Ciarán’s free software notes

A look at the good and bad of four proposed methods for fighting software patents.

1. Prior art
2. Patent pools
3. Our licences
4. Lobbying

1. Prior art

The plan: build up a catalogue of ideas that exist in free software projects so that when a free software developer is threatened with patent infringement, we can point out that the patent is invalid because we thought of the idea before the patent was filed.

Effectiveness: Low, potentially negative. If the database is publicly viewable, people applying for software patents could use it as a guideline for drafting valid patents. They could see what they have to avoid, and write their patent application so that it doesn’t overlap our prior art. If the database is not publicly viewable, the negative effect will be avoided, but the potential gain is low. Most patents are for ideas that we can’t prove we already thought of.

Richard Stallman recently explained this in an article about OSDL’s implementation of this idea. There was also a good article on CNet about Stallman’s essay.

2. Patent pools

The plan: Encourage holders of software patents to make an agreement not to sue free software users.

Effectiveness: Low. Three reasons:

1. The contributors may be mostly friendly companies – in which case the pool will contain patents that would not have been used against us anyway.
2. The contributors may have already cross-licensed those patents with the other large patent holders – in which case, the contents of the pool will not be available for us to use in counter-litigation.
3. "Patent trolls" – parasitic companies who write no software and make their money from buying patents and charging license fees from people who are writing software – will be wholly unaffected.

3. Our licences

The plan: Add terms to our licences saying that if you bring patent litigation suit against someone for using the program, you lose the right to distribute the program in future. Many software licences developed from the late-90s onward have attempted this. GPLv3 attempts this too, to some extent.

Effectiveness: A bit good. This action cannot solve the problem – it can only protect people from patent litigation in the case where the patent holder is also a distributor of the software whose licence contains such a clause. So this is a good idea, but it’s not enough on its own.

4. Lobbying

The plan: Try to influence legislation, inter-governmental treaties, global agreements, and patent office policy so that software patents are neither approved by patent offices or upheld by courts.

Effectiveness: Good, but hard work. This action can potentially solve the problem, and it has proved practical. See FSFE’s software patents page, FSFE’s WIPO page, and my August summary of software patents in the EU. No other action has the potential to completely save free software users and developers from patent litigation.