[arin-ppml] Change of Use and ARIN (was: Re: AFRINIC And The Stability Of The Internet Number Registry System)

Owen DeLong owen at delong.com
Thu Sep 9 15:35:30 EDT 2021

> On Sep 8, 2021, at 18:15 , Fernando Frediani <fhfrediani at gmail.com> wrote:
> On 08/09/2021 17:56, Owen DeLong wrote:
>> <clip>
>> ICP-2 has no relevance once an RIR is accredited. It is a document defining the process for accrediting a new RIR and it only governs ICANN’s process for doing so.
>> If ICANN is replaced and the IANA role is granted to some other institution, ICP-2 would entirely cease to be relevant unless somehow adopted or imposed upon the new organization performing the IANA function.
>> This continued citation of ICP-2 as more than it is doesn’t make it so. It remains a document governing a particular ICANN process and nothing more.
> That your view and only yours as some others that it seems to be exclusively from you. But fine, as usual you view.

Can you cite anyone other than a handful of malcontents from the AFRINIC list who has backed this up? Can you cite anyone from an RIR staff or board outside of AFRINIC that will back this claim?

> ICP-2 is not something to be used one off and be trashed right after and contain principles to be observed an followed not only during the creation of a RIR, otherwise we, as community would be in a very bad shape with that new RIR being created, empowered with such     critical functions and their Board afterwards doing whatever they like at the expense of shared resources and functions that were recognized for them not just at the moment of their creation but at any point.

No… It is a policy and procedure to be used each and every time a new RIR is accredited by ICANN after it was enacted. So far, that’s been exactly twice IIRC. Once for LACNIC and once for AFRINIC.

> If it says there that it must have a bottom-up self-governance structure for setting local policies, how can you imagine that would be valid only at the moment of the creation of the RIR ?

Since it is a policy and procedure for accreditation of an RIR, those requirements are present at the time of accreditation. Since there is no policy or procedure to review such an accreditation after it is issued and no way to revoke or alter such an accreditation, once the accreditation is granted, ICP-2 becomes meaningless in the context of each already accredited RIR.

Is this a defect in ICP-2? Perhaps.
Should there be some such process for revoking accreditation? Perhaps.
Should RIRs have some criteria or ongoing requirement that conforms to some of the provisions of ICP-2 that they must continue to comply with in order to keep their accreditation? Perhaps.

However, at the current time, no such mechanisms or requirements exist, whether you like that fact or not.

If you want to start working on creating them, I’m probably willing to help in that effort if they are along what I would consider rational lines.

Personally, I think that the RIR boards have too much power and too much control over their respective RIRs and that the RIRs are not sufficiently accountable to their communities. However, I also recognize that there are issues and concerns with the potential for organizational takeover and that there’s a bit of a delicate balance between those two.
>>> They are entiteled to their opinion but I do not believe that corrensponds to practical realitty.
>> Then your “practical reality” is divorced from actual reality, which doesn’t really come as a surprise to me.
> RIRs legal prerogatives at not above certain and wider mechanisms the community have to refuse unilateral and not bottom-up decisions that concern the community. If that ever happen there are different mechanisms to overwrite that regardless the legal system where the RIR is placed. That is the practical reality I was referring to.

The only mechanism that I see for the community to refuse unilateral and not bottom-up decisions that concern the community is to convince a sufficient critical mass of those who run routers to start believing a different registry system. After all, nothing prevents anyone. who wants to from creating a new registry system and issuing any addresses they wish. Getting those who run routers to listen to your registry system instead of the RIRs is, however, a rather fraught exercise to be sure. The absolute worst possible outcome is competing overlapping registries with different subsets of people running routers listening to different competitors. If we are to replace the RIR system with a new registry, it must be done in such a way and through some form of process that avoids a loss of uniqueness of registrations and that would require convincing at least a significant critical mass (say 80%+) of routers to switch allegiance on a particular flag day or similar process. I think that would be difficult to achieve even if every RIR did something truly egregious (which is unlikely).

>>> I sincerelly hope that not only ARIN Board by any other RIR Board never void the bottom-up process and respect the ultimate power of community to choose how policies will be, not the Board unilaterally at their will.
>> There’s a whole lot of grey between the black and white extremes you describe above. The reality is well inside of that grey area and neither of your extremes represents reality.
>> In every region (except possibly RIPE), the board has the ultimate authority and control over the policies. In every region, the board is elected by the membership. In every region, while the board has ultimate authority, there is a PDP which to varying degrees provides input into how the board manages the policies, but ultimately, (with the possible exception of RIPE as I have not reviewed their precise process), the board must ratify all policy proposals prior to implementation.
> That's a narrow view. In practice they don't have. As mentioned there are mechanisms available to overwrite Board decisions on community exclusive matters. This doesn't confuse with the prerogative of the Board to ratify policy proposals prior do implementation.
> I am also not referring to emergency  policies. Those are fine to exist for justified reasons but they still are subject to community evaluation and final say and if the community doesn't want that they shall cease to exist at some point, despite the most noble and good intention from the Board which created it may have had.

What, exactly, are those mechanisms short of outright rebellion and the creation of an alternate registry system?

This has been done in DNS and it has not worked out particularly well for the alternative root zone advocates.

>>> Obviouslly this doesn't confuse with the prerrogative of the RIR Board to care about the organization protection and legal protection and I support that including the prerrogative of the Boards to ractify proposals that reached consensus.
>> What about the prerogative of boards to not ratify harmful proposals which reached consensus?
>> If they can’t choose not to ratify, then it’s not a prerogative, it’s a rubber stamp.
>> I believe that on RPD, you were one of the people arguing that the board must ratify the “board prerogatives” proposal despite it being harmful to the organization and an end-run on the process of updating the bylaws which is where most of the intended policy belonged.
> Not exactly.
> What I say is that if the Board is going to deny the ratification of a proposal that community reached consensus they must give detailed justification for that specially on how that may be harmful to the RIR. Ratification process is not for Board members do a second round on policy evaluation and they may choose to not ratify just because most of them 'don't like it'. They are not there to like the proposals or not, but simply to protect the organization. Therefore they may reject ratification if they are able to justify that may be a harmful thing to the RIR. I agree that is a very subjective thing, but the main point that the duty of the Board to provide clear and enough justification for such event as we know how long and detailed can be a process to reach consensus on something.

Which has been done in each and every case that I am aware of. Yet I seem to recall you still advocating against the AFRINIC board’s legitimate denial of the “Board Prerogatives” proposal.

So on one hand, you say that the board does have the prerogative to reject a harmful or out of scope proposal and on the other hand, you are arguing that you should be allowed to substitute your (and possibly the community, though I think only a small minority of the community actually backs the proposal at this point) judgment for theirs and override their rejection. I think it is absurd to claim you can have it both ways.


More information about the ARIN-PPML mailing list