Freedom of Speech in Software

akpoff writes " I've been struggling with the question 'what's wrong with software patents' but haven't been able to find the right words. I was over at John Gilmore's website and found a link to John Salin's 'Freedom of Speech in Software' letter to the USPTO back in 1991! This is one of the best explanations I've seen. He reminds us that computer programs are essentially like literature or music -- they are expressions of ideas. Just because they run on a computer doesn't make them uniquely different from other creative mediums. We should think player piano (patentable) vs the music (copyrightable but not patentable) it plays. Europeans -- put this letter into the hands of your MEPs!"

How to bring across your message to MEPs

[Sven Tuerpe]

Europeans -- put this letter into the hands of your MEPs!

This article [theregister.co.uk] suggests that free speech might not be the prime issue from a MEP's point of view.

Re:But if code is like music...

[ralphus]

It isn't about encouraging code to be blatant rip-off's of existing ideas. Did you RTFA [philsalin.com]? I link [philsalin.com] it [philsalin.com] again [philsalin.com] for emphasis.

It's about me or you or anyone having the natural right to be able to stand on the shoulders of giants. Imagine if you didn't have the benefit of standard libraries because all the concepts and processes in them had been patented.

The ariticle [philsalin.com] lays out all sorts of other seemingly reasonable analogies that I'm too tired to type at 4 am.

FYI, I think this post [slashdot.org] does quite a good job of laying out the systemic issues.

That argument won't work!

[Serious Simon]

The "free speech" card is useless in case of the proposed European Software Patents directive.

Article 5 of the proposal says:

Member states shall ensure that a computer-implemented invention may be claimed as a product, that is a programmed computer, a programmed computer network or other programmed apparatus, or as a process carried out by such a computer, computer network or apparatus through the execution of software

and this is explained on page 15:

... It should be noted that the proposal has not followed the practice of the EPO in permitting claims to computer program products either on its own or on a carrier, as this could be seen as allowing patents for computer programs 'as such'.

And on the bottom of page 7, it says

An abstract algorithm can be defined in terms of pure logic in the absence of any physical reference points. It is possible that such an algorithm may be put to practical use in many different functions in apparently unrelated domains, and may be capable of achieving different effects. Thus, an algorithm which is considered as a theoretical entity in isolation from the context of a physical environment, and in respect of which it is accordingly not possible to infer its effects, will be inherently non-technical and thus not susceptible of being regarded as a patentable invention.

It is a consequence of the above that an abstract algorithm as such cannot be monopolised. The normal rules for patentability mean that a patent claim to an invention which is founded on a particular algorithm would not extend to other applications of that algorithm.

The way I interpret this is that "free speech" objections to the proposal are effectively countered. The proposal denies patents on algorithms and on software 'as such'.

In other words, your rights to write and publish software are not affected (free speech), but you are not allowed to run any software that allegedly contains patented technology, without paying for a license!

I think the only useful (and powerful) objections to the directive are economic ones. Patents as allowed by this directive stifle innovation rather than promoting it, and can easily be abused for anti-competitive purposes. The directive allows over-broad patents that pose a risk to the software industry (although the "explanatory memorandum" sounds very reasonable, the actual articles of the directive provide hardly any limitation to the scope of software patents or guarantees that they are not too easily granted).

For example, the broader version of the "Amazon one-click patent" that was recently granted by the EPO, would be allowed by this directive.

In the long run, the negative effect on innovation would not even benefit the big software companies (who initially may profit from software patents as anti-competitive tools). It will only be profitable to a small group of patent lawyers (at the EPO) and a number of patent sharks, at the cost of the European citizen.

Read the proposed directive for yourself and shudder:

http://europa.eu.int/comm/internal_market/en/ind prop/com02-92en.pdf

5 steps to solve the problems with patents

[jonwil]

1.Applicants must show a working example of their patent that implements all the claims contained in the patent. If they dont, they dont get the patent. (this would probobly get rid of some of the shonky patents)

2.The patent office should do better searches for prior art and if things are found that should have been found by the applicant, the applicant must pay the patent office money (to discorage the filing of "obvious" patents and so on)

3.If you dont enforce your patent, you risk loosing it (i.e. it would be similar to what happens with trademarks now). This would actually be good since it would prevent things like what happened in the Unisys LZW case.

4.Once a patent has been submitted, it cant be changed. Also, the patent must be published publicly as soon as it is submitted. (as a Pending Patent or something) Also, the date a patent expires goes from when the patent was submitted not when it was granted. (this would prevent e.g. patents that are submitted then modified then modified again then delayed and so on in an attempt to make the patent last longer)

and 5.You cant patent a formula or algortihim (genetic, biological, chemical or mathematical).
This would mean that patenting individual genes or whole organisims would be illegal. As would patenting software algortihims like encryption, compression and so on. An example here might be a test for breast cancer that looks for a breast cancer gene. You can claim a patent over your specific way of finding that one gene from all the others (assuming that there is no prior art of course) but you cant patent the gene itself or patent "testing for breast canser by finding this gene".