Path: utzoo!utgpu!jarvis.csri.toronto.edu!rutgers!cs.utexas.edu!swrinde!zaphod.mps.ohio-state.edu!sunybcs!sbcs!stealth!brnstnd
From: brnstnd@stealth.acf.nyu.edu (Dan Bernstein)
Newsgroups: comp.misc
Subject: Re: Lempel-Ziv Implementation
Message-ID: <4161@sbcs.sunysb.edu>
Date: 5 Dec 89 06:26:20 GMT
References: <10690@venera.isi.edu> <1989Nov30.144334.7215@alzabo.uucp> <56000@looking.on.ca> <45686@sgi.sgi.com>
Sender: news@sbcs.sunysb.edu
Reply-To: brnstnd@stealth.acf.nyu.edu (Dan Bernstein)
Distribution: usa
Organization: IR
Lines: 19

In article <45686@sgi.sgi.com> bron@bronze.wpd.sgi.com (Bron Campbell Nelson) writes:
> Ummm ... it has always been my understanding that an *algorithm* is not
> patentable.

> Could some lawyer type clarify this?

I'm not a lawyer, but anyway:

You are correct that a ``method'' may not be patented; the law says so
explicitly. Nevertheless, Brad is correct that Unisys holds a patent
covering LZW. Apparently, IBM holds an intersecting patent; in any case,
until either is tested in court, the patents are valid.

On the other hand, Brad is incorrect in his claim that there is ``no
such thing'' as a public-domain LZW implementation. There are a few
non-copyrighted LZW implementations running around; and until a court
enforces the patents, the situation won't change.

---Dan