[2115] in Commercialization & Privatization of the Internet
Trickle Down? How About Bubble Up?
daemon@ATHENA.MIT.EDU (Brian Lloyd)
Mon Jan 27 14:21:46 1992
Date: Mon, 27 Jan 92 11:17:01 PST
From: brian@lloyd.com (Brian Lloyd)
To: jhaverty@us.oracle.com
Cc: daveh@csn.org, com-priv@psi.com
In-Reply-To: Jack Haverty's message of Mon, 27 Jan 92 09:17:48 PST <9201271717.AA14102@rivendell.us.oracle.com>
Reply-To: brian@lloyd.com
Date: Mon, 27 Jan 92 09:17:48 PST
From: Jack Haverty <jhaverty@us.oracle.com>
Yes, but why should I? What advantage is there to running uucp over
TCP versus running uucp over raw dial-up with V.32/42 modems? Which
approach costs less in terms of equipment and operations expense?
Does anyone have any hard data about these alternatives (Scott Bradner
- a new frontier?!)?
Jack
Now you get into one of the major reasons that I designed the
NetBlazer: so that we could use it to learn the answers. It is not
clear that dial-up SLIP or PPP with SMTP and NNTP is more
effecient/effective than uucp for mail and news delivery. My gut feel
is that we will reap benefits from not having to have protocol
conversion gateways (ala your local UNIX system doing conversion from
SMTP and NNTP to uucp). I do know that being able to reach out and
FTP something when I want/need it is a big win over what I had when I
lived with uucp. But that is anecdotal and does not form a
statistically valid universe. So we use the scientific method:
hypothesize that dial-up IP is better, build the tools, try it out and
collect the data, then make a final determination.
Brian Lloyd, WB6RQN Lloyd & Associates
Principal and Network Architect 3420 Sudbury Road
brian@lloyd.com Cameron Park, CA 95682
voice (916) 676-1147 -or- (415) 725-1392
