🚀 go-pugleaf

We're developing some software here at Stanford which utilizes interior-
point methods for optimization. Karmarkar's famous LP algorithm is an
example of such methods, but it is not the one we use.

...however, I have heard that AT&T has a patent whose scope extends
over all interior-point methods; on the other hand, I know at least
some countries who specifically disallow patents on 'algorithms', such
as France. So I may not be correct on that. I do know that, after the
initial fervor over Karmarkar's algorithm in the press, AT&T quickly
squelched when the realized what a money-making tool was being given away
to/by the academic community!

Can anyone shed some light on this issue? It certainly would affect the
distribution of our work if AT&T breathes down our necks. It certainly
has enough new, original work that maybe we could escape from under the
patent, but a bunch of grad students/professors can't afford to go up
against a giant.

(Yes, I'm going today to the library to check out the patent).

(I posted to gnu.misc.discuss because of the League for Programming Freedom
announcements that occaisonally show up---so I thought this might coincide).

Thanks a lot,
Michael C. Grant