[ppml] Proposed Policy: IPv4 Countdown

[Kevin Loch]

David Conrad wrote:

> http://nutss.gforge.cis.cornell.edu/pub/ieee-nutss.pdf

This suggests that automatic tunneling technologies and associated
id/locator isolation would make IPv6 adoption unnecessary. I don't
know about that but it does have the potential to improve routing
scalability regardless of the type of address space used.  Of all
the routing scalability ideas that have been explored this seems the
most promising to me, though not necessarily the methods outlined in
that paper.

It could also be sold as a feature upgrade to replace NAT, even with
ordinary IPv4 space. That alone makes it worth pursuing.

But if you're going to tunnel anything it might as well be v6 for the
extended addressing.  You could also tunnel a "2nd" fresh ipv4 space 
that is only valid within the tunneling system.  That would sill be
too limited for future needs and would also be very confusing.

- Kevin