no-op?

[Martha Steenstrup <[e-mail]>]

Hello John,

Nimrod, when completed, will be a scalable architecture for routing
and addressing in a large and diverse internetwork.  If we had to
define a network-layer packet header for Nimrod we would, but there is
so much good work going into packet header design currently that it
would be superfluous for Nimrod to offer yet another packet header
design.  We are instead concentrating our efforts on mechanisms for
expressing and distributing routing and addressing information in a
hierarchical internetwork and for selecting and using routes that meet
user service requirements.  In general, we are working on mechanisms
for reducing the amount of routing-related information in large
internetworks and hence the amount of resources necessary to carry out
routing-related activities.

We Nimrod folks will work with any internetwork packet format
presented to us, provided that the information in the packet header
permits packet forwarding in a very large internetwork.  Thus, Nimrod
could work with IPv4, CLNP, SIP, PIP, SIPP, whatever; in some cases,
Nimrod may attribute different semantics than those originally
intended for certain header fields.  For example, Nimrod may interpret
IPv4 "addresses" as endpoint identifiers; Nimrod addresses would be
distinct from IP "addresses".

m