Summarising the problem
I was recently asked to give an introduction to our perspective on software patents. Below is a quickly edited copy of what I sent. The purpose was to help someone prepare for a meeting, so these are starting points, not an overall summary.
Why shouldn’t they exist?
Software patents block useful freedoms which computer users can make use of individually or in groups. Economically, they confer an unjust amount of power to large companies who can combine patents with a strong market position to exclude others from large areas of software development.
On freedom: Every software patent is a restriction on what you can do with your computer. Patents on car manufacturing are a restriction on what you can do with your car manufacturing plant, but since you don’t own one, it’s a meaningless restriction in terms of personal freedom – it’s just an industrial regulation, one among many. Software is different because writing software is as accessible as writing music or writing a book, and because these fields don’t have existing financial and legal restrictions, introducing the restrictions of patents is unacceptable. You (or any particular person) may not plan to write software, and most people will never write a book or any substantial amount of music either, but everyone can understand that society benefits from the general freedom of others to do these things, and everyone understands the value of having the freedom to do these things one day if they really want to.
Most of the GNU/Linux operating system was developed by individuals or groups of people who aren’t directly paid for the software development they did, but when it’s all put together the result is of massive social benefit.
On economic problems: If someone holds a patent on a common data format for email, images, or video, they can block me from using that data format. Since an email reader would be useless if it can’t read the date format that emails arrive in, that patent equates to a document with the power of law banning me from writing a functional email reader (or image viewer or video player).
For large companies with cash, lawyers, and their own portfolio of patents, these problems can be circumvented by cross-licensing, or if necessary, by paying for use of the patent. For small companies, the legal and financial requirements are too high, and their small patent portfolio is probably useless. Cross-licensing can be done if the small company has a patent which is infringed by the large company, but with one or ten patents, this is unlikely, and the large company can easily check this. In the other direction, when a company like Microsoft or IBM, each of which holds more than 10,000 patents, most of which are software patents – when they tell a small company that the small company is an infringer, there’s no way to check all those patents, and there probably is something in that mountain of patents which really does cover some technique used by the small company.
Some organisations say that the problem is the number or the quality of patents being granted, but this is a mistake. If Microsoft and IBM had 5,000 instead of 10,000, the problem wouldn’t diminish – and no one has proposed any method that would cut patent granting by 50%. Finding ways to judge “obviousness” and “usefulness” is not easy. A subject matter exclusion is much more practical.
Another example is the MPEG video format. 22 companies claim to have essential patents on this format. Negociating with 22 companies, and paying each of them would be frustrating and costly, but reducing this to 10, or even down to 1, doesn’t change the fact that someone has a 20-year veto on whether or not my company is allowed to sell a video player. In software, where interoperability is so crucial, these powers go too far.
When patent offices and government bodies hold “public consultations”, it’s always the lawyers groups and the large multinationals who hear about them and have the procedural knowledge necessary to participate. Recent examples are Europe, the USA, and Australia. For some industries, that might be ok. If all product development is done by those multinationals, then that’s ok – maybe the car manufacturing industry or the pharmaceutical industry is like that. I don’t know about those industries, but for software, the cost of software patents is borne by individual computer users, companies that use computers but which wouldn’t call themselves software companies, and many small to medium sized businesses. There are also large multinationals involved, but they produce only a minority of used software, and they have mechanisms such as cross-licensing to ease the problem. So, for software, these public consultations fail to reach the people whose opinion should count. We have to try to help those people participate.
When patent offices launch procedures to consider whether or not to continue granting software patents, we participate, but in terms of democratic legitimacy, substantive decisions that affect society shouldn’t be made by the patent offices.
When patents *only* affect areas of industrial activity, then the question of harmonisation is an economic/political one and could be the subject of an international treaty.
Software patents are different in that they affect what you and I can do with our computers. They affect individuals, and they affect communities such as the communities that have developed the GNU/Linux operating system and other widely used free software.
An international treaty banning software patents would be legitimate. It could say that patent law cannot be used to prevent people from writing software or from making a computer do input-output and computations – that would be protecting a freedom, like press freedom. But an international treaty requiring software patents would be illegitimate since it would be imposing restrictions on people’s daily lives and their ability to work with each other as a community. Sometime restrictions on people’s daily lives are necessary, but this should be policy set by the government which is answerable to he people concerned. It shouldn’t be set at a business negociation between large corporations and diplomats.