[arin-ppml] v4 to v6 obstacles

[Joe Maimon]

Lee Howard wrote:
>
>>
>> NAT arrived late to IPv4 and it appears to be arriving late to IPv6 as well.
>>     
>
> Different issues.  I don't see a need for NAT66.  But Address Family Translation
> (AFT, or NAT46 and NAT64) would be useful.
>   

In the context of depletion, address family translation is what I was 
primarily referring to. It certainly is not only late, but has already 
been ejected from the party previously.

In the context of end user networks deciding to deploy ipv6 or not, 
unavailability of NAT66 may very well be a deciding concern, and it very 
possibly will also arrive late to IPv6.

It may also cause NAT46 to gain prominence should enough networks choose 
to skip out on internal deployment for whatever reason. Also late.

> NAPT-PT, being deprecated, is poorly supported.  
And seems to have been removed from various versions of IOS.
> Seems like people are holding
> their breath for NAT64, and maybe someday somebody will submit a draft for 
> NAT46.  Which is too late to get gear fielded.  And doesn't solve most of the
> issues identified in RFC4966.
>   
Even the NAT haters should be able to come together on this. Because the 
only non deprecated standardized workable alternative is dual stack with 
rfc1918 and more NAT.

And while I think current AFT efforts are being well spent, its from 
behind the 8-ball.
> Is it possible to "reprecrate" something that has been deprecated?  Can we 
> obsolete RFC4966?  Anyone have a pointer to the right list for that?
>
> Lee

I suspect that only in the absence of proper user demand to help steer 
the ship do vendors pay such close attention to RFC document status. I 
suppose anyone can go join an IETF WG and write a RFC draft.

Joe