[e2e] end point identifiers, name/identifyer/address separation, end-to-end ness..

[George Michaelson]

This is probably a silly question, but *why* is the endpoint identifier and
identifier/address/name separation discussion taking place on everywhere except
e2e?

isn't this almost 100% on-charter for this forum?

-George

PS my personal un-informed opinion is that we'd be much better off
modelling this explicitly in a dumb network, end-smart way before diving into
discussions on modifying the base IP architecture right now. I also see
parallels with the 'locater' models which are popping up in enum/uri/NAPTR
space, making it explicit that you ask about a service, then find instances of
it as an explicit stage.

hmm. the return of the dreaded flow bits wars a-coming?...