Stallman on Software Patents

An Anonymous Coward writes "On Monday Richard Stallman gave a speech at the Cambridge University (UK) Computer Lab. Over at ZDNet UK they have a transcription of the speech - the most eloquent discussion of the subject I have yet seen. Software patents victimise developers, he says, but there are ways to get around them. The best part is his comparison of writing software to writing symphonies: 'Oh Beethoven,' they would have said in 1800 if there had been patents on music, 'you're just bitching because you've got no ideas of your own.'"

Comments

Other Talk: Copyright vs Community

[thick_sliced]

Stallman also recently gave a talk at Queen Mary, University of London.
An audio version of the talk can be found at:

http://www.odl.qmul.ac.uk/stallman/

Allow an "independant discovery" defense

[redelm]

There is a solution to this morass. Simply allow the fact that an inventor discovered/invented a thing independantly of the patent as a defense against patent violation. Of course, there are evidentiary problems. But there always are.

The issue is that if something was independantly discovered, the patent should hardly be granted since it doesn't induce anything that wouldn't otherwise happen.

Re:Maybe something new?

[nagora]

Patents promote a developing market. It encourages you to get your ideas out in the open.

It seems to me that what they do is strike fear into the heart of all but the richest inventors, the legal fight to prove that you are not infringing a patent is more than most people can afford even when they are right. Patents encourage the small inventors to scrap the whole thing and just keep their day job, then everybody loses.

This is not an effect we need in the software industry where so much progress is made by individuals.

From what I gather the biggest flaw with patents are the clerks not the laws.

Software moves so fast that no prior-art system is ever going to be able to keep up. The only way to fix the clerk problem is to slow down the entire field to the point where they can cope. Sacking the idiot in charge of the USPO would help, too; paying clerks on the number of patents they approve is hardly professional or responsible.

TWW

Public key cryptography patents, not "PGP patent"

[Xilman]

I was present at this event. Despite what ZDNet claims, RMS did not use the term "PGP patent" but, correctly, described it as the public key cryptography patent.

The "transcript" is abbreviated which is acceptable, maybe even laudable, but it is also an inaccurate record of what RMS actually said --- which is not.

IMO, Stallman gave an extremely good talk aimed at those who were not experts on how the patent system works in practice. Far more eloquent than I'd anticipated from his previous writings on this and related subjects.

Paul

Having written a patent application

[f00zbll]

For a previous job, I had to research patents and write up a patent application. The instruction from the lawyer was "make it more general than it needs to be." His reasoning is the patent application shouldn't go through the first try. It should take atleast 2 or 3 tries, to make sure the patent is as broad as possible.

Now of course the lawyer gets paid each time you file, so it's just fine by him. The management tends to side with lawyers, but honestly, patenting ideas, especially in software is stupid. People confuse patent with copyright. Software should use copyright only, because you're protecting the actual work that went into building the application. It's both impossible and idiotic to patent ideas. Unfortunately lawyers run the country and now we have things like patented business process. Like the company that patented the use of Prozac for treating PMS for women. Ideas should not be patented, since it's not possible to police thought. Well that doesn't stop corporations and the government from trying.

Just because RMS can be a raving nut at times, that doesn't automatically discredit every word out of his mouth. Now if only there were more clear thinking people in government, we can fix this damn problem.

Music comparison is an interesting one

[jvmatthe]

Although I didn't see him mention this specifically, I think it's worth pointing out, given his discussion of music and the borrowing of material from other sources. In particular, how borrowing is important culturally.

One moment I recall quite clearly from my college years is the day in class when the teacher of the survey course in music asked the question "What makes American music American?" This was something I hadn't considered before, and the answer wasn't clear to me until he asked more precisely "Why is the 1812 Overture so distinctly Russian? Why is Appalachian Spring American?" The answer is that both borrow themes from the folk music of each country. The Russians knew that Tchaikovsky was writing Russian music because it had recognizable themes that reminded them that they were Russian. The Shaker melody in Appalachian Spring is something many Americans had heard many times before, when that piece was written. (And even moreso now, as "Spring" is itself a common piece of music to hear performed.)

So without the ability to borrow ideas and themes and work them into new compositions, music would be a barren landscape.

One could argue, similarly, that if we impose stronger and stronger restrictions on what we can build into software, then we are resisting the natural cultural synergy that results from algorithms and features being shared among the community members.

But four notes is enough to get sued

[yerricde]

Sure there are a small set of notes, and only so many ways you can arange any two notes in any tempo. After two notes, it is all in the arrangement, and composition.

The Yes! We have no bananas! case set the precedent that four notes is enough to get a songwriter sued in the United States. Given that there are only about 30,000 ways to combine four notes in the Western music theory (reply if you want a more detailed explanation of the math), it appears that the only reason songwriters haven't exhausted the melody space is that the big "all your right are belong to us" publishers have entered into cross-licensing agreements with one another. This is part of why you should write your legislators and request a repeal of the Sonny Bono Copyright Term Extension Act.

Too broad

[yerricde]

If you want to make nice, solid, constantly evolving software, go with Open-Source. Otherwise, if you're like the rest of the worl, you'll want to make money along with nice software (hopefully). Then, you'll go wtih Closed-Source proprietary, patented software.

The problem with patented software is that the patents that the USPTO has issued in the last 20 years are so d*ng broad that instead of "promot[ing] the progress of science and useful arts," they have precisely the opposite effect. For instance: data compression by dynamically building a character-to-string dictionary? Patent 4,558,302. Falling blocks puzzle game whose goal is to remove a specified initial set of colored or shaded blocks from the playfield (in other words, B-type Columns)? Patent 5,265,888. Image analysis by blocks against a smaller version of the same image? Patent 5,065,447. Heck, even topological sorting and XOR drawing were once patented in the U.S.

And don't count on waiting for the patents to expire. Just as Hollywood managed to get a Sonny Bono Copyright Term Extension Act passed with tons of soft money and (possibly mandatory) individual contributions, watch the pharmaceutical industry propose a Cherilyn LaPierre Patent Term Extension Act.