[arin-ppml] Discussion Petition of ARIN-prop-125 Efficient Utilization of IPv4 Requires Dual-Stack

[Leif Sawyer]

Jimmy Hess wrote:
> in response to Leo Bicknell whom wrote:
> [snip]
>> I strongly object to this petition for the same reason I 
>> objected to the original proposal, the dual stack requirement.
> 
> This should be discussed further. 
[snip]
> IPv6 network deployments may be structured differently and 
> require fewer V6 interfaces than interfaces on the V4 network.
[snip] 
>  - for every 100 IPv4 addresses requested, at least one 
> pre-existing interface is dual stacked, OR the IPv4 address'  
> purpose is to provide at least one network service that is 
> identical in purpose and content to a service available over IPv6.

This is an interesting quandry, and yet it does nothing to help
service providors whom are at the mercy of vendors that have done
nothing to support IPv6 within their equipement.


"Buy new equipment"  you say.    Certainly, for a handful of routers,
this is not an insurmountable task.

But again, for service providors,  what is the drawn line for replacing
cablemodems or DSL modems?  10,000?  100,000?  1,000,000?

And what about your mobile phone?  Are you going to march down to
AT&T, T-Mobile, Sprint, or any other carrier and demand that they hand
over a shiny  IPv6-capable phone?  Can you find any within the US?
I'd love for my Nexus One to support IPv6 (sure, the kernel -could-,
but the rest of the stack doesn't)

And what then of the Ericksson GSM stack?  Do you expect them to forklift
upgrade every GGSN, SSGN, and the rest of the infrastructure?



If anything, it would be best served to have the condition be graceful and
hopeful:

At a minimum: IPv6 address space assigned and deployed to the core infrastructure, with
either public facing services or test-only services active.