[arin-ppml] Policy Proposal 2008-6

Stephen Sprunk stephen at sprunk.org
Wed Dec 31 23:02:37 EST 2008


Joe Maimon wrote:
> Suppose class E,

This has already been discussed to death, and the consensus is that we 
could not get patches to routers, CPE, hosts, embedded devices, apps, 
etc. out and applied widely enough in time for it to matter.  It takes 
5-10 years for major changes in IP stacks to be rolled out, and that's 
simply too late for IPv4 at this point.  Perhaps if someone had started 
this effort in the mid-90s...

> legacy unannounced swamp reclamation,

Once 2007-14 gets through last call and makes it into the NRPM, I will 
be drafting a proposal to expand it to cover legacy space; we will see 
how well the community accepts that, given how long it took to get the 
non-legacy version to be adopted, not to mention the potential legal 
problems...

> tighter policies with smaller allocations producing a fraction of current burn rate

Not really possible; the mega-ISPs are already 80-90% of the burn rate, 
and they're also pretty darn efficient according to policy (though 
reality isn't always perfect).  There's not much to be gained on this 
front, short of actually charging them the same rate per address that 
smaller ISPs pay, and nobody seems to have the balls to do that.

> and incentivized returns/transfers would extend ipv4 life far longer than anybody would want.
>   

If nobody wanted it, then nobody would pay for the addresses.  Despite 
over a decade of work, IPv6 still isn't usable yet, and probably won't 
be for several more years.  I predict it will _not_ be possible to go 
IPv6-only prior to or even a few months after the IPv4 pool is depleted.

> Including long enough to kick in more space from ipv6/NAT conversions.  So in effect, ipv4 integers forever being available is actually possible.
>   

If there is some incentive (or other mechanism) to recover them, yes.  
Markets are very good at producing supply where there is demand, though 
we may not like the price...

> Technically speaking, 25-50% of ipv4 may still be available to be used on the internet, even after pool runout.
>   

How?  That much of the pool isn't even available today.

> Picture explaining the justification for unavailability of every single 
> /8 before a congressional hearing or similar, anyone want that job?
>   

Not me.  That's part of the reason I support reclamation: it's a lot 
easier to defend being out of addresses when you've made some effort to 
get back the ones that we all "know" aren't being used.

> (If there is even the remotest of possibilities of using E space, than reclaiming it should start sooner rather than later)
>   

See above.

S
-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/x-pkcs7-signature
Size: 3241 bytes
Desc: S/MIME Cryptographic Signature
URL: <https://lists.arin.net/pipermail/arin-ppml/attachments/20081231/d59de5c5/attachment.bin>


More information about the ARIN-PPML mailing list