* Tore Anderson
(Just to be clear: I am not stating neither support or opposition to the proposal.)
I've slept on it, and I support this proposal. Allowing End Users access to the last /8 is, in my opinion, "The Right Thing To Do". While I think the low cost of PI space compared to PA space is going to cause a high risk for abuse/circumvention of the "one piece each" restriction, this is ultimately something that needs to be solved outside of address policy. If a new version of the proposal is going to be spun, I have the following suggestions for modifications: 1) The proposal should «voice an opinion», as Gert put it, that the fees for obtaining a PI prefix from the last /8 should be comparable to obtaining a PA one through becoming a LIR. I hope that this would make it less likely that the "one piece each" restriction would be abused/circumvented - at least, not more likely than the current restriction being circumvented through multiple LIRs. 2) The /24 limit for assignments should be raised to /22. While the current proposal does reduce discrimination of End Users compared to LIRs, it does not remove it completely. I believe that "Doing The Right Thing" would be to make access to the last /8 completely equal, instead of continuing to give preferential treatment to LIRs. As I've pointed out earlier, both of those amendments would make the RIPE last /8 policy (and pricing) more aligned with APNIC's, which appears to work fine for them. They've used 8.7% of their last /8 over the 17 months it's been since they ran out. As I understand it, in their last /8 policy, assignments and allocations are both capped at /22. Regardless of it being an assignment or an allocation, it costs the same - and significantly more than the RIPE €50: http://www.apnic.net/services/become-a-member/how-much-does-it-cost http://submit.apnic.net/cgi-bin/feecalc.pl?ipv4=%2F22&ipv6=&action=Calculate Best regards, -- Tore Anderson Redpill Linpro AS - http://www.redpill-linpro.com