set them aside and hold them there for that purpose, make a reserve today.

So how many people think we should wait for the last /8 in order to do this? Nigel?

MR. CASSIMIRE: Nigel Cassimire, CTU. I'm seeing in the policy it says -- and there is a note that the policy proposal hints that the /10 should come out of the last /8 received by ARIN from IANA.

However, it does not say so explicitly, leaving the final decision up to the ARIN staff. And I'm cool with that, because I think, given that ARIN staff would be the one who would be -- what, actually doing the allocations -- I would be cool with leaving them the discretion as to what's the neatest thing to do, whether they take it from the last /8 or the -- or from someplace before.

MR. PLZAK: How do others feel like on that subject?

MS. HUGHES: I have a comment.

MR. PLZAK: Go ahead.

MS. HUGHES: I think we should do this now rather than later. I don't think we should wait for the last /8. I don't know if I should be doing this from the podium or not.

MR. PLZAK: That's fine.

MS. HUGHES: Okay. But if people are trying to bootstrap their IPv6 networks now, they should be able to acquire IPv4 resources for this reason.

MR. MANNING: This may be a really silly question for ARIN staff: has anybody ever thought about or had communications with, say, the defense networks to simultaneously use, say, 22 or 26 that are being used for private purposes?

MS. HUGHES: V6 or v4?

MR. MANNING: Twenty-two /8, 26 /8.

MS. HUGHES: Oh, oh.

MR. MANNING: The defense network /8s that are currently used for private purposes that will never be used in public -- potentially having a conversation with them to use it for purposes like this.

MR. PLZAK: The answer to your question is no. We are concentrating more on getting them to give space back. Bernadette?

MS. LEWIS: Thank you. Bernadette Lewis from the CTU. I am inclined to agree with Stacy. I think that we should do the reservation now rather than later. We don't know what sort of eventualities could arise at the last -- when you do get the last allocation. And my preference is we have identified a situation. We have a solution. Let us make the provisions and we'll see what happens two years hence. Thank you.

MR. PLZAK: Is there anyone that doesn't think that the decision of where to take this address space from should not be left to the ARIN staff? In other words, is it okay for the ARIN staff to figure out where to do this, where to get this address space? Anyone think that that's not a good idea? Okay.

One other question I'd like to ask you then and get an idea is -- well, actually let me ask for a show of hands on it. How many people would be in favor of the ARIN staff making a decision about where to take the -- to reserve the address space for this purpose? How many people are in favor of the ARIN staff doing that?

Raise them high, please. (14 hands raised)

MR. JIMMERSON: Okay, I've got it.

MR. PLZAK: How many people are opposed to the ARIN staff being given that decision authority? (0 hands raised)

MR. PLZAK: Thank you. Now, one last question I'd like to ask.

This proposal also talks about a contiguous /10. How many people will think that this should come from a contiguous /10? (1 hand raised)

MR. JIMMERSON: Got it.

MR. PLZAK: How many people think it should not come from a contiguous /10? (9 hands raised)

MR. JIMMERSON: Got it.

MR. PLZAK: So to iterate the information that we passed on to the Advisory Council, the number of people that think that this is a good idea, that something like this should be done is 16; nobody opposed. The number of people that think that the staff should be the one that decides where to take this address space, 14 yes; nobody opposed. And the number of people that think this should come from a contiguous /10; 1 in favor, 9 against. So this information will be provided to the Advisory Council for this consideration.Thank you, Stacy.

MS. HUGHES: Thank you on behalf of the AC. (Applause)

2007-14: Resource Review Process

MR. PLZAK: What's next, Einar?

MR. BOHLIN: 7-14.

MR. PLZAK: Okay. Policy Proposal 2007-14: The resource review process. This was proposed about a year -- over a year ago and it's already been discussed at three public policy meetings: the ARIN XX, XXI, and at our meeting in Kingston earlier this year. Community consensus has been to revise this proposal. The current version is current as of the 14th of August this year. It's under discussion. We are the only region that are discussing this.

The understanding is that it provides a clear policy authority to audit or reclaim pool resources, guidelines for how it should be done, and a six-month minimum grace period for the user to stop using the space. The shepherds are both in this room. That's Leo, who's going to present the proposal, and R.S., who's sitting in the back.

So some history on the discussion. At the meeting in Denver, with 123 people in the room, there were 29 in favor of saying that the policy should explicitly state ARIN has a right to do audits, and there were 35 against doing it as is written. In Kingston, with 50 people in the room, 34 supported the general idea -- and in support of the current text, it was pretty much at an even split, 5 for, 6 against.

So with that, I'll turn the mic over to Leo.

MR. BICKNELL: Hi. My name's Leo Bicknell. I'm the chairman of the ARIN Advisory Council, and I work for ISC in my day job.

The first thing that I want to tell you about this policy proposal -- do we have the text on a slide -- is -- it's in your manual on page 4, if you're looking for it. Yes, if you've got it there.

I'll talk over here. This policy, one of the reasons it's been around for so long is there's been a great deal of confusion about the policy. And so I urge you, if you haven't looked at this, to look at both the policy in your packet, but also, as a reference, you need to look at the Registration Services Agreement that you would sign with ARIN if you have resources.

So either the contract that you have or there is a copy on the ARIN website fairly easy to find. You can just search for RSA in the box and get a PDF version.

And I say that because an awful lot of people in the discussion of this policy felt that this created a situation where they could be audited or -- he's getting the now over-loud mic. Let's try for a middle level, how about that?

An awful lot of people felt that this policy created a situation where they might be audited more than in the past or might have their resources revoked more than in the past.

And the reality is the current situation for when ARIN might audit and/or revoke resources is pretty much governed by the RSA, because there's no significant policy language around it. And the RSA is, in many people's view, a little bit lopsided on this in that it basically allows ARIN to audit you whenever ARIN wants, for any reason that they want, and to take whatever action they deem necessary to correct any problems that are found, including revoking space.

So a bunch of people stood up and said they don't like this policy proposal, for instance, because you could be audited every two years, when the reality is right now, in a particularly bad set of circumstances, you could be audited every day if somebody really wanted to. So it's important to very carefully read and keep in context.

You know, we've had a very good situation where the ARIN staff has done their job remarkably well and hasn't done anything evil, which I applaud them for. And I don't suspect that's going to change any time soon. But Owen really wanted to close the gap here with this policy, so he's approaching it from two sides.

One is he felt that it shouldn't be possible that you could be audited every day or every week or every month. And so he wanted to place some limits on the staff in that respect and define it for both parties.

On the other side, I think there are certain people who felt that ARIN staff may, in some cases, be hesitant to audit people and reclaim space because of the vagueness of the relationship and probably, in part, not wanting to appear evil. You know, that's not what ARIN's out there to do either.So I think what Owen really wanted to do was close the gap in both directions. On certain bad parties he actually wanted to get them looked at a little more closely, but, at the same time, he wanted to ensure that there were better policy protections for the average ARIN member so that abuse couldn't happen at some point in the future under some future regime or administration, or however you want to view it.

So that's kind of the history of this policy. It has been through two sets of revisions now. As with anything that affects audits and reclaiming space, there is a lot of attention to the details in this and so it's been through a couple of steps of revisions. Hopefully, it is fairly close to acceptable, but obviously, this is a forum where if there are still issues, that they need to be brought up and we need to get the text of this corrected.

So with that, does anybody have any comments, questions, concerns?

MR. HUGHES: Hi. Aaron Hughes. I support 2007-14 as written.

MR. PLZAK: The microphones are open.

MR. SAMUELS: Carlton Samuels, UWI. I support the proposal as written.

MR. PLZAK: Anyone else? Anyone want to get up and say why they think this shouldn't be done or why it's not needed?

Okay. Let's go for a show hands on this then. How many people in the room think that this policy is a good idea as written?

Please raise your hands. (15 hands raised)

MR. JIMMERSON: Got it.

MR. PLZAK: And how many people do not support this policy? Please raise your hands.

MR. JIMMERSON: Got it. Oh, is there one over here?

MR. PLZAK: Yes. Yes, one over here.

MR. JIMMERSON: Where?

MR. PLZAK: Right there. (1 hand raised)

MR. JIMMERSON: So there's one there? Okay.

MR. PLZAK: With the number of people saying this is a good idea and support it, we have 15. The number of people that don't like this idea, one. This information will be provided to the Advisory Council for further discussion and further information.

Thank you, Leo.

MR. BICKNELL: Sure. (Applause)

MR. PLZAK: Well, I don't have a slide for this.

MR. JIMMERSON: 2008-2?

MR. PLZAK: We're at open mic.

MR. JIMMERSON: It's on Wednesday PPML, 2008-2.

MR. PLZAK: No, we're not ready to do that yet. You've got to go to open mic, right?

MR. JIMMERSON: You want to do open mic?

MR. PLZAK: Well, let's see what happens here first. We may go ahead and go with that proposal.

Open Microphone (Morning Session)

MR. PLZAK: Right now on the agenda, we have an open microphone session scheduled. So I'm going to throw the microphones for anyone that wants to discuss any topic relevant to policies. And this could include some of the carryover from some of the discussions from yesterday that you had with the regulators, that regulator session, and so forth. So the microphones are open for anyone.

Is there anything that came up yesterday in your discussions that you think should be turned into a policy or some things that ARIN should look at in a way it does its business with regards to resource allocation processes? Nigel?

MR. CASSIMIRE: Nigel Cassimire. What I'm saying isn't really related to what you just asked, but it's me trying to think of how maybe beyond the outreach efforts that have been made so far, can we involve the -- I would say more of the practitioners in the policy development process, right? Because some of these things that are happening in the room here certainly will impact on or could impact on the way that, say, ISPs do their operations. And you know, we still have a long way to go, I think, before we can say, well, we have active participation of the ISPs in the Caribbean in considering these issues and making inputs and so on.

So I was just -- I'm just really trying to think of maybe what else we can do to reach them. And maybe not reach them in a formal way like this, but reach them -- you know, kind of where they live, in their working environments, and let them know -- you know, the way you're doing this now with respect to routing and address allocations and usage, things could be better or in the very near future you're not really going to be able to do this anymore.And you know, basically get them engaged, you know?

And I'm really not sure what the answer is. So I think that's a question that I just want to put out there that I'd like to get some other thoughts on and suggestions.

MR. PLZAK: Okay. I can tell you this, we are actively looking at a number of ways.

Richard, I believe, this afternoon is scheduled to give a discussion or a presentation about what our reach out activities we have been doing. As I've stated in the past, we are more than willing to send persons to attend any type of a session to discuss what we do and how we do it and the relevance of ARIN to the persons involved.

We are also actively engaged in expanding the participation in ARIN meetings in general and in the Caribbean sector meeting in particular. As I said, we will have -- every year, at some point, we will have a regular policy meeting. There will be, like, three full-fledged, 100 percent ARIN meetings.

There won't be a Caribbean sector meeting. It'll be an ARIN meeting in the Caribbean. There won't be that little tag "sector" on it as it has today.

In addition, we're looking at ways of increasing the remote participation to include JABBER chat rooms where people can talk about what's being discussed; a separate room where people can actually vote on the proposal. So while we're doing a show of hands in the room, the people that are remotely participating can also have their opinions stated and known. So those are kind of the active things that we're looking at.

We're looking forward to the day in the not too distant future when someone from the Caribbean will attend every ARIN meeting, and some of those they will attend in person and some of those they will attend online, but the experience will be basically the same. The only difference on the remote online, you'll have to find something to do on your own for the social.

Other than that, you'll be able to do everything to include participate in Plinkos and Wheels of Fortune. You'll get to be able to draw prizes. You'll be able to engage in conversations and networking conversations with other persons that are remote participants.

You will be able to ask questions real-time in terms of we'll have a separate microphone, a virtual microphone, queue. So we're looking at all that. So that -- you know, it's very difficult to travel in the Caribbean just given the way the airlines operate. And now with the increased travel costs, it becomes even harder to do. So we're looking at ways of increasing that participation.But that being said, you're right, Nigel, there are other things that we have to do. And so we are always open to suggestions and ideas. We are willing to try almost anything once just to see if it'll work. And if it does work, then we look at how we can build on that success.

And so we do have a suggestion program. You can just drop a little suggestion into the box and see what happens. So there are a number of avenues to get that information to us. But -- you know, we can only have so many good ideas sitting in Chantilly, Virginia.

The real good ideas have to come from the Bahamas and Antigua and Barbados and all the other islands in between, and Jamaica and the Caymans, the Turks and Caicos, so all of them. That's where the ideas are. That's where the community is going to grow, so we need to have that.

Any other comments on that?

Anybody have an idea or something they'd like to throw out in the direction of -- please, the microphone.

MR. WALLACE: Good morning. In reference to the --

MR. PLZAK: Would you say who you are, please?

MR. WALLACE: Oh, Romel Wallace, Bureau of Information and Technology.

MR. PLZAK: Okay.

MR. WALLACE: The VA. To follow up with that, in terms of I was looking at the membership of -- members with ARIN. How many members in this sector is signed up with ARIN?

MR. PLZAK: I'm sorry?

MR. WALLACE: How many members do we have in this region?

MR. PLZAK: Yes, in this area. Megan, do you happen to know that?

MS. KRUSE: I don't.

MR. PLZAK: It's under 50, I'll put it that way. I know it's greater than 25, but I think it's under 50. And part of the reason is the fact that you either become a member of ARIN by paying $500 or you become a member of ARIN by receiving an allocation. And one of the barriers to receiving allocations in this part of the ARIN region is that the allocation size is too large. So that's why the policy proposal on the table for this afternoon, the minimum allocation size for the Caribbean, is important.

Another reason why it's important is because there are not the people in the Caribbean that have sufficiently large potential customer bases to qualify under the current policy. And also, you've got -- you know, some dominant telcos in the region and a few other things that regulators are aware of as further problems.

So the straight answer to your question is it's not as many as there should be. And there are policy avenues to make it better.

MR. WALLACE: And just another quick comment. I think maybe some of the members, like, over 25, when it has, like -- it comes to voting, maybe it could have a certain amount of members that may have to be present. I mean, that might be difficult, but I don't know if we have any members at all present for this region. Is that the case?

MR. PLZAK: Let's differentiate between voting. If you're talking about voting on policy proposals, you don't have to be a member.

MR. WALLACE: Okay.

MR. PLZAK: Anyone, anyone can be there. I can go out right -- well, actually this gentleman sitting right here transcribing notes, if he's got an opinion one way or the other on a policy proposal, he can raise his hand and it will be counted. I mean, that's how open this meeting is, okay?

Now, when it comes to voting for the governance structure of ARIN, for members of the Advisory Council or members of the Board of Trustees, you must be a member of ARIN. Not just a member of the community, but a member of ARIN. And in order to do that, you must either pay the membership fee or have received a direct allocation from ARIN.

So you have to differentiate between those two kinds of voting, what you're voting for. The first one I would argue is probably more important, which is to get people to participate and actually vote on policy proposals. And there's no qualification to do that other than be interested in it.

MR. WALLACE: So when it reaches that level with voting, how many people from this region participate?

MR. PLZAK: We don't track the number of people that vote on -- in that regard. We do -- we can tell you, for example, how many people are subscribed to the Public Policy Mailing List because by virtue of the discussion on the list, the Advisory Council uses that discussion, as you would, to help determine consensus. And I can quite frankly tell you that the number of people from this portion of the ARIN region that participated on a Public Policy Mailing List is very few.

And this goes more along the line of what we were discussing earlier in terms of getting more participation. So wherever we can encourage people to subscribe to the policy list and participate in the discussion, that's something they can do from Internet cafe. That's something they can do from anywhere they can gain access, you know. It's an open mail list. You just have to have an e-mail account.

As we increase the remote participation other ways of participating in that process, and so that's where we need to go. But we have to raise the level of awareness, and I think that's where Nigel was going as well. And that's where we need the ideas of things that we can do to help raise the level of awareness. So any ideas that you may have in that regard are certainly welcome.

MR. WALLACE: Okay.

MR. PLZAK: Megan?

MS. KRUSE: As of April, I don't have a more updated number, but as of April, we had 32 Caribbean members.

MR. PLZAK: Okay, thank you. Anybody else want to discuss anything?

Okay. What I would propose then is that we go off to the break and we will start the discussion of Policy Proposal 2008-2.

And we will discuss this up into the break time.

MR. JIMMERSON: We've only got a few minutes until the break.

MR. PLZAK: What time is the break?

MR. JIMMERSON: It's quarter to.

MR. PLZAK: Oh, I thought it was 11:00. Okay, disregard that.

MR. JIMMERSON: Okay.

MR. PLZAK: We only got a few more minutes, let's just take a longer break, okay? Rob, I'm setting the stage for you, R.S. They want to see you up here.

MR. SEASTROM: Building the anticipation, right?

MR. PLZAK: Building the anticipation. So we will continue back in here at 11:00 a.m.

Thank you. (Recess)

2008-2: IPv4 Transfer Policy Proposal - Recent PPML Survey Results

MR. PLZAK: Anyway, at long last, Policy Proposal 2008-2.

MR. SEASTROM: Are you sure this time?

MR. PLZAK: Well, I'm not ready for you quite yet. Anyway, at some point in time we want to find out if Kingsley Smith ever made it back in here. Anyway, 2008-2, the IPv4 Transfer Policy Proposal. This was designated a formal proposal in February of this year. It's been discussed twice: Once in Denver and once in Kingston. The consensus in the community has been to revise it. And it's gone through several revisions. In fact, I think another revision will probably be released to the PPML very soon, within the next 24 hours or so.

In the other regions, it's being discussed -- a similar proposal -- I will characterize it as being a much more liberal proposal -- is being discussed in the APNIC and RIPE NCC regions. And in those regions as well, the consensus has been to continue the discussion. The general understanding is this allows an organization to transfer excess IPv4 addresses to an organization with a need for addresses and creates a listing service. The shepherds are Scott Libron (?) and Stacy, who happens to be in the room.

So far to date, discussion statistics: in Denver with 123 people in the room, there were 49 people, 7 against, ARIN continue to work on a relaxed transfer policy. And should it have a need-based qualification, 54 to 3 in favor of doing so. And assuming that the policy retained a needs-based qualification should the policy take effect sooner than depletion of the IANA free pool, about a 2 to 1 on there, 35 for and 13 against. In Kingston, with 52 people in the room there were 25 in favor and 3 against. Should it be needs-based, 32 -- overwhelming 32 to nothing in favor of being needs-based.

There was a survey conducted between the 14th and 20th of August this year. The persons who were directly contacted were the 1,805 subscribers at the time to the public policy mailing list. The number of persons responding was 202, or 11.19 percent of the subscribers to the list responded to the survey. The survey was conducted by the Advisory Council to get further information as to the opinions of the community in various aspects of this.

So this is some more stuff on this area. We'll do a presentation and discussion. I'll put the understanding slide back up here. Do you need a copy of the survey results?

MR. SEASTROM: Yes, if you've got it.

MR. PLZAK: So I'll turn it now over. If you haven't figured out yet who's going to be doing the presentation for the AC, it's R.S., Rob Seastrom.

MR. SEASTROM: Hi, everyone. I'm Rob Seastrom. I'm on the ARIN Advisory Council.

If you don't already have it open, you probably want to open up your packet with the discussion guide in it for this part of the discussion, because this may well be the longest proposal we've ever had in ARIN history.

So let me give you a little bit of background on this. There's been some discussion for, oh, probably a couple of years at this point. What are we going to do to get IPv4 addresses for people who need them once there's no more free pool to take them from? And a solution which has been widely considered is to create some sort of a way that organizations that have addresses that are surplus to their needs can transfer those addresses to organizations that need the addresses.

And there's a whole bunch of concern over the ins and outs of how something like that would work, starting with the fact that addresses aren't property. They're assigned to an organization by ARIN in consideration of need for those addresses that has been demonstrated. But the organization does not own the addresses and can't buy or sell them. It's not like a piece of property.

One of the concerns that was raised is that this has the possibility of creating a huge windfall monetarily for organizations that have address space from many years before, when the assignment criteria were more relaxed.So this proposal comes from an ARIN Advisory Council meeting that happened -- was it January of February of this year? January of this year -- in Reston, Virginia, where the ARIN Advisory Council got together for a weekend to come up with a proposal that would allow for transfer of IPv4 addresses between organizations.

And what we came up with is one, two, and three-quarters pages long, including the rationale, and has been through -- I think as of yesterday -- 31 minor revisions. So to be frank, there's a lot of concern that we may not actually be able to get consensus on something that's this large.

But we're interested -- you know, everybody seems -- if you take a look at the survey results, there should be some sort of transfer policy, but people are very divided on what the details of that policy ought to be.

So if we go through and look at the individual points of the transfer policy 2008-2, Section 8.1 reminds us that number resources are not sold. By exchanging money, optimally, it would be an amount of money that covered the organization that was providing the addresses with the costs of renumbering out of that address space into other address space. Number resources are issued based on justified needs, so there's a requirement that an organization would have clearance from ARIN that would justify to ARIN their need for the address spaces before getting the address space.

It covers, in Section 8.2, the requirements for showing that one is not just selling a company for the purpose of moving the address space the company has assigned to it. This also covers -- let's see, simple way of passing address space back and forth between, as covered in Sections 8.3 and so forth, that the transfer and transferee have to both be in the ARIN service area.There has to be a registration services agreement signed, conditions on the transferee and transferor, and the fees and prequalification required, and concern for de-aggregation, and other such stuff.

So this is a very large proposal. And we're interested in getting feedback -- I guess we're interested in feedback on the same questions we asked at the other --

MR. PLZAK: Yes, feedback on those questions; plus, we can propose new questions here as well, which would then be carried over in some form to the next meeting.

MR. SEASTROM: Right. So we're looking for both answers on these questions from the transfer proposal survey and for suggestions for new questions to be added to the survey for L.A.

MR. PLZAK: Yes, we can do that, and also when we frame the questions looking for a show of hands, we can frame those any way we want to in this meeting as well.

MR. SEASTROM: Okay.

MR. PLZAK: So I'll allow you -- since the AC is more involved in this, I would allow you to do that as well.

So I guess the first thing that I would do is open the microphones and see if there is anybody who would like to clarify or draw specific attention to certain parts of the proposal, or contradict anything I've just said as being completely incorrect, or anything like that before we start talking about the proposal -- transfer proposal survey questions.

Bill Manning is always happy to contradict me as being completely off in the weeds.

MR. MANNING: I think this was actually set up for Aaron, so we'll move the microphone.

To my mind, the particular set of tradeoffs here is efficient utilization of the resources we've been given -- the IPv4 pool -- and the impact on the routing system. The discussion tends to focus on those two areas, and those tend to be, at least at this point, diametrically opposed to each other. The more efficiently we utilize the address space, the greater the impact on the routing system. And that is a subtext that needs to be actually discussed, I think.

MR. SEASTROM: Do you want to provide any more commentary on that, like the two sides of that?

Or do you want me to do that?

MR. MANNING: What would you prefer?

MR. SEASTROM: I would prefer that you do it.

MR. MANNING: Thank you. The particular issue here is efficient utilization of the address space. Allow an organization -- we will presume for purposes of example that Mr. Seastrom is organized.

MR. SEASTROM: That would be a big assumption.

MR. MANNING: And he has acquired a block of address space in 2003 from ARIN that's a /16 or 64,000 addresses, at which point he figures out that he only really is actually using about 1,000 of them. And so he recognizes that Stacy Hughes needs some additional address space. And Bill Manning needs some additional address space. And Bernadette needs some additional address space. And so he can carve off small pieces out of this /16 and give Stacy some, and Bill some, and Bernadette some. And then we all get the address space that we can use -- effective utilization.

But then that turns around, and each of us will announce into the routing system our own smaller piece of this once large block. The routing system, depending upon when you look at it, is constrained to approximately a million entries. And the passing of this policy proposal and the expected outcome of turning a large block into a bunch of small blocks, each of which will be independently announced, will derive the entries in the routing system past the capability of most current hardware.

Did I get it right?

MR. SEASTROM: Yeah. Now, a couple of things that should be considered are, number one, we've driven past the capability of all currently deployed hardware on several occasions in the past, and the solution has been to get more hardware. But at the 1 million or 1.5 million, or where that number happens to be number of routes in the global routing table -- which, by the way, we currently have, what, 280,000 or something? About like that.

Yeah. There end up being problems that are not solved simply by putting out routers that have more CPU and more memory in them. There are constraints on BGP being able to actually handle it in its current form.

And while people have made BGP handle 2, 3, 4 million routes in the lab, as we know the difference between theory and reality is that in theory, they are the same, and in reality they are not.

So once we go -- I would say that Bill's comment about being constrained around a million routes is saying that we are reasonably sure that the Internet will still work okay if we apply appropriate technology to it at a million routes. And beyond that, it becomes increasingly unsure.

So we want to be careful that we do not put too many routes out there that are eligible to go into the routing table because once we have done that, there is no way to effectively back out of it.

MR. PLZAK: Leo?

MR. BICKNELL: Yes, Leo Bicknell, ARIN AC, ISC.

I think Bill, perhaps, oversimplified the situation, and R.S. approached a more complete picture of it in that there are two curves here. Over time, there is the capability of equipment which is always increasing. Increasing at different rates. Sometimes that curve is very flat for long periods of time, and other times it goes up rather dramatically in short periods of time. But I believe it is an always increasing curve.

And then there is also the curve of new routes being added. The size of the route really isn't important because a route is a route. So there is a curve of new routes being added. The important thing to recognize with the second curve is that with no transfer policy, when the RIRs have run out of space, in theory, the rate of new routes being added to the table goes very close to zero.

It won't be exactly zero because some people will subdivided for traffic engineering and do other things, but that curve that has been marching at a fairly constant rate always upward is going to come very close to flat lining.

And so when we look at proposals like the transfer policy that may inject some additional prefixes back in the table, I think the key is that the policy not inject them at a faster rate than we've done historically. That is, if there were more free IPv4 space and that curve could continue at the same rate, the transfer policy needs to aim for that line or under. And as long as that occurs, the people who are building router hardware already have decades of history of how to meet this curve. I'm not saying there aren't challenges in their future, but they'll be able to meet it.

MR. SEASTROM: What you're saying is no surprises.

MR. BICKNELL: No surprises, and here particularly so because the hardware refresh cycle for a lot of large providers is five years. Vendors have to develop a product and get it out, so you probably have to have a real good prediction 7 to 10 years out for the vendors to be able to meet it. But if we can keep the prefix growth line predictable for 7 to 10 years out, then I have confidence that the vendors will find a way to meet the solution. And in fact, there is -- right now I would argue a larger buffer in hardware capacity compared to what it's being asked to do by percentage than we've ever had in the past. We are operating right now at 280,000 routes out of a million capability, so 20 percent. There are many times in the last decade where people have been running their routes at 90, 95, 98 percent of their capability. So I think we've got a little wiggle room.

MR. SEASTROM: One thing that Leo touched on is in the absence of a transfer policy, a lot of people have said why have a transfer policy at all?

Once we are out of IPv4 address space, won't that create a serious incentive for people to get with the program and use IPv6? Well, the counter argument to that, which I am all for getting people onto IPv6 -- the counter argument is that there is a compelling public interest in having the data in WHOIS reflect some form of reality.

And there is already a shadow market, a nonsanctioned transfer market in IPv4 space. Sometimes you can even see it popping up, sort of like the tip of an iceberg in places like eBay. There have been /24s listed on eBay. But it's mostly in a much more shadowy sort of situation involving perhaps people who have hijacked blocks of IP address space -- put in unsanctioned changes into WHOIS, and splitting it up that way.

And in many cases, I'd say probably in just about all cases, you can assume that somebody who buys one of these questionable IP address blocks is going to be very interested in, as we say letting sleeping dogs lie, which means don't make any changes to WHOIS at all.Don't disturb the status quo.

So a situation in which there are transfers being done that ARIN is not involved in in some way, leads to a situation where the data in WHOIS is increasingly unreliable. And my opinion, speaking personally rather than as the presenter, is that that doesn't serve anybody's best interest.

Bill?

MR. MANNING: Who? That's much nicer this morning this time around. There is another perhaps more insidious problem here. In the absence of a transfer policy of some flavor, we end up in a situation where an unscrupulous operator -- Richard, the unscrupulous operator, de jour, sees that the address space is running out and creates plausible business plans to go to ARIN and suck up all the remaining address space, effectively removing the opportunity for competition for new entries into the markets, et cetera.And if that occurs, that cries out for the aggrieved party who cannot participate, to go talk to the local regular and say my rights have been abused. I cannot enter the market. Please fix this. And so a transfer policy actually allows the industry the opportunity to continue to allow and foster growth and innovation as opposed to locking into the incumbent players.

MR. SEASTROM: In the shadow of the comment that you just made, not having a transfer policy would be played by the people who do not want our consensus community based governance model to continue and believe that they can do a better job.

That would be cited as an example of us not being able to govern ourselves, and, of course, they should be given the reigns.

Leo?

MR. BICKNELL: I guess this is a disagree with Bill day for me.I see the transfer policy as unnecessary, personally. And the reason or that is that I believe the situation in IPv4 to IPv6 transition is remarkably analogous to the Y2K transition. They're both the situation of the numbers just get a little bigger and everything has to change to deal with it.

And when I looked at Y2K or other similar transitions that have happened in the past, the ones that were successful were successful because they had a defined date on which the old technology ceased to work and the new technology worked.

Because you get the power of the crowd. Everyone goes to Cisco and Juniper and whatnot together and all says in unison we must have this by this date or the entire industry fails. And lo and behold, pretty much every time, 98 percent of the cases, the vendors step up and make it happen, and the IFPs and -- you know, the banks, and Y2K, whoever, it's a non-event.

The cases where it fails is where there is a long, prolonged transition where different parties transfer at different times. Where someone is able to reap a first mover advantage, or a last mover advantage, or any of those sorts of things. So anything that prolongs the transition will actually create more disparity in the marketplace and create more winners and more loses.

And specifically to Bill's point, if you look at somebody -- if your worry is someone cornering the address space, your two scenarios here basically is -- in the standard scenario we have today, ARIN goes around and does a really good job as R.S. pointed out reading his 98 pages of document, and ensuring that he's not lying about his immediate need -- that he has contracts, and customers, and all of this other stuff.

In the potential world of transfers for money, the rich can just buy them up. And having been on the AC and talked to a number of providers, I have heard from at least one person working for a large multinational corporation who said, you know what, throw in two or three hundred million dollars of buying IP addresses to keep our competitors from using it, that's something our executives would consider. And that to me is amazingly frightening.

So for those two reasons --

MR. SEASTROM: Just to interject here, there are no transfer proposals currently in the IRPEP that do not include a requirement that one show need.

MR. BICKNELL: I understand that. I believe these companies could create business cases that show need. There are many companies -- Comcast was one of them who as just here, talking about if they did everything they're talking about, they need 100 million IP addresses. They could decide to spend the money to do that next year as opposed to over the next eight years or whatever their current schedule is. And they may see the money to accelerate those deployments as money that allows them to buy IP addresses and keep their competitors from deploying those services. Not necessarily Comcast specifically, but since they were here and presented 100 million, I'll use that as an example.

So I think need is something these people can show if they want to. And the fact that they can then just throw money at the problem, that's something corporate America understands very much -- how to throw money at a problem and how to buy out resources that your competitor needs.

MR. SEASTROM: I would say that the Comcast representative who was here yesterday, when he talked about the 100 million addresses being something that is simply not going to happen was very correct in his assessment -- the amount of scrutiny that such a request would get would basically tear down any kind of veil of plausibility.

MR. BICKNELL: It's an opinion.

MR. SEASTROM: I understand. Does anyone else have any comments?

MR. CASSIMIRE: Yes, Nigel Cassimire again. Until you spoke, I thought I might have been the only one in the room who was still at that place where I was trying to figure -- you know, why we needed this thing. I mean, I can understand efficient utilization of address space. And sometimes when I speak, I guess my telephone background shows, because certainly in terms of telephone numbering and so on, it was a very structured ordered hierarchical kind of thing, which is totally different than what we have today.

But certainly I understand the concern, as well, about limiting the proliferation of routes. But -- you know, in my mind, we're running out of IPv4 addresses and we have a solution. The solution, of course is, well, we start transitioning to IPv6. And to me, creating a market for transfers -- you know, I really don't see what the benefit is in doing that if it is that we have a clear solution in terms of transitioning to IPv6 because more addresses are available in a practically unlimited supply. So one would not expect there to be financial incentives for people to do things like hoarding and blocking customers and that sort of thing.

So I'm wondering, so what is the real incentive? Usually it's economic. And I'm wondering, are people seeing it that it would be cheaper to buy IPv4 addresses rather than spend money to transition to v6? Is it costing less to be in the market?

MR. SEASTROM: Let me speak to that for a minute, because you raise a couple of very important points there. One of those is I used to be very much against the notion of having any kind of transfer policy. And I sort of changed my opinion on that over time. And my reasons were very similar to the ones that you just brought up. When we're out, people will move to v6. That's good. The problem is that we have an installed base of IPv4. And all of the requirements out there -- when my father gets on the Internet to send mail -- when my father gets on the Internet to check his stocks or to buy an airline ticket, he doesn't care what's underneath. He only cares that he's able to get to the website that he wants.

So there is no value in selling IPv6 rather than IPv4 to people who are like my father. All they care about is the number of websites and e-mail addresses that are reachable. E-mail is easy, because you just have mail server that has both kinds of addresses on it and you're done.

But he cares about the number of sites that are reachable via his computer. And right now the vast majority of them are not on IPv6. So a computer that only has IPv6 will have trouble getting to 99.999 percent of the Internet. Now, if there were some sort of transitional technology in there, like NAT PT, that helps a lot. But there's still a lot of stuff out there, particularly in the enterprise sector.

And the enterprise people do not understand at all what all the fuss is about because for them, they only need one or two IP addresses for their entire company. And they use RFC1918 space internally. We, as service providers, have networks that are large enough that even if the 1918 space didn't potentially collide with our customers' 1918 space, we wouldn't have enough of it. You can run out of NET10 with an ISP that is really not that large.

So this is a problem that is more in the direction of service providers than the enterprise people, which causes a problem with what Leo was suggesting, where you just push on the vendors and make it happen. Well, who is buying most of the equipment these days from the large vendors like Cisco, and Force10 and Foundry, and what have you? It's not the service providers. It's the enterprise people.

We buy a lot of equipment that is large and very neat in terms of -- you know, the capacity through it. But in terms of their revenue streams, which is what the router manufacturers are going to look at, it mostly comes from the enterprise. And the big push for IPv6 is not coming from there.

In any event, when we talk about having a market in IPv4 address space announcement rights -- not in the address space itself -- do I really think that this will extend the life of IPv4 for any appreciable amount of time? No. What am I trying to accomplish then when I say that I advocate it? Well, I'm trying to accomplish two things. One is if we don't have this, there will be a lot of people pointing at it and saying, well -- you know, you could have gotten a lot more time out of IPv4. You were remiss in not doing it. Without quantifying that and realizing that in this case "a lot more time" means three to six months.

And that creates a big problem, because in the Internet governance arena, we have to constantly defend ourselves against that kind of accusation that we didn't do everything we possibly could. The other, which is far more important, is -- I don't know how familiar you are with the level of inaccuracy of data in WHOIS, but it has problems already. And just as Bill was talking about the problem with the routing table getting huge all of a sudden by having policies that encouraged de-aggregation, if we have no transfer policy and address space is sort of changing hands behind the scenes, the amount of inaccurate data in WHOIS is also going to go way up.

And as somebody who deals with spam coming to my customers; as somebody who deals with routing anomalies, it's very important that I have information and WHOIS that I can count on to get me to somebody who can help. When there's stuff that I don't expect in the routing table, when there's a flood of spam coming in from somebody's server, and I think that as rationales go for having a transfer market go, that that is the number one concern, at least from my stand point -- preservation of accurate data in terms of what organizations have what IP addresses.

MR. CASSIMIRE: And have there been any other suggestions for solving that WHOIS issue?

MR. SEASTROM: I would love to hear any other suggestions you have.

MR. CASSIMIRE: Give me a minute.

MR. MANNING: If you want to know more about WHOIS --

SPEAKER: Could you use the mic, sir, please?

SPEAKER: Microphone please, sir, so we can capture your sterling words?

MR. MANNING: My sterling words. With regards to WHOIS information, Cathy is friendly with people who have had those kinds of discussions, sometimes aggressively, for the last few years, so she -- if she's willing -- could talk at length about WHOIS.

I think that, for me, this particular policy actually talks to a fundamental change in the way people think about their use of addressing. Historically, we've thought of addresses as something that goes into every end system. Your computer has got an address. My cell phone has got an address. My glasses frames might have a couple of addresses. One for the left lens; one for the right lens. Right?

And in an IPv4 context, there aren't enough. The 2007 population survey said that there were approximately 6.5 billion people on the planet.And if you look at the entire IPv4 space, there's only about 4 billion addresses. There aren't enough. Even if we could give one to everybody and could figure out how to route, there's not enough IPv4 space to give everybody their own IP address.

There's certainly plenty of IPv6 space. The transition issues are the most interesting ones.

And as we discussed and experienced yesterday, the possibility of actually having a translation function and using IPv4 now as to translate between an IPv6 universe and an increasingly diminishing IPv4 universe in a clean way, in a sustainable supportable way, allows people to transition gracefully. And so that's a change in the way you think about the use of IPv4.

Picking on Leo day, if every enterprise purchasing manager went and knocked on the door of Force10 and said we need a box with these characteristics in the next three months, no way that Force10 or Juniper or Cisco or AT&T or anybody else can build, test, and deploy. Right? So we can't expect, even if there is a tremendous unified market demand, we can't expect people -- equipment people to actually deliver in a timely manner. It's got to be spread out.

MR. SEASTROM: Thanks, Bill. I'd like to ask a few questions of everybody here. These are questions that we had in a survey that we had on the public policy mailing list earlier this year. And we'd like your input -- this is in broad terms -- what form 2008-2 should evolve into before we get some sort of transfer policy into place.

So the first question is --

MR. PLZAK: R.S., they all have copies of that.

MR. SEASTROM: Okay, everybody has a copy of this in your handouts. So if you want to get that out and look along with me.

MR. MANNING: Is this a test?

MR. SEASTROM: This will be graded, Bill. And if you fail to pass, we'll hold you after class.

SPEAKER: Detention.

MR. SEASTROM: So the first question is should a transfer policy have an expiration date?

That would be to say two, three, four years after its adopted.

MR. MANNING: Does anybody need one?

MR. SEASTROM: So I'd like a show of hands for yes, that a transfer policy should have an expiration date. (11 hands raised)

MR. JIMMERSON: I have a count.

MR. SEASTROM: Okay, thank you. Show of hands for no, that a policy -- that a transfer policy should not have an expiration date. (4 hands raised)

MR. SEASTROM: Thank you. Mr. Hughes?

MR. HUGHES: Just real quick on that particular subject. I'm not sure that it matters because at some point we're going to exhaust the v4 table capacity anyway. So having an expiration date four years down the road is kind of mute.

MR. SEASTROM: Yes. Okay, so the second question. Should a transfer policy allow transfer of Legacy space, which would be to say space that was issued before the RIRs were created, circa 1997?

Ordinary allocations, that would be space from a RIR? Or both?

Legacy only space. I'd like to see a show of hands please. Legacy only.

SPEAKER: Can you repeat the question?

MR. SEASTROM: Should a transfer policy -- this is repeating the question -- should a transfer policy allow transfer of Legacy space? That would be pre-1997, pre-RIR space? Or ordinary allocations? That would be RIR issued space. That would be ARIN, RIPE, LACNIC, APNIC, AFRINFC. Or both?

For Legacy space only, a show of hands please.

MR. JIMMERSON: I don't see any hands for Legacy.

MR. SEASTROM: Okay, no hands for Legacy space. For ordinary allocations only, a show of hands please. (2 hands raised)

MR. JIMMERSON: I've got a count.

MR. SEASTROM: Thank you. For both Legacy space and ordinary allocations, a show of hands please. (12 hands raised)

MR. JIMMERSON: I've got it.

MR. SEASTROM: Question No. 3: If an address holder wants to transfer part of their Legacy address holdings, should they be required to sign a Legacy registration services agreement?

Briefly, a registration services agreement is an agreement with ARIN that formalizes the relationship between ARIN and the address space holder. And anybody who has space that has been issued by ARIN is already a signatory to one of those. There has been a Legacy registration services agreement in two versions so far, and subject to further evolution I'm sure for people -- for organizations who have address space that was pre-RIR that wished to formalize their arrangement with ARIN.

So the question is if an address holder wants to transfer part of their Legacy address holdings, should they be required to sign a Legacy RSA -- that would be to formalize their relationship with ARIN first?

If the answer is yes, they should be required to sign a Legacy RSA, please raise your hand. (10 hands raised)

MR. JIMMERSON: I have a count.

MR. SEASTROM: Thank you.Can you use the microphone, please?

MR. SAMUELS: My concern with the question is that I don't know what the Legacy RSA entails now. Is there a different -- is there a qualitative difference between the Legacy RSA and the regular RSA?

MR. SEASTROM: Yes, there is.

MR. SAMUELS: What is the difference?

MR. SEASTROM: Well, there are many of them and we could have a discussion -- a sidebar discussion on those parts right now. But if Ray can go over it briefly, I'd appreciate it.

MR. PLZAK: A quick summary, Carlton. First of all, it has a fee associated with it which is basically there to make sure we can keep in contact with the person that signs anything. So it's $100 a year. The second thing is that it provides a guarantee to the person signing this RSA that any future policies that may be adopted in the ARIN region which would affect these particular holdings would not have an effect on them. Those are probably the two biggest.

It does include a provision for ARIN to be able to revoke the space if the person does bad things with it, and is so found to do those bad things by a court of law. So those are probably the biggest things to it.

MR. SAMUELS: Can I then ask for the question to be put to the vote?

MR. SEASTROM: Sure.

MR. PLZAK: Go ahead. He wants to hear the question again.

MR. SEASTROM: Okay, so the question one more time is if an address holder wants to transfer part of their Legacy address holdings, should they be required to sign a Legacy RSA and formalize their arrangement with ARIN first?

If the answer is yes, please raise your hand. This is a recount. (14 hands raised)

MR. SEASTROM: Mr. Hannigan?

MR. HANNIGAN: I'd just like to qualify that with --

MR. JIMMERSON: I have a count.

MR. HANNIGAN: I'd say yes, but not the current RSA. I'd like to just say a RSA.

MR. SEASTROM: Let me speak to that as a holder of Legacy space who has not yet signed the RSA. I agree with you. And I believe that it is incumbent upon all ARIN members who are dissatisfied with the current RSA, just as anybody who is dissatisfied with any aspect of ARIN governance, to work diligently towards making it better.

MR. HANNIGAN: I'm not making a statement about the existing RSA. I'm just qualifying my vote to say what I think --

MR. SEASTROM: I agree with you 100 percent. If an address holder wants to transfer part of their Legacy address holdings, should they not be required to sign a Legacy RSA? This is no, they should not be required to sign a Legacy RSA.

If you agree with that statement, please raise your hand. (1 hand raised)

MR. JIMMERSON: I see one.

MR. SEASTROM: Okay.

MR. MANNING: Let me explain why.

MR. SEASTROM: Okay.

MR. MANNING: This is Mr. Manning. And I believe that it is perfectly reasonable if you like the existing registration services agreement -- not the Legacy one -- and I wish to bring my Legacy address spaces under ARIN administration, if you will. With the current regular RSA, I ought to be able to do that.

MR. SEASTROM: So you would reword the question to say a standard RSA or Legacy RSA?

MR. MANNING: Kind of like Martin, but yeah.

MR. SEASTROM: Can you put a note on that, Richard?

MR. JIMMERSON: Yes.

MR. SEASTROM: Current or Legacy RSA. Standard or Legacy.Okay, Question 4: How long must a recipient of space under the policy hold it before the space can be transferred to another party? The intent of this sort of restriction was to eliminate speculation and eliminate flipping space back and forth for the purposes of just raising money. We wanted to turn it into just legitimate transfers between organizations that actually had actual space and organizations that needed extra space.

So we have four possible answers to this question: Two years, one year, six months, and no time restriction. So this is for the minimum amount of time that a recipient must hang onto the space before they can transfer it again.

If you believe the correct answer is two years, please raise your hand. (5 hands raised)

MR. JIMMERSON: I have a count.

MR. SEASTROM: Thank you. If you believe the answer is one year to hold onto address space, please raise your hand. (4 hands raised)

MR. JIMMERSON: I have a count.

MR. SEASTROM: Thank you. If you believe that six months is a more appropriate time to have to hold onto address space before you can offer it up for transfer again, please raise your hand. (No hands raised)

MR. JIMMERSON: I see no hands.

MR. SEASTROM: No hands. Lastly, if you believe that no time restriction is more appropriate, please raise your hand. (2 hands raised)

MR. JIMMERSON: I have a count.

MR. SEASTROM: Thank you.

Question No. 5: Should a policy require prequalification of address holders wishing to transfer space? That is to say should an address holder, not a recipient, need to pre-verify with ARIN that all is in order with their address space, meaning verify that they actually are the ones to whom it was issued, et cetera, before they can offer it up for transfer under this policy proposal?

If you believe that the answer to that is yes, please raise your hand. (15 hands raised)

MR. JIMMERSON: I have a count.

MR. SEASTROM: Thank you. If you believe the answer is no, an address holder wishing to transfer their space should not be required to prequalify with ARIN, please raise your hand. (No hands raised)

MR. JIMMERSON: I see no hands.

MR. SEASTROM: No hands. Thank you.

Question No. 6: Should a policy require prequalification of address need for requesters wishing to receive space via the transfer policy? This would represent no change from our current situation where an end site or an ISP requesting space must show need before they're eligible to receive space from ARIN.

If you believe the answer is yes, a requestor should be required to show need for space, prequalification, before they are eligible to receive space, please raise your hand. (13 hands raised)

MR. SEASTROM: If you believe the answer is no, that there should be no requirement to show need or other prequalification before an organization can receive address space under this policy proposal, please raise your hand. (1 hand raised)

MR. JIMMERSON: I see one.

MR. SEASTROM: Thank you.

Question No. 7: Should ARIN staff be given discretion over limited de-aggregation of blocks to be transferred in order to ensure an adequate supply of smaller blocks for those requesting space? To clarify, an organization that receives space can deaggregate it to whatever degree they wish to that they can get their upstream ISPs to carry, but this would allow de-aggregation of a block of addresses so that it could have the pieces assigned to different organizations, thus ensuring an adequate supply of smaller blocks for organizations that could not qualify for larger blocks.

If you agree that ARIN staff should be given discretion over limited de-aggregation of blocks that are offered to be transferred in order to ensure an adequate supply of smaller blocks for those requesting space, please raise your hand now. (6 hands raised)

MR. JIMMERSON: I count six.

MR. SEASTROM: Thank you. If you disagree -- if you agree that ARIN staff should not be given discretion to allow limited de-aggregation of address space, that is to say that a block that is offered must be transferred in the size that it is offered only, please raise your hand now. (No hands raised)

MR. JIMMERSON: I see no hands.

MR. SEASTROM: No hands? Thank you.

Question 8: Should a current holder be allowed limited de-aggregation, full de-aggregation down to a /22, or no de-aggregation at all? This is slightly different from the previous question. This would be a holder who was attempting to transfer space, deciding up front that they are going to chop up space in a certain way when they offer it into the address transfer marketplace.

If you believe that an address holder should have some degree of de-aggregation, please raise your hand now. (9 hands raised)

MR. SEASTROM: If you believe that there should be no limits on possible de-aggregation, that it should be at the discretion of the holder who is offering that block for sale, please raise you hand now? (0 hands raised)

MR. JIMMERSON: I see no hands.

MR. SEASTROM: Thank you. If you believe that there should be no de-aggregation allowed, that blocks should only be transferred in the size that they are currently issued, please raise you hand now? (0 hands raised)

MR. JIMMERSON: I see no hands.

MR. SEASTROM: Thank you.

Question 9: Should a requestor be limited in how frequently they can request new address space? For example, in order to limit de-aggregation. If the answer is yes, that you agree a requestor should be limited, please raise you hand now?

MR. JIMMERSON: Please raise you hands high.

MR. SEASTROM: Should I re-read the question?

SPEAKERS: Yes.

MR. SEASTROM: I don't see many hands. Okay, let me explain this question a little bit further.

There was concern that organizations might be coming back as often as -- say, for instance, once a week saying we need more address space, we need more address space, we need more address space. And getting whatever they could, particularly in the case of a large user, that could qualify for a large block of address space.

The problem is if that large user were able to come back once a week and get address space for two or three months, that would involve a new route in the routing table for every new piece of address space they got as opposed to making that large provider wait until an appropriately sized address block came on the market, which would only add one route to the routing table.So the decision would be, of course, up to the large provider. They could come back, they could say, well, if we can only get a /18 instead of a /16 now, we will take that and make due. But they would not be able to come back for a certain amount of time if that was put into this proposal.

On the other hand, if they were completely free to come back as often as they wanted, they could wait for a /16 to become available or they could get several /18s with the attendant addition to the routing table size.

So if you agree that requestors should be limited, which would create pressure to have the right size block assigned to them as opposed to a collection of smaller blocks, please raise your hands.

MR. JIMMERSON: Hands high, please. One, two, three, four, all right. (4 hands raised)

MR. SEASTROM: If you believe that there should not be a limit in how frequently an organization can come back to ask for more IP address space, please raise you hand now? (8 hands raised)

MR. SAMUELS: Can I just make a clarification?

MR. SEASTROM: Sure, sure. Absolutely.

MR. SAMUELS: I am voting this way because, to me, it doesn't matter how many times they come back -- what is whether or not they going to be granted? And every time that they come, they have to present a business case for need. Now, you don't have to accept the case --

MR. SEASTROM: But the business case might be we didn't get enough last time.

MR. SAMUELS: Oh, that's not a good business case, definitely. That's not good.

MR. SEASTROM: Okay.

Question 10: Should ARIN operate a voluntary listing service to help qualified address holders and requestors find each other?

Does anybody want clarification on that? This would be like a stockbroker, where you go -- who facilitates the trades in the middle between somebody who wants to sell or buy. Or like a car dealer, you know, that if you want to buy a car, you can find a place where there are plenty of cars for sale by going to your car dealer.

So the question is: Should ARIN operate a voluntary listing service to help qualified address holders and requestors find each other? If you believe the answer is yes, please raise your hand? (12 hands raised)

MR. JIMMERSON: I have a count.

MR. SEASTROM: Thank you.

Question 11: Given a policy of --

MR. JIMMERSON: You didn't do --

MR. SEASTROM: I'm sorry, no. I apologize.

Question 10: Should ARIN operate a voluntary listing service to help qualified address holders and requesters find each other? If you believe the answer is no, please raise you hand.

MR. JIMMERSON: I don't see any hands. (0 hands raised)

MR. SEASTROM: Okay, no hands. Thank you.

Question 11: Given a policy proposal consistent with your answers, that is, say, the answers that you gave above, would you support a liberalized transfer policy? If the answer is yes, please raise your hand. Nice and high.

MR. JIMMERSON: Raise your hands high, I'm almost done counting -- five. One, two, three, four, five.

MR. SEASTROM: This is, if things went the way that you voted, would you support a transfer policy? (6 hands raised)

MR. SEASTROM: Thank you. If there was a policy proposal consistent with your answers above, would you not support it? If the answer is no, I would not support a liberalized transfer policy, please raise your hand now? (1 hand raised)

MR. JIMMERSON: I count one.

MR. SEASTROM: Thank you.

I have one more question that was submitted to me from the audience, which is: If you took the survey online as part of the PPML survey that went out last month, where these questions were put to ARIN PPML readers, please raise your hand. (3 hands raised)

MR. JIMMERSON: I see two people who took the survey online.

MR. SEASTROM: Nobody else took the survey online? We've got three. We forgot Lea.

MR. JIMMERSON: Lea? Okay, three.

MR. SEASTROM: Okay. Thank you very much. I'm going to step back here for a moment.

MR. BICKNELL: R.S., if I may, since Bill and Marty brought up the same point, I did a quick bit of research on Question No. 3, where they both commented on any RSA. And the standard RSA qualifying -- and I realized that the survey question may slightly mislead what was in the policy.

In the policy text, on page 5, the proposed language right now says, "The transferor has signed an RSA, and/or a Legacy RSA, covering the IPv4 addresses transferred." And it was written that way because anybody with ARIN space already has an RSA by definition. So the question that the AC was probing was whether the "and/or a Legacy RSA" should be included in there or not? But if you just read the question in the survey, you might think that the RSA was precluded from the policy as proposed.

So I wanted to see if the language on page 5 made the two of them happy and/or anybody else, unhappy?

Bill gave me a thumbs up.

2008-6: Emergency Transfer Policy for IPv4 Addresses

MR. PLZAK: Now to Proposal 2008-6, which is related to this. We still sent a form of proposal less than a month ago -- the correct version is as of the 15th of August -- that's under discussion.

Similar type of proposal in the APNIC and RIPE NCC in that it relates to transfer policies in general. It says that transfers will be allowed for three years. Recipient must document the need. Original prefix may not be de-aggregated to more than four pieces, each greater than or equal to the current minimum prefix size.

And so I'll turn it back to R.S.

MR. SEASTROM: Thank you, Ray. So if you'll turn with me to page 11 in the same packet, this is the text of the Policy Proposal 2008-6. If you are thinking about writing an ARIN policy proposal, a policy proposal that looks a lot more like this one in terms of size versus 2008-2, is a lot more likely to get a consensus because it is short and concise and only covers a few things.

This policy proposal is of a limited time, three years; allows transfer from somebody who's shown that they're legitimate, which means that they have to pre-qualify; takes place to a recipient that has documented need; and puts a limit on de-aggregation. Does not put pretty much any other restrictions on the transfer. Does anybody have any comments on this proposal?

I'd like to add that the author of this proposal, Bill Darte, is on the ARIN Advisory Council and was concerned, and probably rightly so, that we would never get to consensus on 2008-2. So he figured he would offer up something that was sharp and concise and maybe we could get consensus on.

Stacy?

MS. HUGHES: Stacy Hughes, TW Telecom. And I support 2008-6 over 2008-2 for that very reason. I think we should stick with simplicity and I think we shouldn't micro-manage.

MR. SEASTROM: Thank you. Leo?

MR. BICKNELL: Leo Bicknell. Specifically regarding .03 of the policy statement that limits de-aggregation to no more than four resultant blocks. I believe that if you consider a host IPv4 run out world that, as written in this policy, were to be adopted, would not be an effective restriction.

Because what you will find once IPv4 runs out is there are plenty of companies that will have need because there will be no more to get. So the number of people having need will grow over time. And that some of those companies will find that buying an address block, chopping it up into 4 bits, and selling it immediately back out to other people with need is more profitable than using it for their own need.

So you have an entire set of people who have need, but because there's no limits on hold time or anything like that, you just create a situation where some middleman's making profit and the limit doesn't actually end up being an effective limit.

MR. SEASTROM: Okay. Bill?

MR. MANNING: Bill Manning. How many people in the room come from the telecoms background?

One, two, three, four, five, six, seven -- I'm not going to count you guys back in the corner. All of us, more or less, are comfortable and familiar with the concept of the North American Numbering Plan. And about eight years ago or so, when you went and got a block of numbers, how big was the block?

SPEAKERS: Ten thousand.

MR. MANNING: Ten thousand. How big is it now?

SPEAKER: About 1,000.

MR. MANNING: It's 1,000, right? So somewhere along the line in the North American Numbering Plan, we went through this exercise where we started to find out that numbers were scarce, and so the size of the block that people got, got smaller. And what we're talking about in these two proposals talks about taking perhaps large blocks and shrinking the size of the block. Both this 2008-6 and 2008-2 talk about the current minimum allocation size, which today is /22, which is a thousand addresses. That may be too large.

And so I'm concerned with either of these two policies in that the blocks that we're talking about may be too big.

MR. SEASTROM: The policies both refer to the current minimum allocation. And I agree that the proper thing to do may be to address that minimum allocation and that the right time to do that might be now, before address roll run out.

MR. MANNING: Somebody want to help me write a policy?

SPEAKER: No.

MR. MANNING: Thanks, Mark.

MR. SEASTROM: Anybody else have any comments?

MR. PLZAK: So let's take a quick show of hands about this proposal.

Oh, sorry, Nigel?

MR. CASSIMIRE: R.S., could you help me again? I'm just trying to grasp fully the difference between this paired down version and the earlier one. What does this one not have that the earlier one had? Do you understand?

MR. SEASTROM: One thing it does not have is a requirement that people keep addresses for a certain period of time. Another thing is, it is not micro-managed, what counts as documentation. It does not require documentation if you sell the organization in the case of mergers and acquisitions. Let's see, it does not -- if you take a look, especially in --

MR. CASSIMIRE: It is not on a distinct service?

MR. SEASTROM: 8.2 -- actually, all of 8.2, there's all sorts of little concerns that individuals had, like 8.3.7. Safe Harbor. It was concerned that ARIN would take back addresses if you were offering them for transfer on the grounds that clearly you didn't need them, so you shouldn't have them, so we'll take them back. So 8.3.7. was specifically to deal with that kind of concern.

Basically, 2008-2 is what we call an omnibus proposal that includes a roll-up of all sorts of people's concerns and little requirements.

And I think 2008-6 tried to simplify that and make it just down to the bare minimum that's required to have some sort of transfer mechanism in place.

MR. PLZAK: Okay. One more?

MR. CASSIMIRE: It's a kind of an aside question. Has ARIN tried or had any success in getting Legacy space back from people -- try to get Legacy space? Have you had any success with that? And especially given the imminent run out of IPv4.

MR. PLZAK: Okay.

MR. SEASTROM: There has been substantial effort on the part of ICANN to get the space back. I think Lea can probably speak to that.

MR. PLZAK: I'll take care of it.

MR. SEASTROM: Okay.

MR. PLZAK: Speaking specifically for ARIN -- and also I wanted to make a general comment about the impact of recovery of Legacy address space upon extending the life of IPv4 and what it does to possibly impact on deployment of IPv6 or the implementation of IPv6.

First of all, there is no active going out and hitting people over the head and saying if you're not using your IPv4 address space, give it back, with one possible exception. And that is when we made the allocation of IPv6 address space to the U.S. Department of Defense, that we wrote and the U.S. Department of Defense agreed to, they agreed in that RSA to return Legacy address space to ARIN as they converted those blocks over to IPv6. That notwithstanding, that is the only real active thing that's been going on.

Now, have people been giving address space back to us? Yes, some have, but that's been here and there. In addition of looking at it from the perspective of what does it do?

And this is me speaking personally, I don't see, given the rate of consumption of IPv4 space and even if we were to recover two /8s a year in the ARIN region, what it would do would --

MR. SEASTROM: That's two months of world usage.

MR. PLZAK: It's two months of world usage. You know, it doesn't do much. On the other hand, an active program like this where you're actually getting it back might cause people who are getting ready to do IPv6 to say, well, I guess this could wait a little while longer.

So if nothing else, it would push that bar a little further out for getting IPv6 space or to implement IPv6 space. So like I said, we have got some back. It's not been a significant amount and it has been incidental to people saying I don't need it anymore.

Yeah, Lea?

MS. ROBERTS: Lea Roberts, Stanford University. As R.S. alluded to, we did give back our /8, but that was back in the days that the Internet was, shall we say, a more cooperative kind of focused institution. I'm personally aware of some of the commercial people who have /8s, not necessarily in very much use, and I think their business people are very much looking forward to being able to sell the space rather than giving it back.

MR. SEASTROM: Can I ask a question of you? And I'm sorry for putting you on the spot.

MS. ROBERTS: Sure.

MR. SEASTROM: But I know it was kind of difficult to get it through at Stanford when you did it at the time. Do you think you'd have any chance of doing that today, in today's climate?

MS. ROBERTS: Do you mean to do the space thing?

MR. SEASTROM: Having the notion that we ought to renumber out of 36 /8 and give it back?

MS. ROBERTS: It's hard to say. I mean, probably, I think most -- although I'm not sure Stanford -- you know, would be concerned about the monetary thing. Although I do hope that I'm retired before they find out how much it might have been worth.

MR. SEASTROM: Okay, thank you.

MS. ROBERTS: Thanks.

MR. PLZAK: I'm going to ask two questions. The first question I'm going to ask is a yes or no on people in the room that are in favor of this proposal. And then after that, I'm going to ask a second question, which is are you in favor of this over 2008-2, the broader omnibus transfer proposal?

So two questions. First one, are you in favor of this proposal? And second one, are you in favor of this over 2008-2?

So first question, show of hands. All those that are in favor of this policy proposal, please raise your hand. (6 hands raised)

MR. JIMMERSON: I have a count of six people.

MR. PLZAK: And how many in the room are opposed to this policy proposal, please raise your hand? (6 hands raised)

MR. JIMMERSON: Hold your hands high, please. One, two, three, four, five, six. Thank you.

MR. PLZAK: Now, regardless of how you voted on the previous question, how many in this room are in favor of this proposal over 2008-2, the previous proposal we discussed? (9 hands raised)

MR. PLZAK: How many are -- regardless of which way you voted on the previous question, are not in favor of this proposal over 2008-2? (3 hands raised)

MR. JIMMERSON: I've got a count.

MR. PLZAK: In regards to the proposal being in favor: 6 yes, 6 no. That pretty much says no consensus at all.

And in favor of wanting to do this one instead of 2008-2, in other words, over it: 9 yes and 3 no. Thank you very much. This information will be provided to the Advisory Council.

MR. SEASTROM: Thank you, Ray.

MR. PLZAK: We'll begin one more proposal and see how we do. And that is 7? 2008-7?

SPEAKER: Yes, that's correct.

2008-7: Whois Integrity Policy Proposal

MR. PLZAK: This is Policy Proposal 2008-7, entitled WHOIS Integrity Policy Proposal, another recent proposal, and the current version is dated the 15th of August, that's under discussion. There is a similar proposal to this in the APNIC region, which has been adopted, and it's not even anything like this under consideration in the other three regions.

Basically, the summary of this proposal is that a resource must be covered by a valid RSA before it can be updated. In other words, before you can update a record in the WHOIS directory, you must have an RSA with ARIN.

And so with that, I'd like to ask Leo Bicknell to come up here and lead the discussion -- present the proposal.

MR. BICKNELL: Hello, my name is Leo Bicknell again. So this proposal's fairly self-explanatory. The one thing that I want to caution people on about this proposal -- and it's not so much about the proposal -- is there are, I believe now, four different WHOIS-related proposals being talked about on PPML. Several of the others haven't been fully considered by the AC yet, so they don't have policy numbers. So if you're wading into discussion on the mailing list, in particular, you will see an awful lot of subjects that have to do with WHOIS, and they don't all necessarily apply to this specific policy. So you need to kind of be careful as you read it.

But I do believe in this case, Heather was inspired by the APNIC policy that they passed. Their situation is a little different from ARIN's, in that to update records with APNIC there are various website log-ins and cryptographic controls involved that they, more or less, force you to use. It's -- you know, use the automated system. And so part of it was in an effort to secure access to changes in that region. And you can find the text of the policy on their website.

But so the understanding today is, if you have a Legacy net block and you've never had any other involvement with ARIN, and you move or change phone numbers or otherwise need to update your WHOIS information, you can send an e-mail to ARIN hostmaster, and they may or may not ask you some questions, and then they will process an update to your Legacy WHOIS record.

And Heather believes that for that to happen, there should be a formal relationship between the two parties in the form of a Registration Services Agreement, which she leaves open here to include any Legacy or traditional RSA that exists now or may exist in the future.

So any questions or comments, whether this is good or bad?

MR. HANNIGAN: Martin Hannigan, NRO. So I support this policy, notwithstanding the issue I raised related to the RSA. Let's just say that's independent for now. But all this transfer activity that's going to take place eventually is going to cost money, and we need to recover our costs and this is the only way to do it. And that's why I support it.

MS. HUGHES: Stacy Hughes, TRA Telecom and ARIN AC. I support this policy because if you're asking ARIN to do something for you, you should have a contract with ARIN.

MR. SEASTROM: I'm Rob Seastrom. I'm on the ARIN Advisory Council. I have Legacy space that is issued to me. I have a Legacy ASN that is issued to me. I have not signed a Legacy RSA yet, although I intend to. And I intend to continue working vigorously toward having one that is more palatable.

However, should I move, which I intend to do this fall, it would indeed be very annoying if I were not able to update my contact information on that net block and that ASN and that org ID, without signing a Legacy RSA. And so I oppose this proposal.

MR. BICKNELL: Anyone else have any thoughts on whether it's good or bad? Questions about it?

MR. WALLACE: Yes, I have a question. Suppose it's a -- Romel Wallace, Bureau of Information and Technology, Virgin Islands. I have a question. Suppose it's implemented -- I read -- was reading, it says something about 60 days, you contact the person -- point of contact via e-mail address. Does that mean if the person doesn't respond at some point in time, that ARIN can just reclaim that space?

MR. BICKNELL: No, with this proposal, there has been a suggestion that a timetable for implementation and its -- and right now it's just an estimate because the Board and staff would make the final determination on timetables after this was adopted, were it adopted -- is that it could be implemented within 60 days of the policy passing.

And that to be fair to all of the Legacy RSA holders -- sorry, to be fair to all Legacy space holders -- since this is a change to how they interact with ARIN, that they should all receive an e-mail notification of the change. There is nothing in here that if that notification bounces or doesn't go out or anything like that, that anything would happen.

This is similar to a notice of a building being constructed being posted in the town square. It's just a method of letting people know that the rules have changed. But this proposal would take no action were those e-mails to come back or bounce or people to reply in any specific way. It's purely a one-time notice.

MS. ROBERTS: I'm Lea Roberts from Stanford University and on the ARIN Advisory Council. I don't happen to have any personal Legacy space, but certainly Stanford does.

I'm a little concerned about the message that this policy would send to Legacy holders at a time when we're trying to encourage them to come in.

It seems like we're pulling out the sticks a lot sooner than we've been pulling out the carrots.

So I oppose this policy.

MR. BICKNELL: If I may ask you the same question that I've personally asked on the mailing list several times, what carrots could we offer? Because you're right, there have been many, many sticks and I've seen almost no carrots even proposed.

MS. ROBERTS: Well, I think that the transfer proposal, if it passes, is the largest possible carrot. I mean, the opportunity to money for Legacy space is something that I'm very concerned about, but I think -- anyway, I think -- given the tone of a lot of Legacy people on PPML and so forth, I'm concerned about the fact that the Legacy RSA still really does have some issues with the legal side of it until there's like a really clear consensus that the Legacy RSA is a no-brainer.

That only people -- you know, that people are only holding out because they're trying to prove a point or something, then it's just too early.

I mean, the carrot, as far as I'm concerned, would be continuing to tune up the Legacy RSA to remove those hurdles that people see in signing it.

MR. BICKNELL: Okay, Bill?

MR. MANNING: So Bill Manning, holder of Legacy space and signature to Legacy RSA for said space. I'm not comfortable with this proposal for some of the justifications that -- on the existence of WHOIS and other things that Rob talked about earlier.

If you want accurate information in the ARIN databases, and you pass this proposal and someone who holds Legacy space -- someone not currently covered by an RSA -- and they move, they change, whatever, the Legacy holder doesn't have a recourse, per se, to actually go in and actually keep that information up to date because ARIN has removed an avenue of communication, unless the Legacy holder agrees to whatever terms and conditions are in the RSA that they find objectionable.

So I'm not comfortable with this policy. I think this will actually increase the entropy in the database before it will improve it. And for that reason, I'm not real happy about this, but I'm not the community.

MR. BICKNELL: But you are the community.

MR. MANNING: No, I'm not. I'm a trustee.

MR. BICKNELL: You're still part of the community. We let you come to the meetings.

Marty?

MR. HANNIGAN: Martin Hannigan again. So now, all of this transfer stuff is making all this other stuff really complicated because it's all kind of intertwined. So I can understand what Bill's saying, but if you want to play in the transfer process game, then eventually, in my opinion, it's going to turn into some kind of market where some people are going to end up with windfalls, you got to pay. And I don't know what order we should be looking at these policies. I think that some of them may be a little premature. Some of them may need to be readdressed after we figure out exactly what we are going to do. But I think the bottom line is you got to pay to play, and I still support this policy.

MR. SEASTROM: Rob Seastrom again. I am very much in agreement with the points that Bill just made. And as somebody who pays ARIN membership in his day job, which is to be separated from the Legacy address space that I hold personally, it is very much in my interest to have valid and working WHOIS. And part of my money that I spend as part of my dues is supposed to go towards that. So anything that puts a step away from that is as a DMR from my organization that is actually a member, I can't get --MR. BICKNELL: DMR?

MR. SEASTROM: Designated member representative, even though that isn't quite -- I'm not voting for the Board or for the Advisory Council here, but I'm saying with my hat on as a representative of a member as opposed to as personally.

MR. BICKNELL: That's okay. I thought you actually said DNR.

MR. SEASTROM: No, no.

MR. BICKNELL: Do not resuscitate. (Laughter)

MR. SEASTROM: That would not be me.

MR. BICKNELL: Lea?

MS. ROBERTS: Lea Roberts, Stanford University again. I guess, let's see, what caused me to get up this time is I think this is, again, just a little restrictive. As R.S. just referred to, it's possible for someone to have a relationship with ARIN, to be a member. I actually got Stanford to be a member even though our Legacy resources aren't under an RSA. So you know, we're paying a lot more than we would have to pay for the LRSA to be a member of ARIN. And so perhaps, this could be expanded to include -- you know, some recognition of membership as a relationship with ARIN as opposed to having an RSA.

MR. BICKNELL: That's a good suggestion. Anyone else have any other comments?

MR. CASSIMIRE: Nigel Cassimire. I'm not a Legacy holder or anything. But I was sitting there and I was saying to myself maybe we should let the Legacy holders speak for themselves, and I think that they did. But I agree most with what I heard Bill say about the possibility of unintended consequences from this.

Because it's a nice show-up policy and everything, and the name of it is WHOIS Integrity Proposal Policy, and it once said it intended to improve the integrity of the WHOIS database, but I'm not sure that the recourse is the right thing in order to do that. And it comes across to me as more of an incentive to sign an RSA rather than ensure integrity of WHOIS. And I think I like what Bill's recent suggestion was that maybe it should in fact make some reference to some other legitimate recommendation qualification rather than forcing people to sign an RSA.

MR. BICKNELL: You know, reading the APNIC policy, whose name I forget, but it's pretty easy to locate in their archives, may provide a little bit of history as to why this one's titled the way it is at a really high level because I don't know the specifics of interacting with APNIC. One of the impetus for that proposal was APNIC was accepting e-mail templates and authenticating them based on e-mail address. The mail from authentication as many people call it.

And as I'm sure many of you are aware, that is a less than reliable way of ensuring that somebody is who they say they are. They actually had some problems -- not necessarily with WHOIS updates -- of people abusing that and sending in invalid updates. So APNIC moved across the board to disable mail from authentication and require PGP keys, s/MIME certificates, I don't know the full spectrum. They have more than one way you can authenticate yourself. I believe there's three or four methods available, and require that of everyone who does business with them.

And some Legacy holders sit up in the back of the room and said we don't do any business with you. The only way we've had to update is mail from, so when you take it away, we will never be able to update our records again. And so APNIC passed a policy that basically said, fine, you're just going to sign our standard contract and get a PGP key or an s/MIME certificate like everybody else.And they went, okay, in that region. So there, there was a lot about security of updates regarding the mail from. So the policy proposal there is something like update security or something if you go looking for it. I forget the exact text.

And again, that was the inspiration here, so that's where I think you see the integrity coming from. But I think Bill's point about whether or not in this situation it accomplishes integrity are quite valid.

See, I agreed with Bill.

It happens.

MR. PLZAK: So anybody else? Any other comments one way or the other or clarifications needed?

Mr. Hannigan?

MR. HANNIGAN: I just wanted to point out that I am not a Legacy resource holder in any way, shape, or form. And I just want to make it clear that my view of this, actually, is that WHOIS is only going to get better if there is any kind of transfer policy because you're going to have to -- and if there is any kind of value added to -- you know, assigned to IP address because you're going to have to have clear ownership. And the only way to have clear title to the resource is to authenticate yourself, and it's going to end up in WHOIS, anyway.

I really don't think this is a big problem. Thanks.

MR. PLZAK: All right. Time for a show of hands. I'm going to ask two questions. First question is, how many people are in favor of some kind of policy like this? And then the second one will be, how many people are in favor of this specific policy?

So first question will be how many people are in favor of some kind of policy like this? And then the second one will be how many people are specifically in favor of this proposal?

So Question No. 1: How many people are in favor of some kind of policy like this?

Please raise your hands. (19 hands raised)

MR. JIMMERSON: We have a count.

MR. PLZAK: And how many are opposed? (0 hands raised)

MR. JIMMERSON: I see zero hands.

MR. PLZAK: As to the specific proposal, how many people in the room are in favor of this specific proposal? Please raise your hand.

MR. JIMMERSON: Raise your hands a little longer.

MR. PLZAK: One more time, please. (5 hands raised)

MR. JIMMERSON: Okay.

MR. PLZAK: How many people are opposed to this specific proposal? (5 hands raised)

MR. JIMMERSON: I got a count.

MR. PLZAK: There were a large number of you that said we should do something like this.

So for those of you that said we should do something like this, would you mind stepping up to the mic and saying what kinds of things you would like to see included in this proposal? Particularly since the fact that so few of you are in favor of doing this specific proposal.

MR. HANNIGAN: I've had enough, thanks.

MR. PLZAK: Okay.

Well, if no one has anything specific to say, I guess the AC is going to have to rely upon the discussion that was in this meeting and discussion on the list to try and figure out what kind of language needs to be there and they can work with the author on that.

So the information from the show of hands will be provided to the AC.

Thank you very much.SPEAKER: Thanks, Leo.

MR. BICKNELL: Sure.

MR. PLZAK: Okay, let's take our lunch break.

It's going to be the same as it was yesterday. You get it out in the hallway and you go next door. (Whereupon, at approximately 12:38 p.m., a luncheon recess was taken.)

Internet Exchange Points

MR. PLZAK: No, I'm not Carlton Samuels and I don't pretend to be him on television, but Carlton is going to be the first one up this afternoon to give a short presentation -- not that short really -- but on Internet exchange points for the Caribbean.

Before he starts, I'd like to point out that there is a policy in the ARIN region that if you are operating an Internet exchange point, you can get an allocation from ARIN specifically for the operation of that. Also, it's the same policy that applies to people that happen to be operating ccTLD addresses for the servers. So this is where you can get provider independent space directly from ARIN.

And as we all know, Internet Exchange Points are very important for the development of the Internet, so I'm sure that Carlton's got a lot of good things to tell us.

So Carlton, the show is yours. And we won't mention to ARIN that if he was in here. He would have had a chance to win $100.

MR. SAMUELS: Thank you, Ray. What we have to say about this Internet exchange point idea -- give you some background as to why it became an issue for us and why we are seeking the support of the community to see if we can move the idea a little bit forward.

Internet exchange point. It's a physical infrastructure, for those of you who might not know what it is, that allows multiple high speeds to use peering arrangements that exchange Internet traffic between their autonomous systems.

Let me give you a background as to why it is important to us. I, in another life, I wore a hat as a university director of IT for the University of West Indies.

I now only teach. But at that time we had some problems.

The UWI is an ISP. We had probably the first ISP in the region. And we had bandwidth provided, at the time, was only cable and wireless and we had two providers, and so we were doing a lot of online courses and our students have multiple ISPs to deal with. And what we were finding out was that we were getting a lot of the bandwidth, we were flat-lining bandwidth, inbound and outbound, by 10 in the morning and we just couldn't afford more bandwidth, so we started looking at ways to reduce the bandwidth.

And then we decided we'd go to the providers and work with them to pair with us to do BGP, to keep all the traffic that we had local, and we saw dramatic improvement, so we started expanding the idea. We presented it to the Caribbean Internet Forum in several areas to see whether or not we could -- if that would work for the University of West Indies. We figured that there has to be something there for the regular Internet users and it was a way to encourage the ISPs to being able to work together in their own interests and conglomerate against the large providers of bandwidth uplink, bandwidth. There were only, as you would imagine in the Caribbean at the time, one or two of them.

So we started exploring the idea and we looked at some of the benefits that we had achieved from just that simple agreement with cable and wireless on the Flow. We really had a reduction in the uplink provisioning. The efficiency of the bandwidth usage was very much a part of that. We had some fault tolerance to a major corporations and of course we reduced latency. There was a lot of latency when we had local traffic.

You know, trying to get to a local school, like say UTECH, in Jamaica, the latency was as much as if you were trying to get Boulder, Colorado. So we had some definite benefits from it.

Now, we've been trying sell this idea for a while, and we are not getting as much traction as we supposed we would from the ISPs generally, and so we've been kind of going back to the drawing board to see how we could make the case. And we find that there are certain bits of data that we need that's missing, and we've been trying to ask the community -- and I've had some positive response already from members of the community -- to find out some information that we need.

For example, I don't know the definitive traffic profile for the Caribbean what traffic is moving north, what traffic is moving south, what traffic is moving between countries in the Caribbean. I don't know that, and that is a really good piece of information to know. So I've collected all the ASN numbers for the Caribbean and I'm looking for help to develop a profile, a traffic profile, for the Caribbean, going north, calling north to south, and between and among Caribbean countries.

I should tell you that a couple of months ago, I found out that there was an ISP that was based in Curacao, and we've been trying to work with them to see what exactly his experience has been. It's been there not for a very long time, so we are looking towards getting some help from that.

So this is a wish list to members of the community. We want more particular data on Caribbean data traffic patterns and we want to develop a nations document for Caribbean ISPs so that it clearly outlines what the issues are, what the downside, what the benefits are, and then even help them to create a framework business plan for an ISP.

The idea won't work without the ISPs coming together and seeing value in participating and so what we are trying to do here is to kind of sweeten the pot and help them to help themselves.

I really believe from our experience at the university, which is probably the largest user of data and provider of data in the Caribbean that an Internet exchange point would help significantly.

So that's all I have.

MR. HUGHES: Just as a suggestion, you may want to contact Terremark. Talk to Bill Williams. I know they have arbor boxes (?). You can probably get reasonable flow data about what's in and out of the Caribbean and they already have it, so it might be a good place to start.

MR. SAMUELS: Great. Thank you. Any questions? Comments?

MR. CASSIMIRE: Caribbean Net -- oh, Nigel Cassimire. Caribbean Net next meets in Trinidad at the end of October. And I know on the agenda, it is proposed to have a session about this and I've seen some names put up there. I'm trying to remember if yours was one, but I certainly remember Mark Reed's (?) name being mentioned and a few others. So I'm wondering -- do you plan to be there or are you aware of the development of that topic at this thing and --

MR. SAMUELS: Actually, they have asked me to participate. And part of my participation is to provide the profile data because that's what we don't have. We do not have the profile data.

I have a schedule conflict I'm trying to resolve, but it's very important for us to get the data, because everything that -- as you know, this has been on the agenda for the last three cycles of CIF. And we've been working our way towards getting more particular about what it is that we wish to do, the data, the profile data. Let me tell you why the profile data is important.

If you talk to some of the ISPs, they say to you, most of the traffic is from eyeballers, people who are going to people with content in the North, in the Caribbean, in North America and Europe, and they are buying things online. That's where you see the traffic, so it's more data traffic. A lot of it is traffic, and it's -- you know, it's the context against content in the North. Okay.

So if that is the case, and given how the ISPs are provisioning bandwidth these days, there really is no major incentive for them to do that. But from all perspective, users, ordinary Internet users, we are developing some content. There's some content. Content is beginning to appear in the Caribbean and specifically in content we're using for online teaching and learning and that's important to us from the tertiary higher education community.

And so we have a compelling interest in looking at this issue in greater detail because I can tell you from our experience that could help all of us. So that's why we're trying to become more particular in the data that we present. We really need the flow data to help us to see whether or not the central argument for most of the ISPs that you talk to about becoming involved, is credible.

That's a problem.

MR. CASSIMIRE: In your presentation, you appear that there were benefits that you saw, right?

MR. SAMUELS: Yes.

MR. CASSIMIRE: And what you said about -- you know, people's comments about the usage is basically filled with people going to people with content and so on. We've heard that around the Caribbean and I'm sure it is no different for Jamaica. But yet still, you saw benefits. So I mean, I think that is a point that needs to be brought out to the ISPs operating in the other Caribbean islands as well. And I think if you can quantify the sort of benefits that you saw that -- you know, people walk around with notions in their heads without actual data. Okay?

MR. SAMUELS: Let me give you one benefit that we saw immediately. We have over 25,000 students that we serve and we were using a 40 megabyte link for Internet access, and as I said to you, by 10:00 they were flat-lined inbound and outbound.

We instituted BGP with Flow and cable and wireless and our throughput improved by about 60 percent, and the pattern, when you have saturation, changed totally. We pushed it back until probably about 3:00 in the afternoon, 4:00 in the afternoon.

And most of the times when you were looking at access for online courses and all that was happening between 6:00 and 9:00 in the morning and after 7:00 in the evenings, which is when we wanted bandwidth availability.

So we have that information. It's there. We have that data. It's not enough to get our ISPs, small ISPs, because the first thing they worry about is, well, if I peer with this guy, what's it going to cost me? Because the model that you use, well, it's a net-net kind of a situation. You will only pay me if the amount of traffic you send to me is above, so it won't work for them. That model won't work very well for small ISPs in the Caribbean. It will not work. So we have to come up with a different business model to make the case even if end users are going to be beneficial -- from the arrangement. The business model that we use to sell it is going to have to be different.

So that much I'm pretty sure of. Now I just need some data to support the market. So to the members of the community, we are looking for data and any opportunity that you can find to help us, to collect the kind of data we need, we'll be very appreciative.

Thank you. (Applause)

MR. PLZAK: Thank you, Carlton. When you come back from the break, you will find on your table a copy of the survey about this meeting, and so we would ask for you to fill this out and return it out front.

We take these survey comments very seriously, and that's how we make improvements to our meetings. So please take the time to fill it out and drop it out front before you leave this afternoon, so as I said, when you come back from break and you're in here for the prize drawing, the survey will be sitting on your desk.

BoT, AC, NRO NC Activities

MR. PLZAK: The next item on the agenda is a series of reports on activities of the Board of Trustees, the Advisory Council, and the NRO Number Council, and I will do the quick report on the Board of Trustees, and Leo Bicknell will do the one on the Advisory Council, and Marty Hannigan will do it on the Number's Council.

So what has your Board been up to? A lot. Since we were here in this portion of our region, in Kingston, we've had a couple of Board workshops, a couple, maybe three Board Meetings, two face-to-face and one teleconference. In addition, we've been participating in some informal get-togethers with the members of the other RIR boards to discuss matters of mutual interest. So the Board has been quite active and the Board List has been quite active.

So some of the things that we've been discussing -- one of the things that was touched on some this morning was the RSA, the Legacy RSA, and that discussion has been going on amongst the Board as well.

Recently, we drafted a new version of the LRSA, and the Board looked at that and was basically happy with it, and that's why I was released a week ago Friday.

The Board has been monitoring the discussion of the transfer policy, has been looking at some of the implications for ARIN in terms of the structure of ARIN, the responsibilities that ARIN has in this area, the possibility of things we can do to strengthen our ability to do fraud detection. That's where people come to ARIN and try to change the record in ARIN so they can gain control of the address space by being viewed as the legitimate user of the space.

Technically, they can go out and grab it and use it, but to be able to be the person that can pretend to be the person that has the legitimate use of the space and is entitled to derive its benefits. That requires being able to talk to ARIN directly and to get us to change the database, and I will say this, that most of these folks have been doing a very, very good job in this area.And so we've been discussing this and looking at other things that we could do to possibly make it even harder for people to do that, looking at other ways for people, that as they become aware of fraudulent activity, to actually be able to somehow or other report that, and so that some actions can be taken.

The bylaws have been looked at from the standpoint where there's some duplications and some areas where they can be tightened up. We will be announcing at the end of the meeting in Los Angeles, that the Board has gone through this and will be releasing a new version of the bylaws, which basically is not a material change to the bylaws, but it is a restructuring of them so that, for example, you'll find in many places in the bylaws, the same thing applying to the Board and to the Advisory Council.

So every time you want to do something through the Board or the AC in a particular area, you have to go back and do it to the other party as well, so you have to make two changes, make two adjustments, two refinements. So we've gone through that.

If you look at the article that's miscellaneous, you'll find that it's become extremely long, so some of that stuff needs to be folded and wrapped up and those kinds of things. So it's more of an overhaul of the structure of the way the bylaws are being presented as opposed to any substantial material change. But again, look forward to that in the future.

One of the big things in the discussions this morning dealt with it as well is what is the real status of Legacy Address Base, and, more importantly, what is the authority that ARIN has as a successor of interest or derivative in some other manner over Legacy Address Base? This is becoming more and more important and actually is a matter of profound interest in other parts of the world besides the ARIN region, particularly in South America and Africa, and to a certain degree in Asia, and to a lesser extent in Europe.

As you know, the bulk of the Legacy Address Base exists in the ARIN region and so it's probably ARIN's biggest problem to deal with it.

And so that is a matter of discussion that's been going on for some time. This is a matter that the RIR Boards jointly are looking at because there needs to be some determinations made in regards to the status of this Legacy space as it exists between regions, and in the ARIN region, almost anything we do is going to require some kind of a policy action and so we need to look at, what are the policy implications, what things are going to have to be done from a policy perspective, what things will the Board have to ask the Advisory Council to look into from a policy perspective?So that's another part of the discussion.

I haven't talked about the external structure outside of ARIN at all, and I did make mention of it, and Marty is here from the Number Resource Organization. No, Martin, I'm not calling you up to the table, so you can keep on moving. But the NRO exists today as a result of an MOU. And it really is nothing more than a really loose confederation that allows us to work together. There is no budget for the NRO. There is no staff for the NRO. That may sound familiar to a few of you in this room, but there's a lot of work that gets done.

The NRO was put together to do a couple things. One is to allow the RIRs to work together jointly on such things as communication so that we can speak as one voice in international forums such as the IGF. Also to function as a collaborative type organization so that engineering efforts can be coordinated more directly such as the running of the reverse DNS, things of that nature.

Also, it allows us to act as a single focal point in dealing with folks like ICANN so that the address supporting organization MOU which establishes the ASO basically says a couple of things. One, the NRO will do the task; and two, you deal only with one person. You deal with the NRO, you don't deal with five RIRs. And then in conjunction with that another one of the objectives is to protect the bottom up policy process.

When we were negotiating the Address Support Organization MOU, ICANN wanted to inject into it the ability of the ICANN Board to make policy and we drew a line in the sand and said, no, that's not going to happen. And it did not happen.

We feel it's very, very important to continue to protect that bottom up policy process and using the NRO as a speaking voice in that regard is very important to us.

And then there are organizations like the IETF who write documents that affect, or if they want the RIRs to do something. Well, instead of writing in an RFC go to XYZ RIR or try and figure out which registry you want to contact first or whatever, a single point of contact, the NRO, so. Now, there has been some problems with the way that are organized in that some people, since it isn't really a organization with a staff and budget and so forth, is it really a real thing? Can it sign a contract, for example? What are some of the other things that require some sort of a stature?

So we have been discussing for some time the idea of incorporating the NRO and we actually moved pretty far along on this and then we kind of like dropped it about a year and a half or so ago because other things overtook it and we kind of had to put it on the back burner. We're bringing it back up, so there's a couple real interesting things about this. We still want it to be a very lightweight organization without staff, without budget, really, but we need to be able to put it together and give it an identity, a legal identity from the standpoint of the corporation.

So where are you going to do this? There are some people that would strenuously object to it being in the United States, so we basically threw that out. There are other people that would object to it being in Europe, so that went away. There in places in Asia that went away very quickly, and so finding a place to do that where the law is also favorable to a not-for-profit type organization that is structured the way we want it to be structured is another problem, finding a friendly country that has laws that would allow you to incorporate that way.

So anyway, we're putting that back up on the table and so the joint RIR Boards will be discussing that as well.

One other matter that our Board has been very concerned about and is also discussing it with the -- will be discussing it with the joint Boards is DNS-sec for the reversed allegations. As you know, DNS-sec is kind of spottily being implemented and at the top, there's some political discussions about who and when and so forth, well, the RIRs are responsible for a large number of domains. They are all the reverse domains both in the IPv4 and IPv6 world. And at some point in time, we need to really be serious about implementing DNS-sec, and so we are doing that.

As far as the discussions go, and actually the Engineering Coordination Group of the NRO, one of these collaborative little groups that we can have because of the NRO, is actually looking at the technical details of where we can do it. And then also discussion is going to be at what point in time do we approach the manager of the Arbor (?) domain and say you must sign this domain because we've signed all of our stuff below you, and so forth. So work being done there.

And then on top of all that, there's the normal day-to-day type of things, if you will, that the Board's concerned about in terms of fiduciary status of health of ARIN, which, by the way, I'll say is good. It's actually excellent. And things that we can do to further increase our outreach. The Board is determined that ARIN must continue to provide outreach both in the matter of IPv6 implementation and education and a lot of other things.

So the Board has issued statements about the implementation of IPv6. We have increased our outreach. Richard and Megan are on the road all the time going to different places for that purpose. I go to different places. The chairman of the Board, John Curran, goes to a lot of different places. We've drafted members of the Advisory Council to help in this.

So we're doing a lot of things. And as you can tell from our comments this morning and particularly the ones that Nigel made, there's even more reason for us to keep on doing this because the developing portion of the ARIN region, which is where we are right now, needs to have a lot more understanding of what's going on and the importance of that for the Internet to grow and develop the way we want it to become a flourishing thing in the Caribbean.

So the Board's concerned about a lot of things and it's been paying a lot of attention to them. That's a summary real quick of what's going on with the Board. Does anybody have any questions or comments or something you would want to bring to the Board for their attention?If not, I would like to ask Leo to come up and give a report on the Advisory Council. And as I mentioned this morning, Leo is the chair of the Advisory Council.

MR. BICKNELL: Hello again. The ARIN Advisory Council consists of 15 members that are elected from the public-at-large. And in fact, one of our elections is going on right now, so I want to call that to your attention. The candidate names were recently posted to the ARIN website, I believe in the last week actually, so you can go there and click on the election link and find out about the AC elections and also the Board and NRO and other things going on.

But we are elected from the public. It's definitely a way you can participate. You can get nominated and run for an AC slot. We do have a couple AC members retiring this year so there's definitely going to be some new blood on the Advisory Council no matter how the elections come out, so check that out.

Our primary function is in moving policies through the policy proposal process, the IRPEP as we like to call it, and we're involved at several different steps. I'm going to kind of start backwards here.

Since the Denver meeting, a couple of policies have been moved through to completion -- 2007-21, which was IPv6 for Legacy holders with RSA and efficient use. There was deemed consensus after the Denver meeting and that has moved through the process. The AC recommended to the Board that it be adopted and the Board has in fact adopted it, and it was supposed to be implemented by the 31st of August.

2007-23 end policy for IPv4 allocations to RIRs. This is the policy that was mentioned before where when IANA has five /8s left, they will give one each to the five RIRs. That passed in the ARIN region and will be implemented upon all five RIRs implementing the policy and the NRO and ICANN deciding that all of the appropriate policy process has been followed to make that happen.

2008-1, SWIP support for smaller than a /29, which is relatively self-explanatory. Previously you could not submit a SWIP template for a /30, 31, or 32, and this policy changes that. The AC recommended that be adopted and it was supposed to be implemented by the 31st of August, according to the Board. And 2007-17, this proposal took a little bit of a more circuitous route to adoption. It is Legacy Outreach and partial reclamation.

And the AC decided that there was consensus for this policy coming out of the Denver meeting and forwarded it onto the Board with the recommendation that it be adopted. And the Board actually remanded it back to the AC, having a couple of questions about the outcome of the Denver meeting and wanting to be sure that the policy process had indeed been followed in all manners.

And the AC actually made a minor edit to the language of the policy that we believe satisfied the Board and sent it back to the Board. And it has, in fact, been adopted by the Board, but I have not yet seen an implementation date for that policy.

So since Denver, we have four policies that have gone on to completion there.

During the meeting, we've been talking about all the policies currently on the table for this meeting and the upcoming meeting in L.A., so I'm not going to go back over those. They're in your packet. We've been talking about them.

There are also three items of new business the AC will be taking up that you may have already seen on PPML. Policies have been submitted all regarding who is -- the first is, who is authentication alternatives, from Michael Sinatra. The second is, who is POC e-mail clean up from Ted Middlestaedt, I think is how you pronounce it, and the last is, annual WHOIS POC validation by Chris Grundemann.

So obviously, there's a lot of interest in WHOIS. The AC will be taking these up at their next meeting and see if they become formal policy proposals or not. We're obviously looking for community input on that. We encourage you to go to the mailing list and put in your two cents as to what you think about them and how you think the AC should handle these going forward. They will not be on the agenda for the L.A. meeting. They all came in after the cutoff date for that, so we'll be talking about them for the spring meeting.

So what else has the AC been doing? Well, the AC takes care of this business in monthly telephone calls, is the primary way we get business done. We have a call every month. It tends to be an hour to an hour and a half. At the beginning of this year, as was mentioned in one of the proposals, we actually had a summit over a weekend. We managed to bring all 15 members together in Reston, Virginia, for a two-day weekend workshop and the primary thing that motivated that was the transfer policy.

As you saw with the proposal and the discussion, there are many facets to a transfer policy and there's a lot of complexities that have to be balanced together to make a cohesive policy. And the AC felt that without sitting down and taking some time to actually have discussions about that and hash out different alternatives, that we couldn't create a really good policy in that area.

So we got together, we created the initial draft. The AC has then been driving the various drafts of the transfer policy since that time. And that's definitely taken up a significant amount of time of the AC members this year and really kept us busier than any year before that I've been on the Advisory Council.

So I believe that pretty much covers the AC activities. We will all be in L.A. if you're coming to Los Angeles. There are four of us here. If you haven't met us yet, there's myself, Stacy, Lea and R.S. Certainly grab us in the hall and talk to us if you want to know anything about the Advisory Council. And unless anybody has any questions, that's all I've got.

MR. PLZAK: And now to provide a report on the Numbers Council for the NRO, Marty Hannigan. He's one of three people that has been elected/selected to the Numbers Council from the ARIN region.

MR. HANNIGAN: Hello, everyone. My name is Martin Hannigan. I'm with the Numbers Resource Organization Advisory Council. I'm also employed by a company called Verne Global out of the states, and I'm a consultant to Bahamas Telecom occasionally. I'm not speaking for BTC today.

So the NRO NC is comprised of members from all five RIRs. Two are elected from within the RIR communities, one is appointed by each Board. I'm an appointed Board member from the ARIN region. Current members are ARIN, APNIC, AFRINIC, RIPE, and LACNIC. Those are the five RIRs. There's 15 members and as I said, 2 elected, 1 appointed.

So what does the NRO NC do? So under the terms of the Memorandum of Understanding, the MOU, which was signed between ICANN and the RIRs in October of 2004, the Number Council performs the role of the supporting organization AC. ICANN has several supporting organizations of which the ASO is one of them. And that leads me to what do we actually do?

So there are some very key and important roles that we play and the main one is we're a check and balance in the global policy system. One of the first things that we do is we appoint two seats to the ICANN Board of Directors that are generally RIR community friendly. We appoint one representative to the ICANN NONCOM. Those are the folks that recruit and appoint additional Board members to the ICANN Board. We process global policies from the RIR community. We're not necessarily a policy creating body. We ensure that when a policy does reach the NRO NC that it has followed each RIR, each region's individual bottoms up policy development process. And we basically ratify that and say, yes, in fact, it did follow that process, and then send it along to the ICANN Board.

So what is a global policy? So as defined by the MOU, Internet number resource policies are agreed by all RIRs according to their PDPs, agreed by ICANN and require specific actions or outcomes on the part of IANA or any other external ICANN-related body in order to be implemented. An example would be IANA allocates IPv6 address to the RIRs in blocks of /12s and would they in turn allocate sufficient space to support us, the end-users. Pretty simple.

So with that, we currently have a single -- in the recent history, we had two global policy proposals, one which was recently ratified by the ICANN Board of Directors on 31 July, and that was 2007-19, IANA policy for allocation of ASN blocks to RIRs. Basically when you apply for an ASN, it has to be allocated. IANA holds the entire pool of numbers. The RIRs make requests to the IANA and this is basically a process for IM to allocate those ASN blocks to the RIRs.

We currently have one active global policy under consideration and that would be the 2007-23 end policy for IANA IPv4 allocations to RIRs. That has been accepted in four regions, I believe, and is currently under Last Call in the APNIC region. And what that means is that there's one final action for the Asia-Pacific region to take and the policy will then be forwarded on and follow the processes that's defined by the MOU and its attachments.

And that's it. That's all I have for you. I do want to mention, as Leo mentioned, there are elections and there is a seat available on the NRO NC, which is mine. My term is up and I would appreciate your support.

Thank you. (Applause)

MR. PLZAK: Thank you, Marty. Leo mentioned in this presentation about some policies that should have been implemented by the 31st of August. They were implemented on the 5th of August, so they've been done.

So instead of breaking early for the break, because we just had lunch, I'm going to ask Richard to come up here and give his presentation on the activities of ARIN in the Caribbean.

So Richard?

ARIN Activities in the Caribbean

MR. JIMMERSON: Thanks, Ray. My name is Richard Jimmerson. I work largely in the outreach area for ARIN. I do a lot of traveling around with several different people inside the ARIN organization and conduct outreach. I do these meetings like the one we're having today and attend other events like booth events where we're exhibiting or we'll give presentations on agendas that are taking place throughout the region.

So this is a bit of information about what we've been doing in the Caribbean region and some of the things that we're going to be looking to do in the near future. Ray had alluded to some of those things a little bit earlier. But in 2007, we attended the events that you see up on the screen here. At the ARIN Caribbean workshop in St. Thomas, that was a precursor to these ARIN Caribbean Sector meetings. We had a meeting for the first time inside the Caribbean as ARIN to provide an introduction to the organization and our activities, what it is that we were doing, and this meeting took place in February of last year.

And just following that meeting, we had a public policy meeting in Puerto Rico in April. That was a public policy meeting. It was the full region public policy meeting where we had people coming from Canada, the United States, the Caribbean, and all over the world to attend that meeting and discuss policy proposals just as we are here. It was a slightly larger meeting.

Of course it's the larger meeting. We had between 100 to 200 people that attended that meeting.

Also last year we attended the CTU ministerial meeting on Internet governance and gala and attended the CANTO meeting in Barbados in June.

The CANTO meeting is one of the meetings where we'll go and we'll attend and we'll actually have an exhibit there. We'll set up a booth. They have floor space for that. And we'll also try to achieve space on the agenda to talk about these types of topics that we've been talking about today to gain additional interest from inside the Caribbean into these topics.

We also attended the ICANN meeting in Puerto Rico in June, a CTU Caribbean IGF in Curacao in August, and the 2007 Caribbean Internet Forum in St. Lucia in November of 2007.

Now, it's quite important that we get additional participation region-wide not just in the Caribbean, but specifically in the Caribbean because there is not a whole lot of participation in the process years past. And one of the reasons why we're conducting more activities in the Caribbean is to try to gain more interest, to gain more participation.

I know that there was some talk at the open microphone earlier about bringing more people into the discussion. One of the things that we've been working hard to do is ensure that there's access to the organization and to our process so that people can participate. And one of the things that we're finding to be a challenge and that we're working on is how do we convince Internet service providers and enterprise companies inside the Caribbean to participate in the process? How do we convince them that there is importance to it, there's value to it?

So that's something that we're continuing to work on and any feedback that you have at the microphone here today or that you can send us via e-mail or talk to us about in person, would be very helpful.

In 2008, we held an ARIN Caribbean Sector meeting, earlier this year, May, in Kingston, Jamaica. It was a meeting very much like this one. They had approximately 70 people in the room for that meeting, so it was slightly larger than this one. And we also attended the CANTO this year that was also here in the Bahamas, over on Paradise Island. We had an exhibit there and we had presentations.

Another Caribbean Internet Governance Forum in Curacao in July and, of course, this meeting here today in the Bahamas.

There are some other meetings that we will be attending throughout the year here in the Caribbean, so we're looking for more opportunities.

But in the future, we're looking to continue this effort and continue to do more and find new ways to reach out to people and speak to them and get them to participate in the process.One of the things that we're going to be doing is possibly changing the way we hold our meetings. You can see in February of next year we have a meeting set up as tentative for Barbados. It will be another ARIN Caribbean Sector meeting. What we're looking to do beyond 2009, starting in early 2010, is to change our meeting structure. Right now we have two large public policy meetings as we've been having for 10 years now, and we've been holding those mostly in continental North America. And we've started to have some meetings in the Caribbean and now we're doing some Caribbean Sector meetings.

But what we're looking to do is eliminate splitting the region into sectors and holding meetings specifically for that sector, but hold three public policy meetings per year. Hold one of those public policy meetings in the Caribbean, one of them in the United States, and one of them in Canada. So no longer will there be this distinction about it's the big public policy meeting or it's the sector policy meeting. We're going to try to remove that distinction and just have three meetings per year so that all portions of the region have equal access in physical proximity to the meetings that are held throughout the year.

We're also looking at increasing our remote participation facilities. Right now when we hold our larger public policy meeting, we will have a video being taken of that meeting and the audio being taken of that meeting, and it's provided via live feed where you can watch and listen to the meeting and participate in that meeting.

We allow you to send in communications to us while the meeting is taking place. And if you do that, we will take the question that you have or the comment that you have, we will walk it to the floor, the staff will, and we will read your statement or we will read your question directly to the floor. You'll see it be asked on the video feed. And you'll see response to that from the rest of the community, they're physically at the meeting or perhaps from a remote participant.

We're looking at increasing that and going even beyond that to make that even more interactive and more available to people because it's understood that not everyone can travel physically to these meetings to participate. Not only can you participate remotely at the meetings, but you can also participate on the Public Policy Mailing List. We're looking for more people to participate there.

When the Advisory Council does their work to judge the consensus of a discussion, they're looking at everything that takes place on the Public Policy Mailing List as well as the discussions that take place here at the ARIN Caribbean Sector meeting and the discussions that take place at the ARIN public policy meeting.

So they're taking all of that into account. So there are more than just one way that you can participate in the process.

Another thing that we'd like to invite -- if you have a future activity coming up, if you have an event that you know about where Internet governance is going to be discussed or Internet addressing is going to be discussed, anything having to do with the Internet, where people may need to know that there's an open process here to participate in, when it comes to the governance of Internet Number Resources, or if they need to know about the depletion of IPv4 address space and the adoption of IPv6 address space, please contact us and let us know. You can send e-mail to info@arin.net. It doesn't cost you anything to just send us an e-mail and let us know about it. And if it's something that we can travel to or something that we can send literature to or something that we can support in another way, we will do that. So I'd like to invite you to do that.

We do have plenty of documentation available for events, so even if you don't want to have an ARIN staff member, for instance, come to your meeting and do a presentation there, perhaps you just want the paper material that we provide, contact us. We would be happy to send you as much paper material that you need to support your meeting. The comic books that we have out there, the one-page factsheets that we have, we can send you electronic copies or digital copies of things. It's all available on the website to download as well.

And if you have something specific that you need, we can even put something together for you that's specific to that meeting or the needs of that meeting.

So really, that is it. I'd like to accept any feedback that you have on this. If you have any great ideas on how we might be able to increase participation inside the Caribbean, any ways where we may do additional outreach or do a better job of doing that outreach, I invite you to come to the microphone now and let us know that in a frank way or a friendly way, any way that you'd like, or you can contact us at info@arin.net if you'd prefer to do it that way.

Yes, Carlton?

MR. SAMUELS: I want to bring to the attention of the community something I think is very good, what ARIN has been doing. Those comic books that ARIN produces, what we've done in Jamaica is we've infiltrated them into the schools and the public library system, and to thank you publicly from the library service of Jamaica because all the books that you send, some 2,000 copies, they were distributed to the school's libraries and in the public library services. And they are even willing to take some more.

MR. JIMMERSON: Thank you very much, Carlton. Any other comments? Any ideas on what we could do to increase participation? If there's something that we're doing well that you want us to do more of, anything that you have -- if not, we'll talk to you in the hallways or get e-mail from you.

All right.

Thank you very much. (Applause)

MR. PLZAK: Okay. Now we can take a break. And so let's go out and take that break. Remember when you come back you will find the survey on your desk. Please fill that out. And also when you come back, we will be having another prize drawing because we have a few more things to give away, assuming, of course, Einar doesn't break it some more.

Anyway, so let's take our break and be back here in about 15 minutes. Okay? (Recess)

MR. PLZAK: Before I start, I should have gotten a survey. One thing I didn't tell you about this survey is that if you fill it out and you turn it in, your name will be put in a drawing for an iPod. Okay? So if you want an iPod, this is one way to get at it. So please fill out the survey, because it's important to us and we're hoping we're making it a little bit more important to you by offering an iPod. So enough said on that.

Okay. Policy -- well, two more policy proposals for the afternoon.

2008-3: Community Networks IPv6 Allocation

MR. PLZAK: First one is Policy Proposal 2008-3: Community Networks IPv6 Allocation. This was designated a formal proposal earlier this year. It's been discussed at two meetings so far the year, and XXI meeting in Denver, again in Kingston.

To date, the consensus has been to revise a proposal. The current version is as of the 16th of September, so talking about very, very recent. It's under discussion.

This is the only region where a proposal of this nature is being discussed.

To summarize it, it provides an IPv6 assignment of a /48 or larger to a community network that has at least 100 users and projects to have at least 200 users within one year. The shepherds are Lea and Stacy -- oh.

You have to update Stacy's name. Or maybe there's something that she hasn't told Aaron yet. Honeymoon's over.

So the latest revision was yesterday. It went from nonprofit, which was a criteria to a specific criteria. And a specific criteria was the staff must be -- at least half of them must be volunteers and it must have an annual revenue of less than $250,000 U.S. per year.

The legal responsibility for the network, it must be run by a nonprofit organization or it must be fiscally sponsored by a nonprofit or a university.

That's the latest revisions to this proposal.

And prior discussion with 129 people in the room in Denver, 4 were in favor it as written, 40 against. Should the AC abandon? Most people said no. In Kingston, two supported it as it was with no one opposing it. And should the AC continue to work on it? Twenty-one said yes, keep on working on it.

So that's what's been going on. You can see by the slated revision that work has been done.

So with that, I think I'm going to ask Lea to come up here and summarize the proposal and lead the discussion.

Lea?

MS. ROBERTS: Okay. Hello, everyone. I'm Lea Roberts from Stanford University. And as was said, I'm on the ARIN Advisory Council.2008-3 was submitted previously. The main driver is to try and enable IPv6 address assignment to community networks whether or not they're multi- homed or meet the other criteria for provider independent addresses.

The proposal posits that the community networks would benefit from the ability to have their own address assignments. That way they could gain connectivity through providers -- different providers and also have their infrastructure numbered in a stable addressing scheme.

So that's the basic concept behind this. And I think in the rationale they also mentioned -- I don't know if it's still there, but initially one of their interests was in trying to put in a suggestion that the fees for community networks be less. That's not part of the policy process. Fees are never addressed in ARIN policy. But they can be addressed in the suggestion process. And someone, I believe, has put a suggestion in suggesting much the same thing even before 2008-3 was submitted. I haven't heard any more about what's happening in that direction.

So are there -- you know, any questions or comments about this proposal? The floor is open.

MR. SEASTROM: Hi, I'm Rob Seastrom. I'm on the ARIN Advisory Council. And I've had a little bit to do with trying to get the language worked properly in this proposal.

And one of the things that has been a bit confusing for us is, we had this notion that, well, it would be a nonprofit organization, of course. But nonprofit organization is something that changes a lot depending on the country that you're in. In countries that don't tax businesses, a for-profit versus nonprofit corporation is a meaningless distinction, for instance.And so we on the Advisory Council would welcome input from people who are residents in countries where the tax code and structure is very different from the United States, which we know about, and could offer us some suggestions on how to appropriately word this so that community networks in your -- I'm going on the assumption that everybody knows what a community network is, even if you can't describe it very well, which we can't describe it very well or we would have just fixed the problem. But what community networks in your country are like and how they are organized as legal entities, we would welcome the input for how to make that -- how to make 2008-3 serve the community networks in your country, if that makes sense.

MS. ROBERTS: Sounds good to me, R.S. Well-said.

MR. SAMUELS: If I could --

MS. ROBERTS: Please.

MR. SAMUELS: Regarding the numbers. The -- that last number there.

MS. ROBERTS: On how many users there must be?

MR. SAMUELS: Must have at least 200 users?

MS. ROBERTS: Right.

MR. SAMUELS: But I think that would be a little bit --

MS. ROBERTS: A little bit steep for this area?

MR. SAMUELS: Steep for the kind of community networks that we see in the Caribbean. And I'll tell you where they are, where I've seen them in the Caribbean region. There are lots of them in the -- in Surinam. Believe it or not, a lot of the services that they provide to the interior, especially with indigenous groups, are from community-based networks that are developed by NGOs and so on. And sometimes, if you get 12, 13, that's a lot.

You wouldn't want to kind of leave them all dry --

MS. ROBERTS: These numbers obviously wouldn't fit that --

MR. SAMUELS: Right.

MS. ROBERTS: That profile. But what -- do you have a suggestion as to what kind of criteria might be appropriate then, like a percentage of the local population or something?

MR. SAMUELS: Yeah, it's the timing that I have difficulty with there. It says within one year -- to grow by 100 percent within one year. That's --

MR. HUGHES: Do your community networks have more than one provider?

MR. SAMUELS: No, but they are going to have very soon. Some of them do have.

MS. ROBERTS: And also -- I mean, Aaron asked do the community networks have more than one provider? Which -- the implication being, well, why not just use provider-based addresses?

I think -- you know, I can answer that as well from the perspective of the people who proposed this, which is they want to build their infrastructure with their own addresses and be able to choose providers without having to renumber.

MR. SAMUELS: I will tell you one thing. You probably know that there has been several initiatives being -- going on for about -- open source YMAX (?), for example, which is open source --

MS. ROBERTS: Oh, right.

MR. SAMUELS: YMAX, which is gaining some traction.

MS. ROBERTS: Absolutely.

MR. SAMUELS: As a matter of fact, in a lot of the open learning initiatives that I've seen, that has always been postured as one possibility for building networks to connect open source learners. So I would want to preserve some space for those kinds of networks to be involved in.

MS. ROBERTS: Thank you for those comments.

Leo.

MR. BICKNELL: Leo Bicknell, ARIN Advisory Council. I just wanted to interject. One of the things we saw in Kingston at that meeting was very wide support from the folks who attended for this type of policy, maybe not the exact wording, but this general type of policy. And as we went back and discussed that amongst the AC. And I think definitely if we discussed it across a wider audience in, for instance, the L.A. meeting, I don't think there's a very good understanding of what a community network looks like in the Caribbean. We know what one looks like in the U.S., and there's some overlap. They're not completely different, but there are some differences.

And so personally I think it would be very valuable if the people in this room who know about community networks could get someone -- perhaps yourself, perhaps someone you know involved with it -- to post a description of what it is and how it works to PPML and help educate the rest of the world as to what a community network in the Caribbean is.

You know, the bar here -- the interesting question that's first asked by those who don't understand is why do they need a provider independent block and/or a slot and routing table? So to the extent that those descriptions can help answer those questions, it greatly improves the chance that such a policy would be enacted.

So that's a way you guys could very directly, I think, help get this policy passed.

MS. ROBERTS: Ms. Stacy?

MS. HUGHES: Stacy Hughes, TW Telecom, ARIN Advisory Council. If we're going to give away IPv6, let's just give it away and take out the 100 users and the 200 users because that's essentially what we're doing with this policy.Do you know what I mean?

MS. ROBERTS: I hear you.

MS. HUGHES: Yeah.

MS. ROBERTS: I mean, I think that in the sense of providing an opportunity for a community network to get it, the bar doesn't have to be necessarily hard-drawn. It's just what we need is -- then, is the criteria independent of users --

MS. HUGHES: Right.

MS. ROBERTS: That makes it clear. I mean, the concern in the public policy meetings has always been that -- you know, that this would be kind of a back door for people to skate into and get their own provider independent addresses.

The challenge is to define -- come up with a definition for a community network that is sufficient -- you know, to preclude that in most, if not all, cases.

MS. HUGHES: Right on.

MS. ROBERTS: So this is -- you know, sort of the one that came up. But again, as Leo was saying, it's very U.S.-centric. So we need to broaden our outlook and see how we can define that kind of a structure that would pass muster for the rest of the ARIN community because it does need to achieve a certain amount of consensus to get through the process.

Are there any other suggestions or comments here regarding this proposal?

I guess it's time for questions.

MR. PLZAK: Okay. I'm going to ask two similar questions, as similar as the ones I asked last time for previous proposals.

First question will be, are you in favor of something like this as a policy?

And then a second one, your opinion on this specific proposal as modified.

And could you flip to the next one? This is the revisions that make it different from what's in your packet that you have handy.

So on the matter of the first question, are you in favor -- all those in favor of a policy that is like this policy or the AC should do work in this area, please raise your hand. Put them up high. (11 hands raised)

MR. JIMMERSON: I have a count.

MR. PLZAK: And how many are opposed to that general question of work in this area, a policy like this? (0 hands raised)

MR. JIMMERSON: I see no hands.

MR. PLZAK: Regarding this specific proposal, how many people are in favor of this proposal as it is presented with this revision which was posted at a list just yesterday? All in favor, please raise your hands. (0 hands raised)

MR. JIMMERSON: I see no hands.

MR. PLZAK: And those that are opposed to this specific policy to include with this modification? (0 hands raised)MR. JIMMERSON: No hands.

MR. PLZAK: Very interesting. So I'm not going to let you guys get off that easy.

So there are a number of you that raised your hands and said, hey, I like this. But you don't like the specific things. So let's put some more ideas up on the table here. We've got some time. And so I'm prepared to sit here and wait until it gets a little bit colder outside.

MS. ROBERTS: You could ask without the number of users, Richard.

MR. PLZAK: Okay. Let's explore some possibilities. Do a little bit of the white boarding, okay?

Could you flip it back to the other one? Okay. How many would like to see this policy proposal without the number of users? Let's start with how many would like to see it that says at least 100 users and projects to add 100 -- without any projection, just have at least 100 users. (1 hand raised)

MR. PLZAK: And how many people say we shouldn't even do that? (4 hands raised)

MR. PLZAK: Okay. So -- thank you. If we have the 100 users, would someone -- how many people are in favor of putting a requirement to have a certain measurable amount of growth, whether it's an actual number or a percentage? So if you're in favor of having a requirement for growth within, say, for example a year, please raise your hand. (4 hands raised)

MR. JIMMERSON: Okay, got it.

MR. PLZAK: And how many are opposed to that notion? (2 hands raised)

MR. PLZAK: Okay. Now --

MS. ROBERTS: I'd like to interject here --

MR. PLZAK: Go ahead.

MS. ROBERTS: One of the things that occurred to me when -- the comment about -- you know, networks in the wild is -- you know, at some point you're going to grow. You know, you can't keep growing because there's not much to grow to. So is -- does anyone have an idea of how we could include some sort of community metric? In other words -- you know, there'd be -- for a certain size of a community there might be a number, but for another community, it would simply be related to the percentage of the community. And how do you define a community in the hinterlands?

I mean, it's -- if it's a real wide-open space with very few people maybe that's just a big community? Or?

MR. SAMUELS: Let me give you --

MS. ROBERTS: But I think -- you know, I understand your point, sir.

MR. SAMUELS: I take the suggestion that you made, because it's useful as I think about it, most of these communities that we're talking about, at least the ones that I'm familiar with, are special needs communities. They are indigenous communities just like the Carafunu (?) communities and Maroon communities and so on. They are not, as you rightly say, they are not going to expand indefinitely and grow uncontrollably, but they have a need to be connected and see themselves as part of this whole structure.

I'm not sure it -- the community definition changes from group to group. For example, in the open learning communities that we see, for example, it involves people who are -- drop out from high schools and who have never been placed in the formal school system, and then only (inaudible) could number couple dozen in some communities and they could number hundreds or thousands in other communities. So it's a very difficult kind of metric to put one-size-fits-all kind of measurement against it.

MS. ROBERTS: Just fishing for metrics here. Thank you.How many people are in favor of a structure being defined in terms of staff? In other words, a characterization of a staff that's running this network. How many people are in favor of something like this? Please raise your hand. (2 hands raised)

MR. PLZAK: And how many people would be opposed to something like that? (2 hands raised)

MR. PLZAK: Okay. How many people think there should be some kind of a revenue thing on as -- you know, doesn't have to be $250,000 a year, but some upper boundary or boundary that if you -- if your community network earns more -- or has more revenue than a certain amount of money, that it no longer qualifies under this policy? So please raise your hand. (1 hand raised)

MR. HUGHES: One. Is this one of the --

MR. PLZAK: Opposed to that? Okay. Aaron, do you have something?

SPEAKER: I'm just a little confused and I'm sure this has probably come up before. But if you're making revenue, why not spend it -- you know? Maybe $300 to file for an LLC and have a real business?

MR. PLZAK: It may be there's some other reasons why they don't want to do that. I can point to one that doesn't have a -- I can point to an organization that doesn't have a budget, doesn't have a staff, but does spend a good chunk of change, and that's the NRO. So there may be other reasons why you couldn't become a legal entity in that perception. Okay?

MS. ROBERTS: And again, it might depend on the country you're in, whether that would make sense or not, too.

MR. PLZAK: Right. Okay. How about the statements regarding the legal responsibility for the network? How many people say that somewhere or other, it must be either a not-for-profit organization or be sponsored by a not-for-profit organization? (6 hands raised)

MR. PLZAK: Okay. And how many people are opposed to that? (2 hands raised)

MR. PLZAK: Okay. Now, you notice I didn't say university. I specifically left university out of that. How many people think that universities should be involved or can be involved in such a determination? (11 hands raised)

MR. PLZAK: Opposed?

SPEAKER: Wait, I didn't finish that.

MR. PLZAK: Okay.

SPEAKER: Okay.

MR. PLZAK: Opposed. (0 hands raised)

MR. PLZAK: Let's drop the ladder a little bit down from university. How about some other type of civic organizations or so forth, like maybe a library system or something like that? What do you think about those kinds of ideas? In other words, the university doesn't support something, but a local library might be able to do that. Anybody in favor of looking at those kinds of organizations as a sponsor?

MS. HUGHES: Are they nonprofits?

MS. ROBERTS: Not necessarily.

MR. PLZAK: Not necessarily.

MS. ROBERTS: But you'll notice that a university here may not necessarily be a not-for- profit university.

MR. PLZAK: Right. So those people who think that we should allow somebody else besides university, and I'm just throwing out as an example a library -- library system type of thing? How about that? Or a church group?

MS. ROBERTS: How about a town?

MR. PLZAK: Well, a town. It -- that makes a community.

MS. ROBERTS: Community?

MR. PLZAK: Right. So some sort of sponsorship other than a specific not-for-profit or the organization being not-for-profit. Okay? Can you think of anything else you want to throw up here?

MS. ROBERTS: Not offhand.

MR. PLZAK: So Leo is approaching the mic.

MR. BICKNELL: I'm going to make the same comment I made in Kingston on this proposal. And that is, in this venue today and in Kingston, I didn't hear hardly any talk of routing slots. And I pretty much could guarantee you that two-thirds of the speakers that are going to come up in L.A. Regarding this proposal are going to talk about routing slots.

And so I'd like to at least ask a question or two there. Because one of the questions that has come up is, does a community network necessarily mean an address block that is going to be seen on the public Internet? Or do they simply need unique space to interconnect to each other or their local university, whatever, but it would never be seen? Because that determines whether or not there is a slot in the global routing table.

And if there's a slot in the global routing table, what is the justification to all of the other people in the world who have to spend money on bigger, faster routers and all that to carry that route, particularly if it's for a very small number of people? We just talked about dropping the requirements from -- you know, potentially 100 to 200 people to a group of a dozen people. Well, if I put a route in the routing table for every dozen people on the planet, the Internet is not going to hold up too well.

MS. ROBERTS: I don't think it necessarily requires a slot in the routing table. I mean, one of the nice things -- again, I'm not speaking for the people who did this proposal other than when we were -- when I was talking with them, they were really clear that one of the things they want is to be able to build their infrastructure with solid stuff -- set of space. And certainly if someone was willing to route it locally for them, then that would be an advantage. They might get some kind of committee to other people using that space.

But I don't think they have had necessarily the belief that all these would have to end up in the DFC. But that that's one of the nice things about IPv6 is that you can -- you know, then get a prefix from someone else and distribute it throughout your infrastructure.

MR. BICKNELL: Right. I just think that's key to communicating the potential magnitude of potential prefixes.

The one that gave me particular pause here was the university tie-in. Because I've worked with two, three universities that have done community-related networks where they've gone off the university campus and helped connect up libraries and students who lived off campus and all sorts of things that could arguably be community networks.

And in each case that I've seen of those, they have taken an existing university net block -- you know, a /16 or whatever -- and carved up a /20 out of that to use for the community network. So there was no addition to the global routing table.

And, for instance, if this policy might encourage people in that same situation to go get a separate block just for the community network that wouldn't roll up, then to me that would be a step backwards. So I think there's a bit of a fine line there.

MS. ROBERTS: I think that's understandable.

MR. PLZAK: Anybody else have a comment or an opinion? Okay. Well, we'll take all this information that we've gathered here today and provide that to the Advisory Council for their use as they move forward on this proposal. Thank you.

MS. ROBERTS: Thank you all.

2008-4: Minimum Allocation in the Caribbean Region

MR. PLZAK: Last proposal of the day, Policy Proposal 2008-4: Minimum allocation in the Caribbean region.

Okay. This was first proposed in May of '08. It's been discussed so far in one policy meeting. That's the meeting in Kingston. It's under discussion. And there is no other region that has a policy discussion going forward like this.

And you wouldn't expect it from AFRINIC, APNIC, or RIPE NCC. And so the only other place that has portions of the Caribbean in it is LACNIC.

And so they haven't initiated such a discussion in that region.

Summary is really simple. It lowers the minimum allocation to a /22 for ISPs in the Caribbean region. The shepherds are Heather and Matt, neither of whom are here.

In Kingston, with 38 people in the room, a lot of people supported the general idea. And a lot of people also said include in user consideration. So those are two big things.

To date, that has not been modified, so it still is as it says there.

And so I would like to call on Leo to come up and present this proposal. We'll have a discussion.

MR. BICKNELL: Hello, again. I assure you, policy proposals are almost over.

So 2008-4, minimum allocation in the Caribbean region. This was very much modeled after what was done for the Africa region about -- oh, that'd be about four or five years ago, prior to AFRINIC being created. In fact, the first draft of this policy was a direct cut-and-paste of that policy in an effort to try and use language that people had already seen.And since then, it has been updated to fit the standard policy proposal framework. So the version you see in your packet is the same intent but just formatted for the policy manual.

So basically the policy today for an ISP is if you are single-homed, you must justify a /20 and then you can receive a /20 from ARIN. If you are multi-homed, the smallest you can go is you can justify a /23 and then receive a /22.

This policy would change that such that if you're single-homed, if you can justify a /22, you can receive a /22. And then I'm going to come back to this in just a moment, but if you're multi-homed, and can justify 2 /24s, you get a /23.

And I actually want to ask my other AC members are just puzzling over this with R.S., it seems to me we've introduced a typo or a mistake or something here. Because 2 /24s is a /23, essentially. So the existing policy is if you can justify a /23, you get a 22. And the new policy is if you can justify a /23, you get a 23.

So I think that was supposed to be if you can justify a 24, you get a /23. That would be in keeping with the pattern. But I did not get time to dig through all my e-mail archives before I came up here. So I don't know if any of my fellow AC members might be able to help and see if that just got transposed in this version of the text that got printed, or got lost in the translation, or if that has been wrong from the start?

So definitely we might want to take a closer look at that aspect of this policy going forward.

But this was all driven out of feedback from this region that the existing sizes of a /20 and a /22 were often simply too large to meet the scale of the local ISPs that could really benefit from having provider independent space. And so we would like to continue to hear from you guys as locals to this region to help us bolster the case for why this would help the region and how it would help the region and, of course, look for any other comments and suggestions you might have.

So any questions or comments? Everybody must think this is a really bad idea, right? No one wants to stand up and say they really want this? Yes?

MR. JIMMERSON: Can you say just to the audience what a /24 -- how many addresses that is, what a /23 is?

MR. BICKNELL: Oh, sure. So making me do math in my head. A -- yeah. A /23, which is the current low-end limit for multi-homed, would be 512 addresses that you'd have to justify. A /20 would be 496 addresses if you're single-homed, that you'd have to justify.

And so the new policy would set the single-home bar to /22, which would be 1,000 addresses, and the multi-homed policy at 256 or 512, depending on how we figure out if that's a mistake or not.

MR. PLZAK: So in effect, what this says is that you only need to satisfy a requirement of one quarter of the amount of addresses as what the current minimum allocation policy says.

Now, I'd also like to point out one thing. When we were discussing this in Kingston, we pointed out that -- you know, that the rest of community is going to listen to what the Caribbean community says about this, and I encouraged people to discuss this on the PPML. I have yet to see one e-mail message from anybody in any part of the Caribbean saying, yes, this is a good idea, this would definitely help us develop and grow the Internet in the region. And this is really important, because the other people that are going to say yes or no to this policy may decide to set aside what may be some technical problems they might have in favor of adopting this policy because it would be healthy and it is wanted by this region.

So I'm being a little chastising here, but if this is something you really need and want, you need to say so. And you need to actively say your voice.

Now, I will say that when we had this discussion regarding the proposals for Africa. There was some considerable e-mails that came out of Africa to the proposal. And actually, people journeyed to the ARIN meeting and made the proposal.

They stood up and said something. So you know, this is one of those places where you've got to stand up and be counted. Your voice has got to be heard. There's things out there that says who is ARIN? And ARIN is you. And this is one real good case for it.

So this is one of those things that if this is something you want that you think is important, you've got to say so. Because the rest of the community can't sit back and guess what you want. They want to know what you want.

So I will now get off my soapbox and go back to being my normal, docile self.

SPEAKER: Stacy?

MS. HUGHES: It bears noting that in the policy text, they have to show efficient utilization of space. They already have from an upstream. You can't just justify, according to the policy text.

MR. BICKNELL: Yes, that is a good clarification. Sorry, I was being too loose with my language.

MS. NOBILE: And I'll clarify for you that the policy is correct as written. Multi-homes says you can get a /22 if you have a /23 that you're officially utilizing from an upstream. What changed in this policy, what lowered the threshold in this policy was the single-homed. Because what we found in the Caribbean was most people were not multi- homed. So by taking away that /20 requirement, which was very high, and lowering it to a /22, that's where that came into play.

MR. BICKNELL: Okay. Well, yeah, just my own comment. If the intention isn't to change the multi-homed requirement, then it's likely the AC would take that feedback and edit that paragraph out. It makes no sense to adopt a paragraph that does nothing.

But I can't emphasize what Ray said enough. You know, the folks with the African version of this policy came out in droves and that was a huge reason it was supported. And really, if you guys want it, even if you can't come to L.A., if you stand up and say something here -- you know, it's all being transcribed in the minutes and it's something that can be pointed to as support. So it really is important.

MS. LEWIS: Bernadette Lewis from the CTU. I think I may have indicated the support of the Caribbean Telecommunications for the proposal. I want to reiterate that we are 100 percent behind this proposal. I think it is absolutely necessary and I am going to encourage my colleagues to put their names to it as well and going on the PPML and adding their comments. We've asked them to do this in the past, and I'm asking them again so that there's no doubt in anybody's mind that we don't -- that we do indeed want it.

Thank you.

MR. BICKNELL: Thank you.

MR. SAMUELS: Carlton Samuels, the University of the West Indies. The facts in the Caribbean are not controversial about this. We have smaller spaces with smaller numbers, and that translates to requirements for smaller allocations.

Those are the facts, they're not controversion. We support it 100 percent, and I'm sure that we're going to gin up the e-mails so you can hear from all the people. But the plain facts are not controversial.

Thank you.

MR. PLZAK: Carlton, it would be very good if someone -- you or someone like you would get on the PPML and just say exactly that.

MR. SAMUELS: Okay.

MR. PLZAK: Okay. How many people in this room are subscribe to PPML? (6 hands raised)

MR. PLZAK: How many -- okay, keep your hands up. Now, how many of you people reside in a country outside of the United States? (4 hands raised)

MR. PLZAK: You don't count, Bernie. Okay. So I saw a very small number of people that are from outside the U.S. that are subscribed to PPML. You've got to get on the list and you've got to be vocal. You know, this is one person, one vote type thing. The CTU is one vote. It's not 30,000 votes or whatever. University of West Indies is one vote, one voice, not all the students. So the more people the better.

MR. BICKNELL: I would like to add one more small footnote to this policy. The policy as written applies to ISPs, and that is how the first draft of the policy applied. There was some discussion in Kingston that the bar for end-users should also be lowered, and the AC is definitely sympathetic to that call. However, in discussing what we should do with this policy proposal, we did not see overwhelming support to modify this existing proposal. So we decided to leave that for an additional proposal should someone want to make it.

So if you do feel strongly that end-users should also have a lower bar of some sort, someone needs to write up that proposal and submit it. And there are instructions on the ARIN website to download a template and write a proposal and submit it, and don't fear if you don't get it perfect on the first try. That's what the AC is here for. We work with staff and the submitters very early on in the process and make sure that the language is in the correct form and those sorts of things.

So don't feel shy to submit a template if you're interested in that or any other change for that matter.

MR. PLZAK: Okay. Let's do a show of hands. And actually, I'm going to ask the same two questions that were asked in Kingston.

And actually Leo can you change the slides here?

SPEAKER: It's right there.

MR. PLZAK: Okay. First question we asked in Kingston was, how many people supported the general idea behind this policy? Not necessarily the exact words, but want this to happen. So if you are in favor of this happening, please raise your hands.

SPEAKER: Raise them high.

MR. PLZAK: Keep them up. (17 hands raised)

MR. JIMMERSON: I count 17.

MR. PLZAK: Okay. All those that think this is a bad idea, please raise your hands. (0 hands raised)

MR. JIMMERSON: I see zero hands.

MR. PLZAK: How many people in the room are in favor of this being -- including an end-user consideration? In other words, this proposal as written is only for ISPs? And to add the ability of end-users to apply into this policy? How many people are in favor of an end-user consideration? Please raise your hand.

MR. JIMMERSON: Raise your hands high. (4 hands raised)

MR. JIMMERSON: Four?

MR. PLZAK: How many people are opposed to that? (2 hands raised)

MR. JIMMERSON: I see two hands.

MR. PLZAK: How many people don't understand what the question is? Okay. I'll explain it and then I'll ask it again. Okay.

ARIN makes two kinds of provisioning of Internet number resources, IP address blocks to parties. One is an allocation, and that's given to an ISP. An ISP has customers and they in turn assign those addresses to people.

And the other kind is for what we -- is an assignment to people we call end-users. Now, end-users are not an individual person, per se. An end-user could be somebody like a university or it could be a government. It could be any kind of an organization. It could be a corporation. So those are people that would be an end-user. It could be a very large company, like a Microsoft or a Kodak. It could be a smaller company, but those are end-users.

We're not talking about individual people here. We're talking about people that have the need for provider independent address space, but are not an ISP. In other words, they don't further give out the addresses. They use them for their own enterprise, okay? So is everyone clear on that?

Now, having said all that. How many are in favor of adding an end-user consideration to this policy proposal? Please raise your hands.

MR. JIMMERSON: Raise them high, please. (11 hands raised)

MR. JIMMERSON: I have 11.

MR. PLZAK: Okay. How many are opposed to that idea? (2 hands raised)

MR. JIMMERSON: I have two.

MR. PLZAK: Okay. Well, Leo, are there any other questions you'd like to get for the AC?

MR. BICKNELL: I think that's it.

MR. PLZAK: So we'll provide that information to the Advisory Council.

Thank you very much.

Open Microphone (Afternoon Session)

MR: PLZAK: Okay. All right. We've now come to a point in time where we can have another open microphone session. Now, we discuss anything here again.

Now, since we are not having a members meeting, I would also throw this open for a members-type forum as well. So if you've got something besides policy matters that you have questions about or would like to ask about as far as the way ARIN does its business, or anything along that niche that is not a policy matter but is a concern of yours, please come up.

The microphones are open.

Okay, Bernadette.

MS. LEWIS: All right, thank you. Bernadette Lewis from the CTU. I just really wanted to, first of all, congratulate ARIN on the outreach program. I think it is very important. It is bringing a tremendous amount of information to people who would not normally be part of the meetings that happen in North America and America.

One of the things that is of a bit of a concern, we have been -- the CTU has been promoting the ARIN meetings, the sector -- Caribbean meetings and so on. And it is -- we are targeting the same set of policy makers, regulators, operators, and so on. And I got the distinct impression, even from the CTU's perspective, because we have had a number of events throughout the year, that our target audience is getting a little event-weary, for want of a better word.

So we really have to look at means -- and, of course, there are limited budgets and the number of events and meetings are certainly increasing.

And so I have committed even from our point of view and we will be looking at how we could be minimizing the sort of event-weariness. Because they would sleep, quite frankly. But we went to this one and we went to this one. They're very good. But -- you know, we have work to do in the office. And those are some of the realities that we're working with.

So I -- we will be looking at how we could promote participation and without the burden of people saying they just have too many things to deal with in the office. And I just wanted to say on Richard's presentation about ARIN's activities in the Caribbean, I think it's really marvelous. And one of the things that really should be focused on is capitalizing on the existing events.

That's my contribution.

Thank you.

MR. PLZAK: Okay. One of the things that we could look at in October of each year for the last, I believe, it's five or six years, we've held an ARIN meeting back-to-back with a meeting of the North American Network Operators Group. They are not merged meetings. They are not combined meetings. They are just held back-to-back for the very reason that Bernadette was talking about, having to go to multiple events and spend extra money.

So we certainly would be amenable to something like that in the Caribbean as well. And so it would -- it actually from a standpoint of people with their travel budgets, it's probably easier to spend several days in a place than to go to one place for a day and then go to another place for a day as opposed to maybe spending three days in a given place or four days, something like that. So there is some flexibility in that we are always open to suggestions of ways to make that better.

As we go forward and try to regularize these meetings, we would be looking at doing it in the early part of the calendar year, somewhere along the February timeframe, so we don't have to confront people with funny names like Ike and Hannah and those kind of people.

So -- but realizing that, then again, that's the popular time for the snowbirds to come south, too. So there are some different challenges that we have in this portion of the ARIN region, but we certainly are not adverse at taking them on and coming up with a solution that works for everyone.

And yes, we will continue to look for events where we can make our presence felt. And we look to you to tell us where those places are, and we are more than willing to travel anywhere. We are more than willing to provide literature and information. We are more than willing to seriously discuss some sort of sponsorship-type activity as well. So that's our commitment to this portion of our region.

Anybody else? So how many people want to leave early?So the rest you want to stay here? Okay, so.

Closing Announcements and Adjournment

MR. PLZAK: All right. If there are no further comments, then let's put a wrap on this. I think it's been a very, very good two days. We've gone through a lot of things. I particularly want to thank the CTU for helping us organize this meeting and also thank other folks that participated in helping us get in contact with persons.

One of the things that helps spread the word is the people that attend these meetings go back and talk to everybody else about what happened.

You know, this is a bottom up process. This is growing in grass roots. Any of you that's ever planted a lawn from hand and tossed seeds out there, it's a hard thing to do. And you really have to watch out when those seeds hit the ground what happens to them -- you know? Some of them land on the rocks and some get eaten by the birds and some of them don't get enough water and so forth. But when they hit that fertile soil, that's where we have to keep going and get it to spread.

So with that, I'll bring this meeting to a close. I'll say thank you very much and hope to see you at the next meeting.

Thanks. (Whereupon, at approximately 3:46 p.m. the MEETING was adjourned.)