[arin-ppml] IPv4 Depletion as an ARIN policy concern

[Scott Leibrand]

William Herrin wrote:
>>> 5. Due to a poor architectural decision by the IETF (IPv6 first, fall
>>> back to IPv4), I can't make effective use of IPv6 *at all* unless I
>>> deliberately ignore #4 and choose to accept degraded functionality on
>>> my system.
>>>
>>> If not for #5, IPv6 would have a much easier time getting past the
>>> "worthy enough to deploy" barrier.
>>>       
>> I disagree with you.  I think there's a larger population who have been
>> hearing about IPv6 for so long that they've decided it will never
>> happen (or who are otherwise ignorant of IPv6), and the only way to
>> convince them that networks will use IPv6 is if networks use IPv6.
>>
>> Otherwise, nobody will know if IPv6 is working until IPv4
>> connectivity is shut off, which is a little late to find out.
>>     
>
> You can force me to choose between IPv4-only and
> IPv6-fall-back-to-IPv4 but you probably can't force me to choose the
> latter and probably won't like the lip-service I give it if you try.
>
> Systemic effects of IPv4-fall-back-to-IPv6 notwithstanding,
> IPv6-fall-back-to-IPv4 is a powerful disincentive to early movement on
> IPv6.
>   

Standardization issues aside, I think there is definitely some 
interesting work being done by pragmatic operator types to work around 
this problem, as was shared at NANOG 47 in Dearborn.  However, that 
seems mostly OT for this list.

In any event, do you see any particular areas where we could potentially 
improve address policy in light of the issues raised in this thread?

Thanks,
Scott