[2629] in Commercialization & Privatization of the Internet
The AUP, Common Sense, and the Emperor's Clothes
daemon@ATHENA.MIT.EDU (Gordon Cook)
Mon Mar 23 21:19:20 1992
To: com-priv@psi.com
Date: 23 Mar 92 20:58:05 EST (Mon)
From: cook@tmn.com (Gordon Cook)
<<MESSAGE from>> Gordon Cook 23-MAR-92 20:58
cook@tmn
Although I haven't yet gotten to the point on my tape recording where the
AUP was discussed directly with Steve Wolff. The sense I got from the
hearings was certainly that the NSF should drop the AUP. As I said here
before Steve replied that the NSF believed that it was statutorily
required to have an AUP. He stated that if it were dropped the amount of
backbone traffic would go up dramatically and implied that the government
would be funding a free good for commercial networking where companies
that desired no communication with the research and education community
would be able to get on the internet and balance their checkbooks or send
spreadsheets back and forth all day long.
I think this reasoning is flawed -- as we have seen people say before on
this list many companies with a presence here WOULD NOT use this network
for proprietary enterprise networking for reasons of security and
stability. Those companies that hop on the network and use it would do so
presumably because they had REAL business to conduct with the R&E
community. I have transcribed almost 45 minutes worth of questions from
the hearing. I have heard nothing so far from what I have transcribed
that suggests that either congressman Boucher or Packard would be upset if
the AUP were dropped.
We hear that the NSF is required to have it, but to the best of my
knowledge no one has been able to cite, the statute, the executive order
or the OMB ruling or anything else on which this requirement is based. I
find this very perplexing. Steve has the general counsel at the NSF or
anyone else been able to find anything "scriptual" yet? If they cannot do
so, given the current environment, why not bow to the commercialization of
the network and drop the AUP.
Asking Congress to tell you to drop it is OK. But sheesh asking the
congress to give you NREN took over 4 years. How long would this take
before you got the statutory OK? Why not say we will drop the thing
September 1 unless someone can come up with an overwhelmingly strong
reason by then for our not doing so?
Dropping it would eliminate the tilt that has been given to ANS, because
ANS would no longer have exclusive commercial rights to the backbone. You
could then get on to building the NREN instead of having to fight the
ANS/CIX/MId-level wars. If ANS fells irreparably injured by this action
let it say so and let it pack its bags and go home. Others I am sure
would fill the gap. After all while the NSF fights the AUP wars, ESnet is
preparing to install a *REAL* bleeding edge backbone, it would be nice to
see you with an unhindered path to the same end.
Here's one of the best exchanges from the hearings of March 12th. Mich
cuts to the heart of things when he points out that the NSFnet may not be
able to serve as BOTH the production and the R&D network for NREN.
Congressman Packard: So you suggest that recompetition would be
successful if it were done on a level playing field. What would have to
be done to make it that way?
Mitch Kapor: I could suggest a couple of options. One option would be to
say that the network manager could not also play a commercial role. . . .
. Another possibility is to move to the other end of the spectrum and do
it in such a way that all commercial carriers would have equivalent access
to whatever backbone were put in place. I won't tell you that I know how
to do that today but I will tell you that it is worthwhile to discuss.
****What stands in the way of that today is really the acceptable use
policy and that is why modifying it or dropping it in order to give all
commercial providers equal access to these federally supported and
subsidized regional networks would be another way of leveling the playing
field.**** The distortion comes when you take one or for that matter two
carriers and partially subsidize them and you give them commercial rights
that nobody else has. . . .
Let me mention if I may a third option which is to separate out the idea
of a production backbone, that is the services which occupy 99% of the
users on a day to day basis. Separate this network from investment in
precompetitive ultra high speed broadband networks. My reading of NREN is
that it needs to serve both purposes. To expand the reach and to develop
a high end. . . . to expand the reach let there be subsidy to the
institutions to purchase services on the open market. . . . . At the same
time develop the high end . . . we support this but we don't want to see
it confused with the day to day production network that serves millions of
users today -- millions of whom are already in the commercial sphere.
Sixty percent the institutions on the US internet are commercial
entities. The rate of growth of commercial institutions far outstrips the
rate of growth of educational institutions.
