Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.1 6/24/83; site sdchema.UUCP
Path: utzoo!watmath!clyde!bonnie!akgua!sdcsvax!sdchema!jwp
From: jwp@sdchema.UUCP (John Pierce)
Newsgroups: net.math,net.math.stat
Subject: New Linear Programming Alogrithm
Message-ID: <311@sdchema.UUCP>
Date: Fri, 7-Dec-84 04:12:55 EST
Article-I.D.: sdchema.311
Posted: Fri Dec  7 04:12:55 1984
Date-Received: Sat, 8-Dec-84 05:52:18 EST
Reply-To: jwp@sdchema.UUCP (John Pierce)
Organization: Chemistry Dept, UC San Diego
Lines: 12
Xref: watmath net.math:1634 net.math.stat:71

About a week ago I caught about the last minute or so of a Nat'l Public Radio
spot that said somebody (at Bell Labs?) has come up with a new alogrithm for
solving linear programming problems.  I'm reasonably certain I heard them say
it was better than an order of magnitude faster than simplex, and even better
than that for large problems.  Naturally, I expected to immediately see some
discussion in one of these groups (along with 6 different implementations in
net.sources).  However, so far...  nothing.

Does anyone have any solid information?  Maybe even a pointer to a journal
article?
				John Pierce, Chemistry, UC San Diego
				{decvax,sdcsvax}!sdchema!jwp