new IPv6 policy draft - real soon now

[brian@hursley.ibm.com (Brian E Carpenter)]

Daniel,

The point is that unless we are collectively foolish, IPv6 will
not have a disaggregation problem - it starts out classless
and provider-based, and dual homing will in the end be solved
by dual prefixes. I really can't see why we would *need* prefix
length based discrimination as we do for IPv4.

   Brian

Daniel Karrenberg wrote:
> 
>   > Bernard.Tuy@urec.cnrs.fr (Bernard TUY) writes:
>   > ====BT: moreover, one can imagine carriers will be able to aggregate ISPs p
>   > refixes.
>   >     If they've got different legnth ones, I don't know how to achieve this.
> 
> Aggregation is governed by topology and allocation policy.
> Hierarchy helps aggregation. I do not see how equal prefix length
> helps aggregation. Please explain.
> 
> Daniel