a6 wasn't even a start of a solution to the renumbering problem.
at best, it was a distraction, and a disaster waiting to happen.
i agree to disagree.
note that a6 wouldn't have done anything to relieve the burden
associated with the case you cited - a host changing its network
interface.
true enough, but you can bet that if there were dns support for
the macroscopic problem, then i would be pushing some kind of dns
support for the microscopic problem as well. as it is, there's
no foundation for it, other than campus-wide /128 "host" routes
which would scale poorly.
--
Paul Vixie