[arin-ppml] Fairness of banning IPv4 allocations to some categoryof organization

[Milton L Mueller]

> The point that seems to me to be missing from the commentary in this
> thread is that the argument Milton and the ITU "experts" make about
> management of IPv6 addresses based on issues with IPv4 addresses is a
> classic non sequitur fallacy.  Because of the huge increase in the
> size of IPv6 addresses, conservation of address allocation is not the
> issue it is with IPv4 addresses.  Arguments about relative scarcity,
> allocation efficiency, recovery, and transfer that are contentious
> with IPv4 addresses simply do not apply when there are so many IPv6
> addresses.  

Ah, so you haven't read the reports either, have you John? 
You might want to look at Section 2.1.2, where various arguments that we don't have to worry about reclamation and scarcity are carefully reviewed and discussed. My assessment that we DO need to pay attention to such concerns in ipv6 was influenced by some persuasive arguments and calculations made by, um, Geoff Huston (or some variant thereof) as well as Thomas Narten. Indeed, I showed that their calculations completely ignored the need for reclamation, and when you don't ignore the problem of reclamation the projected consumption can rise by a factor of four. 

I will agree with you that underlying many ITU attitudes, and especially the idea of block reservations for nation-states and allocation by CIRs, is an retrograde, IPv4-era fear of being left behind. I of course do NOT favor CIRs as the model and explored flexible markets as the alternative. However, the concern about a replay of IPv4-era scarcity received reinforcement from some guy named, um, Geoff Huston (or some variant thereof) in a 2005 report which said of existing v6 allocation policies: "we stand the risk of, yet again, visibly creating an early adopter reward and a corresponding late adopter set of barriers and penalties." 

>The technical issue of conserving routing entries in the default-free 
>zone is still relevant, but plays little or no part in the ITU arguments

So once again, you make statements without having actually read the "ITU arguments". Of course route entry conservation is relevant, extremely relevant. And it is discussed in both reports. If you had read mine, you would know that one idea it explored was that of "aggregation fees", i.e. increasing the fees for address blocks based on the number of routing entries announced within them. This was and remains an interesting idea, because earlier ideas about charging for route announcements assumed that ISPs would have to bilaterally negotiate such charges, which was a non-started in terms of practical implementation. My idea was that RIR fees could provide a lower transaction cost method of doing this. My first draft actually recommended doing this. But we decided not to recommend it for a couple of reasons. One of them was that I consulted with Geoff Huston (or some variant thereof) who referred me to his studies that Moore's Law would easily allow routing tech to keep pace with routing table growth and who said that such a tax was a "solution in search of a problem." 

There were other reasons not to recommend routing table bloat fees, but if indeed such bloat is a major problem an aggregation fee-based solution remains an idea worth looking at. Lots of interesting definition, monitoring and enforcement issues would have to be resolved, of course; for now the issue is whether this community can have intelligent dialogue about policy options or starts screaming and demonizing people and institutions when challenging new ideas are floated. 

--MM