Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!wuarchive!gem.mps.ohio-state.edu!uakari.primate.wisc.edu!xanth!rlb
From: rlb@cs.odu.edu (Robert L. Bailey)
Newsgroups: comp.binaries.ibm.pc.d
Subject: Re: rumor: LZW patented by Unisys
Message-ID: <10166@xanth.cs.odu.edu>
Date: 12 Oct 89 02:08:59 GMT
References: <1989Oct9.205917.13123@attdso.att.com>
Reply-To: rlb@cs.odu.edu (Robert L. Bailey)
Organization: Old Dominion University, Norfolk Va.
Lines: 24

In article <1989Oct9.205917.13123@attdso.att.com> tim@attdso.att.com (Tim J Ihde) writes:
>According to PC-Week's "Spencer Katt" article of October 2, the LZW
>data compression algorithm is patented by Unisys.  Has anybody heard
>anything about this?  Things could get messy if this is true . . .

The LZW compression algorithm was indeed developed by Terry A. Welch
of the Sperry Research Center, Sudbury, Massachusetts.  (Sperry &
Burroughs later merged to form UNISYS).  The algorithm was published
in the IEEE Computer journal (June 1984).  As I understand it, the
algorithm itself is not patented, but, Sperry developed several
proprietary implementations of the LZW algorithm for the Sperry
Univac 1100/60 (which is the machine that was used while designing
the algorithm).  These may be patented or at least copyrighted.
However, since the algorithm was published in an academic journal,
it should not preclude anyone from developing their own implementation
of the LZW algorithm (i.e. ARC, PKARC, Unix Compress, etc.).  
Generally speaking, information that is published in the academic
journals is made available to the public to benefit and advance the
cause of research; it usually carries no restrictions on its usage.

Bob Bailey
Unisys Defense Systems
Chesapeake, VA
rlb@cs.odu.edu