[1718] in Commercialization & Privatization of the Internet
Re: nsfnet as large-scale testbed, pt. 2
daemon@ATHENA.MIT.EDU (Craig Partridge)
Fri Dec 13 12:34:09 1991
To: sgou@seq1.loc.gov
Cc: com-priv@psi.com
Cc: kwe2@bbn.com
Date: Fri, 13 Dec 91 09:29:44 -0800
From: Craig Partridge <craig@aland.bbn.com>
> The NSFnet is *not* supposed to be a testbed, it is a vehicle for
> research and education communication. There are other networks that do
> network research and testing.
I just wanted to follow up Kent's note with a few details. People should
be aware that the Feds fund entirely or in part a number of networks
devoted entirely to experiments. Each network connects a number of
research sites. Examples include the five gigabit testbeds, the
DARTNET testbed, and testbeds for experiments in multimedia conferencing
and (I believe) distributed simulation.
In this context, I believe Kent is right that the NSFNET backbone is
not intended as a vehicle for serious experimentation in networking
technology. Personally, I think that given that NSF is subsidizing the
network, NSF is still entitled to encourage the backbone provider to deploy
the next generation of technology a bit earlier than the provider might
otherwise do itself (in the interests of prodding the world into improvement).
We can debate (should one wish) whether the T-3 backbone is an instance
of such a policy, and whether it has proved a good idea.
Craig
