- address-policy-wg - mailman.ripe.net

In favor of 2012-01
by Milton L Mueller 13 Apr '12

13 Apr '12

> this is exactly the problem. this implies that the ip space is an asset of the > seller, which it is not. it is a commons, and if it is sparse, as any one has the > same right to it, it is to be redistributed according to need, fair and equally. The IP address space is not and never has been a commons. Not for those of us who actually understand the vocabulary of resource economics and know what the term "commons" means. For IP addresses to be a "commons" they all would have to be available for use for anyone at any time; i.e., there would have to be no exclusive occupation of it. And of course that doesn't work technically, does it? IP address blocks have to be uniquely and exclusively assigned to specific users to function on the internet. Which means the address pool is not a commons - end of story. Because IP addresses are exclusively assigned, they can be governed either as common pool resources (in which a governance agency establishes rules regulating the extraction of resource units from a free pool) or as tradeable property (in which holders allocate the resources by making trades among themselves) or some combination of both. All that matters is which method is more efficient and produces more benefits for Internet users. Leave your religious beliefs behind. But after IPv4 exhaustion, common pool governance of v4 space breaks down completely and the best way to ensure efficient utilization of remaining v4 resources is to allow market-based transfers. These transfers should be made as flexible and easy as possible. There is probably no need for holding periods, although they don't seem to do a lot of harm as long as they are 1 year or less. Needs assessment is increasingly arbitrary and pointless in such an environment. I know needs-basis is another item of religious faith in some circles, but the idea that RIR staff can accurately assess "need" given inherent uncertainty about time horizons and technical development, is just wrong. Organizations should be allowed to buy as much of an asset as they think they need, and can afford, in order to advance their business interests. Let the price system sort out who really needs what. It should also be obvious that the market for these addresses should be global, not regional. Thus, I support proposed policy 2012-01, my only reservation being that the policy specifies that "Where the recipient is in another region, the conditions on the recipient as defined in the counterpart RIR's transfer policy at the time of the transfer will apply." In this blog post I raised some issues with that and hope that the authors of the policy would respond. http://www.internetgovernance.org/ Milton L. Mueller Professor, Syracuse University School of Information Studies Internet Governance Project http://blog.internetgovernance.org

11 15

In support of 2012-01, and side questions.
by Tiberiu Ungureanu 12 Apr '12

12 Apr '12

Hello. First, I want to show support for 2012-01. Second, I read the treads regarding this proposal, and I've seen some opposition, but I don't understand the arguments of the opposition, so I would like the people who oppose to clarify their positions: - What are the dangers that we face if we approve this proposal? * Resources that are available in another region will be made available to RIPE members? Isn't that a good thing? * Resources currently assigned to a RIPE member will leave RIPE influence zone, the RIPE member will potentially get some cash that *maybe* will help that member spend money to deploy IPv6? Or maybe that member will take the money and run to the first casino and waste them all (but wouldn't that in the end help the economy?) * Resources that a RIPE member "sold" to an entity outside the RIPE region, instead of returning them to RIPE could have been used by another RIPE member that now needs them and doesn't have them available. This other RIPE member is now forced to use IPv6, because there's no IPv4 left. But isn't that what we all want? It is not my intention to feed the trolls, but I find 2012-01 very reasonable, and I don't understand the opposition to it. So I allow for the unlikely situation ( ;) ) where I may be wrong. Am I? -- Tiberiu Ungureanu ro.ines

1 0

Re: [address-policy-wg] 2011-04 Last Call for Comments (Extension of the Minimum Size for IPv6 Initial Allocation)
by Anfinsen, Ragnar 11 Apr '12

11 Apr '12

+1 MVH/Regards Ragnar Anfinsen > -----Opprinnelig melding----- > Fra: address-policy-wg-bounces(a)ripe.net [mailto:address-policy-wg- > bounces(a)ripe.net] På vegne av Emilio Madaio > Sendt: 4. april 2012 15:22 > Til: policy-announce(a)ripe.net > Kopi: address-policy-wg(a)ripe.net > Emne: [address-policy-wg] 2011-04 Last Call for Comments (Extension of > the Minimum Size for IPv6 Initial Allocation) > > > Dear Colleagues, > > The proposal described in 2011-04 is now at its Concluding Phase. > > You can find the full proposal at: > > http://www.ripe.net/ripe/policies/proposals/2011-04 > > > Please e-mail any final comments about this proposal to > address-policy-wg(a)ripe.net before 2 May 2012. > > > Regards > Emilio Madaio > Policy Development Officer > RIPE NCC >

5 9

Re: [address-policy-wg] 2011-05 New Draft Document Published (Safeguarding future IXPs with IPv4 space)
by Gert Doering 11 Apr '12

11 Apr '12

Dear Address-Policy WG, (cc'ing the EIX wg due to "this is where the proposal came from") after a very lifely debate in the early stages of this proposal, you have been VERY quiet in this review phase. Specifically, *no* comments have been voiced, and this is not enough for the proposal to go anywhere - so unless I see a few more comments really soon now, we'll have to extend the review phase. Please let us know whether this version 3.0 of the proposal is what you want to see become policy, or whether you oppose it. If you oppose it, please make clear whether you oppose the general idea or just specific aspects of the proposal as written now. thanks, Gert Doering, APWG chair On Tue, Mar 06, 2012 at 04:41:23PM +0100, Emilio Madaio wrote: > > Dear Colleagues, > > The text of RIPE Policy Proposal 2011-05, "Safeguarding future IXPs > with IPv4 space", has been revised based on the community feedback > received on the mailing list. We have published the new version > (version 3.0) today. > > Highlights of the changes in version 3.0 are: > > -a new punctuation is used in the first bullet point of the proposed > section 5.6.2 > > > As per RIPE document ripe-500, "Policy Development Process in RIPE", > the suggested change was not considered significant to require a new > Discussion Phase. Hence the proposal moves to the Review Phase of the > RIPE Policy Development Process. > > The draft document for the proposal has been published. The impact > analysis that was conducted for this proposal has also been published > > > You can find the full proposal and impact analysis at: > > http://www.ripe.net/ripe/policies/proposals/2011-05 > > and the draft document at: > > http://www.ripe.net/ripe/policies/proposals/2011-05/draft > > > We encourage you to read the draft document text and send any comments > to address-policy-wg(a)ripe.net before 3 April 2011. > > Regards > > Emilio Madaio > Policy Development Officer > RIPE NCC > > Gert Doering -- NetMaster -- have you enabled IPv6 on something today...? SpaceNet AG Vorstand: Sebastian v. Bomhard Joseph-Dollinger-Bogen 14 Aufsichtsratsvors.: A. Grundner-Culemann D-80807 Muenchen HRB: 136055 (AG Muenchen) Tel: +49 (89) 32356-444 USt-IdNr.: DE813185279

24 29

2011-05 Last Call for Comments (Safeguarding future IXPs with IPv4 space)
by Emilio Madaio 11 Apr '12

11 Apr '12

Dear Colleagues, The proposal described in 2011-05, "Safeguarding future IXPs with IPv4 space", is now at its Concluding Phase. You can find the full proposal at: https://www.ripe.net/ripe/policies/proposals/2011-05 Please e-mail any final comments about this proposal to address-policy-wg(a)ripe.net before 9 May 2012. Regards Emilio Madaio Policy Development Officer RIPE NCC

1 0

2011-05 Decision to move to Last Call (Safeguarding future IXPs with IPv4 space)
by Gert Doering 10 Apr '12

10 Apr '12

Dear AP WG, 2011-05 (Safeguarding future IXPs with IPv4 space) has come to the end of the Review Phase. Sander and I have gone through the mails sent by you in the discussion phase and review phase, categorizing everything into "support", "opposition", "side track discussion", and so on - as usual :-) - all the details are listed below. Overall, there was very strong support for the proposal, with a fairly high number of supportive e-mails. In the discussion phase, we had a number of people that wanted to see textual changes, or required clarification on details. Some voices had reservation or opposed the proposal initially but changed that to "neutral" or "support" after discussion with the chairs. Two people opposed on the principle of "no special rules" - more to that later. In review phase, 18 individuals expressed their explicit support (some of them relaying support from industry bodies, but this did not influence the decision process). One person sustained his opposition, based on "there should not be a special rule for IXP/ISPs" and "only IXP folks are supporting their support, non-IXPs are underrepresented and would otherwise oppose the proposal". Indeed, many of the voices supporting the proposal have come from people working for IXPs, but I see a number of supporters that I know do *not* work for an IXP - and I also know that there are lots of non-IXP people on the address policy list who are well-capable to express their opposition to a policy proposal. So that part of the argument doesn't hold. Remains the counter-argument of non-supporting special-cases for IXPs - and indeed, the address policy could be much more simple if all consumers of IP addresses were equal. They aren't, and thus we allow special cases if there is enough support for it. Counting voices, I see very strong support. So, we think that we have enough support to call consensus, and move to Last Call. Emilio will do the formal announcement from the PDO soon. If you disagree with our interpretation of what has been said, and the conclusion we have drawn from it, please let us know. Gert Doering, Address Policy WG Chair ----------------------------------------------------- Discussion Phase pro James Blessing Babak Farrokhi Brian Nisbet Remco Van Mook Martin Millnert Sascha Lenz (even if "not a fan of special entities") Niels Bakker Tom Hodgson "worried about special cases, but still supports policy" Henk Steenman Wolfgang Tremmel Larisa Yurkina Harald Michl David Freedman Job Snijders Vincent PAGES Barry O'Donovan Michiel Appelman Steven Bakker Tom Bird Erik Bais Jan Zorz Carlos Friacas David Croft Basile Bluntschli Rolf Schärer Daniel Holme Kjetil Otter Olsen Sascha Luck Nick Hilliard (with some reservactions about the "must do IPv6" bits) against Florian Weimer "IXPs shouldn't be special-cased" chrish(a)consol.net "no special treatment... 'no'" "this should not be necessary" Nigel Titley (later agrees with intent of proposal) Jim Reid ("reservations but no strong objection") clarifications Dan Luedtke ("which /16 is used?"), answered by proposer Remco Van Mook (sizing criteria?), answered by proposer and chair, text integrated into next version Martin Millnert (what about multiple peering LANs?), answered by proposer Sascha Lenz (renumbering time?), answered by Niels Bakker Nigel Titley ("shouldn't be the every-lir-gets-a-/22 rule be good enough?"), +1'ed by Jim Reid, answered by proposer Henk Uijterwaal ("will v4 IXPs be needed after v4 runout?") side-track "why does an IXP have to return PI and not PA?" Florian Weimer, Niels Bakker side-track "shouldn't this be limited to non-commercial IXPs?" Erik Bais, Andy Davidson, Jim Reid side-track "address assignment must be fair, no special rules for IXPs, and non-IXP-folks are not properly represented at APWG anyway" chrish(a)consol.net, Nick Hilliard (multiple mails) Review Phase, v3.0 with textual changes suggested in discussion phase pro James Blessing Sascha Luck Jan Zorz Sascha Lenz Remco van Mook Florian Hibler (needs more coffee) Paul Thornton Arnold Nipper Job Snijders Alexander Leefmann Lindqvist Kurt Erik Christian Kaufmann Kjetil Otter Olsen Petr Jiran Jérôme Fleury Jan Torreele John Souter Sylwester Biernacki against chrish(a)consol.net side remark Turchanyi Geza "it's good that the WG affected is supporting the proposal" (side track discussion about organizations or individuals speaking in the APWG) -- have you enabled IPv6 on something today...? SpaceNet AG Vorstand: Sebastian v. Bomhard Joseph-Dollinger-Bogen 14 Aufsichtsratsvors.: A. Grundner-Culemann D-80807 Muenchen HRB: 136055 (AG Muenchen) Tel: +49 (89) 32356-444 USt-IdNr.: DE813185279

1 0

On use of the word "sold" (was: 2012-01 New Policy Proposal (Inter-RIR IPv4 Address Transfers))
by John Curran 07 Apr '12

07 Apr '12

RIPE Policy Folks - I have no comments to make with regards to the merits of policy proposal 2012-01, but will comment for sake of clarity on the use of the words "sell" and "sold" in the proposal. In the policies in the ARIN region, we have made use of the term "transfer" for this purpose, although one can argue that "sell" and "sold" is equally accurate. The challenge with using the word "sell" is simply that for many readers it carries the connotation of 'free and clear' transfer of property to another. I do not know, (nor intend to assert) what transfer of number resources in the RIPE region means, but in the ARIN region we hold that there are multiple parties with rights to each number resource, and this includes the resource holder (who has the right of exclusive use) but also includes the community (who has the right of visibility into the public portion of the registration, the right to set policies by which the resource may be transferred or reclaimed, etc.) We've had parties come to ARIN and indicate that they have "bought" or "sold" their address blocks, and we have to point out that they may indeed have sold their particular rights to the address block, but that any transfer in our region must comply with the community policies, and ask that they put a transfer request (agreed to by the seller) so that we may transfer the number resources to the new registrant. In short, using terms like "sell" and "sold" in the policy text may create the impression for some readers that Internet number resources are just widgets to be bought and sold as desired, but that error is may be unavoidable in any case given the circumstances. I do not believe that the use of "sell" or "sold" is incorrect, but simply of question of personal judgement as to whether the improved readability is worth the connotations that some readers will draw as a result. FYI, /John John Curran President and CEO ARIN

6 9

2012-01 New Policy Proposal (Inter-RIR IPv4 Address Transfers)
by sandrabrown＠ipv4marketgroup.com 06 Apr '12

06 Apr '12

2 1

2012-01 New Policy Proposal (Inter-RIR IPv4 Address Transfers)
by sandrabrown＠ipv4marketgroup.com 05 Apr '12

05 Apr '12

Sorry, in plain text this time.... Thank you for your comments on proposal 2012-01, which I have submitted. I will undertake to speak to several of the points you have raised. My opinion on the question of the use of the word "sold" is that I would also be ok with the word "transfer". I believe this policy can cover two situations: 1) where related companies, for example ABC-NA transfers a block from ARIN to ABC-EMEA to RIPE-NCC, and no money is exchanged; and 2) where unrelated organizations transfer a block. Only in the second case is money likely to be exchanged in order to encourage the sending organization to initiate the transfer. Thus only the second case is truly a sale. And let's be very clear that I do mean for money to be exchanged: in most cases there is work to be done to free up blocks of IPv4 addresses for re-allocation, and the incentive for entities to do so is financial benefit. Otherwise, large portions of their blocks will continue to sit dormant and unused. The entire purpose of the IPv4 address market is to free up dormant, unused IPv4 addresses to fill needs. A number of people have questioned the 15 month window before allowing a buyer to re-sell. I have had suggestions ranging from 12 months to 36 months. I think the point is that there is some feeling that an entity could acquire IPv4 addresses with the intent to flip them for profit. I am not sure this would be possible since at some point, the transition to IPv6 will make IPv4 addresses a declining value asset. I, for one, believe in a free market for IPv4 addresses, where no holding period is required. In a free market for IP's, the price of IP's would move up and down like the price of a commodity. However, there are artificial trading rules at play with IPv4, so some window seems to be needed. Hence, I chose 15 months as a compromise for those who feel that 15 months allows a company to satisfy its IPv6 conversion needs and possibly be done with IPv4 and be ready to sell them. To Mr. Blessing's question about an LIR changing its policy, I believe the answer is that there would be a legal binding agreement between the buyer and the seller that would cover this situation. The RIPE-NCC policy does not need to cover the legalities between the two parties. I am not sure what the question of a reverse policy refers to? Do you mean to reverse the transaction? If so, you could add a clause that states that the two parties can reverse the transfer at any time in the future or something to that effect. I did have a hypothetical discussion with Sanjaya of APNIC for a situation that may require this. I was asking about a potential lease situation where an APNIC member may wish to lease a block from another RIR. He wondered if the APNIC registry could initially be updated to reflect the lease, and then once the APNIC member had completed the upgrade to IPv6, the IPv4 addresses could be returned to the originator. We then wondered how to reverse the transfer to return the IP's to the owner. This clause above that Mr. Blessing is possibly questioning about would take care of lease returns. I believe these transfers would always be between entities, ie. LIR's. Mr. Hilliard is concerned about Afrinic losing IP's. Frankly, I have not looked at oversupply or demand in the Afrinic region. Right now Afrinic does not have an Inter-RIR transfer policy and by including a statement that RIPE-NCC will only transfer with RIR's that have a policy, you allow Afrinic to take care of itself. Mr. Lenz is concerned about the use of the word "business" purposes. That is, perhaps, a semantic point. I believe that the word business should not be taken too narrowly. Business may be defined as an activity a person or organization is engaged in, and I would just take it at that. Mr. Leefman suggested we realign the proposal to a /21 as the minimum size, and I think that is an excellent suggestion. Other RIR's such as ARIN and APNIC work with a minimum size of a /24 and I was trying to keep RIPE in line with them, but frankly that is such as small block, why bother? Mr. Leefman also suggested changing the requirements for proving need to align with the intra-RIR policy. In checking with Mr. Madaio, he informs me that there have been no intra-RIR transfers to date. That suggests to me that RIPE's intra-RIR policy needs some rework, if I may be blunt. My experience with transfer markets in other RIR service areas has shown three things: 1) that entities need to be assured that their transfers will be approved, when they put money on the line, 2) that when they pay, they don't want, or believe that they should have to, go through the same needs justification process as when they apply for free allocations, 3) while the period of a free allocation may be fixed (ie 3 months), when a business has a need and a business plan for which it acquires IP's it is a much longer time horizon, and it is not RIPE-NCC's role to limit its members business planning horizon to 3 months. (If RIPE approves me for a three month supply now, I have no guarantee I will get another 3 month supply in 3 months....I cannot plan or operate this way!). For example, and you may not care about what other RIR's do, but ARIN has a 24 month supply for a transfer, and as you may know, we are working to convince ARIN to modify its policy on needs justification for transfers. So, I would not support a change in the inter-RIR policy to match your intra-RIR policy. I would instead suggest that you consider modification of your intra-RIR policy to match what your constituents need. Lastly, I like Mr. Van Mook's comments, other than that he opposes the policy. It is excellent and simple to outline the transfer in and transfer out options, and as noted above, I concur with him that the RIPE intra-RIR policy needs clean-up. I guess I thought that was a bigger mountain than I could climb, but if it is possible to have a conversation about doing it at the same time, I am all ears. Again, thank you for your support, your comments, and your thoughtful feedback. Respectfully, Sandra Brown

1 0

Re: [address-policy-wg] 2011-04 Last Call for Comments (Extension of the Minimum Size for IPv6 Initial Allocation)
by Sascha Lenz 05 Apr '12

05 Apr '12

Hi all, > > Dear Colleagues, > > The proposal described in 2011-04 is now at its Concluding Phase. > > You can find the full proposal at: > > http://www.ripe.net/ripe/policies/proposals/2011-04 > I didn't have a change of mind lately, still happy to support it in this form. (Yes, i'm selfish, "more bits to brag with^W^Wmake deployment easier in some cases", no downsides - /29 is reserved anyways for existing LIRs) -- Mit freundlichen Grüßen / Kind Regards Sascha Lenz [SLZ-RIPE] Senior System- & Network Architect

1 0