Routing per se only uses locators and forwarding indications, and
doesn't care what properties any identifiers might have (it doesn't use
them). From the point of view of routing and addressing, the motivation
for locator/identifier separation is to REMOVE constraints on routing
architecture, to make it possible to solve the routing architecture
problems (scaling, robustness, adaptability etc.).  The list of RRG
requirements with respect to identifiers is negative, not positive.  For
example, something like: "A locator MUST NOT be constrained to a
long-term association with an object.  Locators name network attachment
points, not endpoint interfaces.".

However, for any routing/addressing/forwarding architecture, RRG should
write up the implications has on identification functions.  In order to
do that, we do need to understand what identification functions are
essential for Internet operations.  We have to be sure we understand the
implications of what we recommend before we can do so responsibly.

I believe we are at the point where we have a good sense of what
identification functions are needed, and now we can keep that list in
our back pocket, as a checklist, and narrow the scope of activity back
down.

Scott



_______________________________________________
rrg mailing list
[email protected]
http://www.irtf.org/mailman/listinfo/rrg

Reply via email to