[1103] in Commercialization & Privatization of the Internet
Re: some dumb questions from the gallery
daemon@ATHENA.MIT.EDU (Martin Lee Schoffstall)
Thu Aug 1 17:01:03 1991
To: peter@goshawk.lanl.gov (Peter S. Ford)
Cc: sob@tmc.edu, com-priv@psi.com, mcostel@kaman.com, nren-discuss@psi.com
In-Reply-To: Your message of "Mon, 15 Jul 91 14:04:08 MDT."
Date: Thu, 01 Aug 91 16:57:13 -0400
From: "Martin Lee Schoffstall" <schoff@psi.com>
Peter,
Sorry for the long delayed response - vacation and all of that....
Marty,
Are you at least willing to agree that we might need gigabits in support
of "operating MegaInstitutions"? To quote the little blue book:
Sure. When we get to kilo-institutions we need need T1, when we get
to 10's of kiloinstutitions we probably need T3 in the retail networks.
Now I need to use a word which has negative connotation to some - aggregation.
How have the phone companies gotten to operational gigabits in the few areas
that they have deployed them - by aggregating all of of those sub ds0
analog phone calls from the mass market and then throwing in the occassional
T1 data pipe.
They've created a long term stable infrastructure/service/market by making
it ubiqutious and agreggating the small users of the world.
Let's emulate this model.
I believe that the NREN as it is being discussed in the political circles
has several concurrent components including, but not limited to,:
1) improve capabilities of existing portion of federally sponsored interne
(both bandwidth, reliability, ease of use, etc.)
Well we certainly need improvements in reliability.
2) help broaden deployment of public data networking
Well the public has 2400-9600 baud modems.
3) invest in gigabit network research since it is believed we
will need it for aggregation AND for high speed applications.
How about useful applications?
Am I on track with this discussion, or am I really missing something? If
we are on track, then perhaps we can start discussing how best we can accompli
all of this. I have a hard time imagining that this all is best accomplished
putting ALL NREN monies into the end users hands. This merely pushes the pro
to the identification of THE end-users. I suspect this would get so broad so
we could argue that the 70-90 million bucks gets diluted to less than $100 per
Not much impact can be had from this from my point of view, and I think most
govt people (congress and executive branch) would agree.
I don't think I've ever argued that all of the money should be put in end
user's hands, I'd like to see some useful AND speculative research done.
But I would like to argue percentages of where the money goes! I'd also
like to have a real shot at discussion which parts of the old model are
now obsolete, what things can be seperated out, what pieces should be
owned by the community, vs being impaled from the top, etc..
I believe this discussion should start focusing on identifying viable targets
NREN funding which can subsequently gain leverage from the marketplace. My
favorite hit list includes, and is not limited to,:
access to information sources: libraries, databases, etc.
name service / security services for authentication
user services, consulting on application of network technologies
operational standards, the network is a "critical resource" to most tha
use it.
Operational Standards, what a concept. Your shaking the foundations of
blue ribbon panels here.
making it possible for end institutions (US govt, companies, etc.)
to enforce their own appropriate use restrictions, not the
networks.
Most commercial organizations already do this, I've seen a lot of heat
and smoke on academic institutions trying to do this.
How to handle success: gigabits, resource reservation, internetworking
instead of watching broadcast TV, etc.
Some of this requires research, some needs seed deployment, some will
just happen. I am interested in what vehicles need to be funded, in what fas
by how much.
Agreed.
Marty
