[1139] in Commercialization & Privatization of the Internet
FOLLOWUP Point, Was -> Re: Forwarded from PACS-L Digest
daemon@ATHENA.MIT.EDU (Martin Lee Schoffstall)
Fri Aug 9 14:41:24 1991
To: "Allan H. Weis" <weis@nis.ans.net>
Cc: Stephen Wolff <steve@ncri.cise.nsf.gov>,
In-Reply-To: Your message of "Fri, 09 Aug 91 14:09:33 EDT."
Date: Fri, 09 Aug 91 14:37:09 -0400
From: "Martin Lee Schoffstall" <schoff@psi.com>
I think broad connectivity is necessary.
Speaking for PSI, we agree that broad connectivity is not only necessary
but mandatory. For PSI that means (1)rich layer 3 interconnections, engineered,
firewalled and operated appropriately,(2) multiple layer two carriers -
[again from the meeting the Sprint representative thought we only used
one IXC, where in actuallity we use 3]. It also means interconnection
to other application service providers and packet providers - for instance
we've been running X.25 connections for a long time for both rfc877 ip packet
access as well as X.500 Application layer interconnection, and pad/telnet
application gatewaying.
Most of these concepts should apply to the broad connectivity of the
macroscopic NREN and/or the Internet. For instance clearly the Internet
is a richer place with application gatewaying into MCIMail, and also
clearly as Steve pointed out MCIMail/Compuserve should not be using
the NSFNet as the transport net.
So from my first message do we plug the "Electronic Mail Exchange, EMX"
into the FIX, CIX, or NEX?
Or is there something else that you, the feds, or the community propose?
The time to explore the alternatives is now.
Marty
