Below is another bit of postscript I have thrown together.
It's a map of a backbone that Peter L. and I have been
telling people is a good idea for ICMnet, especially in
light of the possible disconnection of Asia-Pacific,
European, Latin American, African, and Carribbean
R&E networks from each other as a result of the
NSFNET transition. In such a situation, ICMnet could
potentially act as glue between the various R&E networks
around the world.
It is also an attempt to deal with the differing routing
policies and acceptable use policies applied to ICM vs.
SprintLink and allow for different approaches to scale.
ICM and SprintLink would peer with each other in D.C.,
Pennsauken and Stockton. SprintLink would provide transit
between ICM and the other two NAPs and various peering
points; ICM would provide transit between SprintLink and the
FIXes.
This is not official at this time, and Vadim doesn't like it. 
(He has advanced a counterproposal which will have
SprintLink and ICM merge and share the same routing
policy and physical fabric.)
The principal differences betwen ICMnet now and this drawing
are the connectivity to (and the router at) the Sprint NAP
in Pennsauken NJ and the ICM line between Stockton CA and
D.C.
Sean.
- --
# This is a shell archive. Save it in a file, remove anything before
# this line, and then unpack it by entering "sh file". Note, it may
# create directories; files and directories will be owned by you and
# have default permissions.