[ppml] [narten at us.ibm.com: PI addressing in IPv6 advances in ARIN]

[Paul Vixie]

> So, either the community says: "no ipv6 until multihoming is 'solved'" or we
> suck it up and start deploying with PI space for the sake of deployment of
> v6.  I'm not convinced that routing PI space will be 'short lived'...

neither am i, yet i remain in favour of the "suck it up" approach you describe.

> > 2) the middlebox approach: separate the end point identifier from the
> > routing locator at the core/edge boundary, leaving the IPv6 stacks at both
> > the routers in the core and the hosts at the edge untouched, but inserting
> > a mapping from end point identifier to routing locator on source edge to
> > core and an unmapping the routing locator back to the end point identifier
> > from core to destination edge.
> 
> 'middlebox' means 'nat host'...

no, it could mean rekindling A6 and DNAME under some different, politically
correct guise.  DNS would be the EIP and the 128-bit address would be the RL.