[1919] in Commercialization & Privatization of the Internet
combits
daemon@ATHENA.MIT.EDU (Eliot Lear)
Wed Jan 8 03:52:40 1992
Date: Wed, 8 Jan 92 00:50:20 -0800
From: lear@relay.sgi.com (Eliot Lear)
To: brian@lloyd.com
Cc: com-priv@uu.psi.com, lear@relay.sgi.com
>So it appears that things would be GREATLY simplified if ANS would
>simply subscribe to the CIX and let the CIX be the single point of
>access to the NSFnet for CO users (with the exception of CO users who
>subscribe to ANS directly, of course). Seems to me that this would
>completely eliminate the conflict.
Sorry, Brian. Your model gets complicated quick when you add in the
regionals, who support both research and commercial clients. Your
model doesn't improve their situation one bit, in fact. We would all
still have to have tables of who's who in order to implement the two
flavor policy routing that Yakov mentioned some time ago.
Assuming I'm satisfied with the current speed of the Internet and the
demand warrants the cost that speed, I would argue that the right
model is end user grants.
With respect to such grants, the argument has been made against end
user grants that there may be a lack of knowledge that keeps the users
from spending their money wisely. Perhaps one solution to this
problem is one or more openly bid contracts that purchase end user
connects, allowing the government to essentially order up a line
without the end user ever seeing the money. It's really a compromise
or transition position between where we are now, and where we should
in The End. One could even imagine a system where the user is given
the choice of taking the government provided service or going with
another.
--
Eliot Lear
[lear@sgi.com]
