Subject: GIF/LZW patent
From: "Lawrence Rosen" <>
Date: Tue, 26 Sep 2006 19:57:36 -0700

> > [1]  This is what Ghostscript did for LZW writing, eg, for rendering
> > Postscript to GIF.  Some BSD developer went to the trouble of doing
> > the same for libreadline, since it's not free enough by their
> > standards.  In both cases the API was already developed, and I
> > wouldn't be surprised if that wasn't the case here too.
> That's _very_ different: they knew they wanted to use Readline
> functionality, and that they could not use the existing code because
> of licensing.  They were free to re-implement it, and chose to.  They
> started from a position of knowing what they had to do, and an
> economic choice: use the existing licensed code, or do the work of
> re-implementing it.
> Therefore, because of the risk of legal consequences, there is a
> burden of having to be aware of what methods to avoid accidentally
> re-creating.
> This happened, for example, with the GIF patent.  Everyone writing
> code to create GIF images had to be aware that one particular method
> for encoding them was to be avoided (unless they were in a position to
> get a commercial license), and another particular method was fine (or
> commonly presumed to be).  Both methods were obvious given the public
> domain specification.  That was a burden on programmers, though it did
> have the convenience of being widely publicised so everyone knew about
> it.

Our firm happened to represent the Ghostscript author at that time. Without
disclosing any secrets, I can assure everyone that the solution we
recommended was based on an analysis of the patent claims and the affected
software. It was burdensome, but not horrendously so, and it led to an
inventive solution to the LZW patent problem. 

Designing around a patent whose license terms you dislike is an excellent
incentive for innovation. There's nothing wrong with that.

I've been on both sides of software patent wars. Take it from a bruised
warrior, none of these battles has ever been as contentious or as difficult
for free software as derivative work analysis surrounding the GPL. Patents
are just descriptions of technology, after all, and it's usually not rocket
science. Some of you are grossly exaggerating the difficulty of
understanding, and if necessary designing around or possibly invalidating,
software patents in the areas where you are writing advanced software.

I happen to believe that patents can work FOR open source in much the same
way that copyright was supposed to. Even better, if we do it right. Just the
opposite result if we continue to close our eyes to the reality of software

I'll be blunt about it: If we didn't have copyright on software and only had
patents, we'd be much better off. At least then monopolies would be based
upon innovation instead of mere writings by armies of programmers of mostly
me-too code. Programmers would be rewarded for being innovative, and the
rest of us could otherwise copy and reuse any non-patented or off-patented
software we wanted, regardless of who wrote it. That's how intellectual
property works in every other area of technology! Wouldn't that be a better