[arin-ppml] Draft Policy ARIN-2013-6: Allocation of IPv4 and IPv6 Address Space to Out-of-region Requestors - Revised Problem Statement and Policy Text

Owen DeLong owen at delong.com
Thu Sep 12 22:37:34 EDT 2013


> 
>> Plurality seems like an odd choice of word above. The implication is
>> that if 21% of the equipment for which I use ARIN addresses is in
>> North America, and as long as my use in each of the other four regions
>> is 20% or less, I'm good to go.
> 
> Well more precisely the lowest possible use within the ARIN region is some fraction greater than 20%, with less than or equal to 20% in the other four regions.  While possible in reality, this is much more of a contrived example that something you would expect to see regularly in the real world.  However, if you were only operating within the ARIN region and one other region you would need greater than 50% in the ARIN region and less that 50% in the other region, a simple majority.
> 

As written, I could, for example, have 50.001% in the ARIN region and 49.999% in one other RIR and be within the proposed requirements.

However, I think that in a policy like this, it is important to choose language which does not limit operator flexibility in unintended ways while getting as close as possible to the intended result.

If anyone has language that they believe will better match policy intent (or feels that a different intent would be better for that matter), then please express those ideas here.

>> That doesn't seem to be what the author was trying to achieve, does it?
> 
> I'd agree it wasn't what the authors were originally thinking, but if you review the earlier comments there were several people that objected to a 50% majority, and plurality was suggested as an alternative, as discussed in the Advisory Council Comments sections.
> 

Majority is certainly more problematic than plurality. Plurality might not be the best possible choice, either, but nobody, including myself, has yet proposed a better alternative. The AC would certainly welcome any improved language from the community if anyone has a better idea.

Owen





More information about the ARIN-PPML mailing list