[526] in Commercialization & Privatization of the Internet
Re: ANS Acceptable Use Policy
daemon@ATHENA.MIT.EDU (Richard Mandelbaum)
Fri Apr 5 03:49:10 1991
To: russell@cmcl2.nyu.edu (Bill Russell)
Cc: com-priv@psi.com
In-Reply-To: Your message of Wed, 03 Apr 91 16:27:50 -0500.
Date: Thu, 04 Apr 91 09:30:59 -0500
From: Richard Mandelbaum <rma@tsar.cc.rochester.edu>
In any given case the most restrictive AUP applies.
The insane case Bill alludes to are those where I go from NYU to Widget.com
over a NYSER-PSI-CERFNet route and when my correspondent types r, his
reply gets routed over NSFNet. Thus my rules are different from his.
This is not too bad fro a University, since one can make the argument that
anything a University does is at least indirectly in support of research and education. (Including Presidential yachts and furniture, though representative
Dingell might disagree on that). The situation is more murky if you have
pure .com to .com traffic.
A possible solution is to follow the DASC precedent of allowing supercomputer
centers to sell up to 10% of their time purely commercially. If say any
network (non-profit or commercial) which attached to the NSFNet backbone
had to pay some amount, to cover the cost of the purely commercial traffic
it sent over the backbone theneveryone could get out of AUP's altogether.
Initially the amount could be a rough estimate made by the network and
refined through precedent.
Ultimately this is a purely economic question and any solution which would
convince an auditor that Tax dollars are not subsidizing things they aren't
supposed to be subsidizing is probably a good start.
How about all of us coming up with some suggestions to allow the
government subsidized nets to eliminate AUP's
rather than flaming on about how terrible NSF is and how
this would be a much better world if NSF had never gotten involved and there
was no NSFNet and we had simply allowed ARPANet to evolve into the great
commercial Net. For those who are interested in seeing what happens in a
purely commercial evolution of telecommunications , a good source is the
book, The Telecommunications Industry by Gerald Brock. harvard university
press. 1981 (pre-divestiture). he gives a good history of the beginings and
spread of the telephone and telegraph industry in the US and to some extent
in europe.
I for one, believe the current evolution has been more orderly and more
beneficial to the end user community than waiting for these services to
evolve in the marketplace.
(Note that essentially no US infrastructure, not
mail, not rairoad transportation, not telegraph or telephone or radio or
television or automobile travel or plane travel, has evolved without
massive government intervention in the marketplace. The US government paid
the railroads up to $48000 per mile of track laid in the nineteenth
century. Also the governement granting of monopoly status to ATT, was done
at ATT's request. ATT was , in 1911, regulated because it requested it,not
because it was forced to succomb to regulation)
This marketplace is still in its infancy and I see many suggestions which
amount to throwing out the baby with the bathwater. We have a lot of
evolving still to do ,CIX being an excellent example of an evolutionary
step towards universal interconnectivity. I think we need to understand
that the first regionals started operating in 1987 and the commercials
din't come on the scene until 2 or 3 years later. We should analyze things
very carefully before embarking on radical solutions .
It is an intersting paradox of polical history, at least in New York State,
that the "reform" politicians who campaigned on the cry of "throw the
crooks out" generally wound up ensuring that the public had much less
services at a much higher cost than the
"wicked" regimes that they replaced. (The Lindsay regime in New York City
being a classic example). I'd hate to see that paradym applied to the
National Information Infrastructure many of us are trying to build.
____________________
