[139] in Commercialization & Privatization of the Internet
The Internet and the regionals - use policy implications
daemon@ATHENA.MIT.EDU (Erik E. Fair" (Your Friendly Postm)
Mon Nov 12 23:13:55 1990
From: "Erik E. Fair" (Your Friendly Postmaster) <fair@apple.com>
To: com-priv@psi.com
Date: Mon, 12 Nov 90 20:03:29 -0800
One thing I haven't seen discussed in here is the degree to which the
regionals are subsidized by their commercial members. There is now a
pressure that might make that subsidy go away.
Consider the following: Apple Computer and Sun Microsystems decide to
collaborate on some secret project, and wish to exchange data with IP
(OK, it's unlikely for these two companies; just humor me for a minute).
Right now, the regional network to which we are both attached
(BARRNET) has the same use policy as the NSFNET backbone. So that
means that this collaboration cannot use BARRNET. This leaves two
alternatives:
1. Private IP link (we're close enough - it's probably pretty cheap,
but suppose we want to collaborate with someone on the east coast?)
2. We both get a connection to one of the unrestricted commercial IP
network service providers. Obviously, there are security issues here
that I am ignoring for the purpose of illustration.
If we go with #2, things are more flexible in the long term, since you
never know who you're going to be collaborating with next. And you can
still speak IP under the restrictive use rules with the NSFNET people,
and you get backup IP service in the bargain, if you set routing up
right.
Now, sometime later, the bean counters come down to look at the books,
see redundancy, and say "Why do we need two networks, when we get
all the connectivity we need from one? Nuke one of them."
Which connection gets cut, do you suppose? BARRNET is awfully cheap,
but we can't do that whizzy commercial collaboration with Sun over it.
Erik E. Fair apple!fair fair@apple.com
