[address-policy-wg] Consensus on 2015-01, "Alignment of Transfer Requirements for IPv4 Allocations"

Hi working group,

Lu Heng asked me to rectify my summary. He has also expressed support for the policy proposal during the review phase and his name was indeed not listed. Please consider him a supporter of this proposal as well.

Cheers,
Sander



> Op 22 jun. 2015 om 14:06 heeft Sander Steffann <sander at steffann.nl> het volgende geschreven:
> 
> And this time with a fixed subject line so that it is clearly visible which policy proposal we are talking about :)
> 
>> Op 22 jun. 2015, om 12:10 heeft Sander Steffann <sander at steffann.nl> het volgende geschreven:
>> 
>> Hello working group,
>> 
>> Here is your chair's (singular, Gert has abstained from judging consensus as he became too involved in the discussion on the mailing list and might be seen as non-neutral on this policy proposal) analysis on the review phase of RIPE policy proposal 2015-01.
>> 
>> At the end of the review phase there was a sudden flood of messages both supporting and opposing the policy proposal. Many of these messages were on or after the deadline: the end of the review phase. As those messages didn't bring forward any new arguments they didn't influence my decision making process. I have included them in this overview for completeness' sake.
>> 
>> First the people supporting this policy proposal. There were many people who supported the proposal based on the rationale given in the proposal itself (also known as "+1" messages). Others also stated the reasons why they supported the proposal. These included:
>> - It aligns with original intent (make assignments) of the final /8 policy
>> - It makes it less profitable to overtly act against the original intent of the final /8 policy
>> - It is a good step in the right direction, we may need more steps later
>> 
>> Here is a list of people that supported this policy proposal:
>> - Andre Keller
>> - Andreas Larsen (after deadline)
>> - Carsten Brückner (after deadline)
>> - Carsten Schiefner
>> - Christopher Kunz
>> - Daniel Suchy
>> - Dimitri I Sidelnikov
>> - Erik Bais
>> - Florian Bauhaus
>> - Garry Glendown
>> - Gerald Krause
>> - Havard Eidnes
>> - Herve Clement
>> - Jan Ingvoldstad
>> - Jens Ott
>> - Marius Catrangiu
>> - Martin Millnert (after deadline)
>> - Mick O Donovan
>> - Ondřej Caletka
>> - Radu-Adrian FEURDEAN
>> - Riccardo Gori
>> - Richard Hartmann
>> - Robert Sleigh
>> - Sebastian Wiesinger
>> - Thomas Schallar
>> - Tim Kleefass
>> - Tom Smyth (after deadline)
>> - Tore Anderson
>> - Torunn Narvestad (after deadline)
>> - Vladislav Potapov
>> 
>> David Freedman asked for clarifications about the impact on the Mergers and Acquisitions procedure of the RIPE NCC. These were answered by Marco Schmidt.
>> 
>> Daniel Baeza and Richard Hartmann asked for clarifications on how this policy would be applied to allocations made in the past. Marco Schmidt explained that if accepted this policy would only impact transfers happening after the policy was implemented. Transfers that happened in the past would not be impacted. The policy would be applied to existing allocations though. Allocations made in the past would not be transferrable until they were at least 24 month old. For some people this was a problem as they considered it unfair to those LIRs that had started in the last 24 months with the expectation that they would be able to transfer their allocation from the final /8.
>> 
>> The people opposing this policy proposal because they consider it a retroactive change are:
>> - Sascha Luck
>> - Storch Matei
>> - Vladimir Andreev
>> 
>> There were many messages on this topic. We consider this objection handled because this policy doesn't actually change anything that happened in the past. This policy proposal is about the requirements of transfers. If this proposal gets accepted transfers that have already happened stay happened, and transfers that are about to happen will be checked against the current policy at that time. This is how RIPE policies have always been applied and this policy proposal is no different.
>> 
>> There was a message stating opposition to the proposal by Arash Naderpour, but as no reasons against the proposal were given there is not much we can do with this. Consensus based policy development means trying to address objections until the reasons for the objections are taken away. When no reasons are given this is not possible. Therefore this opposition will not have much weight in my analysis.
>> 
>> There was also opposition because people felt that this policy proposal didn't solve a real problem and/or wasn't solving all problems related to abuse of the current final /8 policy. They were:
>> - Amir Mohsen (after deadline)
>> - Aleksey Bulgakov
>> - Arash Naderpour (after deadline)
>> - Borhan Habibi
>> - Ciprian Nica
>> - Olga @ip4market.ru (after deadline)
>> - Petr Umelov
>> - Sergey Stecenko
>> - Storch Matei
>> - Yuri @ntx.ru (after deadline)
>> 
>> During the discussion it was shown that the number of transfers from the final /8 pool was increasing, especially for very "young" prefixes. This shows that there this policy does solve a real problem. As with all policy proposals it is clear that one policy proposal will not solve all the potential problems all at once. That there are still other potential problems related to the final /8 policy is noted as an encouragement for future policy proposal authors.
>> 
>> There were also people objecting because preventing organisations to open a new LIR and then transfer its address space would mean that the membership growth of the RIPE NCC would be a bit lower, and because the RIPE NCC is funded by its members the lower membership numbers might cause the membership cost per member to increase. These were:
>> - Ciprian Nica
>> - Sergey Stecenko
>> - Storch Matei
>> - Vladimir Andreev
>> 
>> The impact analysis by the RIPE NCC however explicitly mentions that "Considering the overall size of the membership, the RIPE NCC does not anticipate a significant impact will be caused if this proposal is implemented.".
>> 
>> Finally, there were also objections that the final /8 pool was too big and/or not running out fast enough. This objection was made by:
>> - Ciprian Nica
>> - Storch Matei
>> 
>> In the impact analysis however mentions that the current pool will last 5.5 years based on the allocation rate of the last 6 months (up to the writing of the impact analysis). That lifetime may be reduced significantly however if new LIRs continue to join in ever-larger numbers and /22 transfers from last /8 also gain more popularity. As the remaining lifetime of the IPv4 internet is extremely likely to be longer than 5.5 years the lifetime of the final /8 pool seems short as it is.
>> 
>> Based on the feedback I see strong support for this policy proposal. All objections seem to be addressed as well, so I hereby declare rough consensus on policy proposal 2015-01 and ask our friendly RIPE NCC Policy Development Officer to move this policy proposal to the Last Call phase.
>> 
>> Sincerely,
>> Sander Steffann
>> APWG co-chair
> 
>