The CIX and the NSFNET regionals - a dilemma

It is clear to me that one of two things is going to happen:
   The rules will change, and there will be good technical solutions. -or-
   The rules will not change, and we will have split research and commercial
   networks with weak interconnects. Organizations that really need to function
   in both worlds will have two network numbers or two connections....

Good assesment. One thing that I have noticed in life, people don't
always see their way clear to perform preventative measures (oh, money
is too tight, I don't have time, etc.) but they sure scream when the
damned thing breaks. Perhaps we ought to start encouraging companies
to start requesting more than one network number: one for RE and one
for non-RE (I would rather we quit calling the non-RE thing "CO" since
not all non-RE stuff is commercial). This will cause the usage of
network numbers to increase dramatically, hastening the time that we
run out of address space. This would break things very quickly.
Maybe the hue and cry would convince the powers-that-be that things
ought to change. It would really be a shame to be forced to break the
very thing many have spent significant parts of their life to build
just so we could get it fixed.

Brian Lloyd, WB6RQN Lloyd & Associates
Principal and Network Architect 3420 Sudbury Road Cameron Park, CA 95682
voice (916) 676-1147 -or- (415) 725-1392