RE: Consensus on identifier/locator split?

["J. Noel Chiappa" <jnc@ginger.lcs.mit.edu>]

    > From: "Christian Huitema" <huitema@windows.microsoft.com>

    > I believe the question cannot be resolved in a general sense

I don't agree. If you decide it's absolutely necessary to provide a
capability which inherently separates location and identity (be it switching
providers while keeping "connections" open - even if it's sessions, not
classical TCP connections - or mobility, or whatever), they you *have*
answered the question "in a general sense".

Perhaps a more useful question is "do we want to provide one set of
underlying architectural capabilities which can be used to provide separation
of location and identity to a number of different "applications" - such as
mobility, multi-homing, etc.


    > specifically not if you ask for a "locator/identifier split", which by
    > its very wording is tied to the Internet Protocol layer.

Well, not really. I can easily imagine providing the "identification" part at
another layer - e.g. as part of some sort of session thingy.

However, if you want to keep the existing TCP (and I would think that's a
good idea, if you want to try and keep within the bounds of what's vaguely
feasible), then I suppose you have a point.

	Noel