RE: Last Call: <draft-weil-shared-transition-space-request-14.txt> (IANA Reserved IPv4 Prefix for Shared Address Space) to BCP

[John C Klensin <[e-mail]>]

--On Thursday, February 09, 2012 15:40 -0500 Ronald Bonica
<[e-mail]> wrote:

> SM,
> 
> At NANOG 54, ARIN reported that they are down to 5.6 /8s. If
> just four ISPs ask for a /10 for CGN, we burn one of those /8s.
> 
> Is that really a good idea?

Ron,

I've mostly been staying out of this since an earlier round, but
that question/ scenario seems pretty irrelevant to me.    If
four ISPs ask for (and can justify) /10s to serve customers, or
a larger number asks for smaller allocations that add up to the
same thing, we burn a /10 equally quickly.   I think the real
answer to questions like the one you ask is another question:
Suppose there is a strategy that delays, by (say) six months or
a year, the date that ARIN (for example) runs out of
useful-sized blocks to allocate, it is worth doing something
strange or pathological just to gain that extra time?  I think
the answer to that question is "no", regardless of how one
defines "strange or pathological", because a year of traditional
IPv4 allocations from the RIRs one way or the other is quite
unlikely to make a major different to the Internet.

I'm not claiming this proposal is either strange or pathological
(or that it is not).  But I suggest that the decision be made on
the same basis that other allocation decisions are made, not by
trying to save a few months or by deciding that some types of
applications or applicants are inherently more privileged than
others regardless of their actual merits. 

      john