Why patents does not work

December 11, 2009 · Filed under open society, open source, politics · Tagged open source, patents, ruby on rails, software patents

This is my notes from the talk I did at the 2009 Plone Conference in Budapest, reworked into a series of blog posts. This is part 2 of 3. Read Part 1 and Part 3 as well.

Ruby on Rails and patents

Open source is a “Tower of Babel”, an example of people collaborating on building something big and complex. But the liberty to collaborate is not entirely unchallenged. There is plenty of lobbying for software patents, and software patents is a brake on the collaboration that is open software. As a first example, I’m going to take the web framework Ruby on Rails, which came in 2004. It created a lot of interest and inspired many other frameworks, and helped spread many good web technologies, like web object publishing, model-view-controller type frameworks and object-relational mappers. All of these technologies have the last couple of years become standard for web frameworks, and almost all new web frameworks use them.

But Ruby on Rails didn’t invent any of these technologies. The first system to use web object publishing was Zope, created in 1996 by Jim Fulton. The model-view-controller principle was created by Trygve Renskaug in 1979. The first implementation of those principles for web frameworks was done 1999 in a framework called Struts. I don’t know who first made an object-relational mapper, but they existed well before Ruby on Rails.

If any of these technologies had been patented Ruby on Rails wouldn’t have been as good and therefore not as successful. It wouldn’t have helped spread these technologies, and the new web frameworks created today would also be worse then they are. Software patents acts as a hindrance on development in the software industry. The more software patents we have, the worse our software will be.

And almost anything is patentable in software. Microsoft has a patent on using the “is” operator in Basic. That means that no other maker of Basic can have an “is” operator. Microsoft didn’t invent the “is” operator nor Basic. Does it make sense to anybody that they have this patent? It’s a bit like somebody patents the idea of watching TV in bed without having invented either TV’s or beds. If that’s patentable, then almost any idea in software is patentable.

Our only hope is that software patents will be ineffective so we can ignore them, but that’s very unlikely. In 1988 a company called Brooktrout patented the sending of faxes from computerized voice telephony systems, without having invented either the sending of faxes from computers, or computerised voice systems. It’s another case of patenting watching TV from bed. But although Microsofts patent is perfectly valid as they were probably the first to have an “is” operator in Basic, Brooktrout were not the first to send faxes from voice systems. There was what is called “prior art”. So you might think that the patent was invalid, and nobody needed to pay Brooktrout, but that’s not the case. There was some lawsuits and counter suits, but they settled out of court, presumably for less then what a longer court case would have cost. So it’s perfectly possible to patent things you didn’t even invent.

Other patents

But what is it that makes software patents different from other patents? Well…. I don’t know. I don’t think there is a difference. Watt had patents for his condenser, practically giving him a monopoly for some 25 years.
And an indication that this was bad was that steam engine development really took off in 1800, when his patent expired. It’s not unlikely that Watts patents delayed the invention of things like steam trains with several years.

Watt himself was for years hindered to use his steam engine to generate circular motion, because there was a patent on cranks. Yes, cranks. Of course, cranks had been known for hundreds of years before James Pickard patented it, but he still got a patent. Makes sense? No. And Watts wasn’t alone in inventing the condenser either. Humphrey Gainsborough also came up with the same idea, at around the same same time, but he didn’t patent it, so he didn’t get rich from it. Fair?

But isn’t patents needed? Isn’t it so that without patents no research and development would be done? Well, why not? We make money in the open source world by selling our knowledge while still sharing it freely. Why wouldn’t that work for other industries? Sure, the argument is that research and development is expensive. Patents are needed so the the research cost can be recuperated. But developing software also costs money, and Ohloh.net estimates the cost of developing the Linux kernel from scratch today would be $136,593,689. That is a lot of money. And research suggests that in fact, patents don’t increase innovation, and may even work as a hindrance.

As we see with the steam engine, most inventions even in mechanics are small and incremental. Should we really be allowed to block development like that just because we made an incremental improvement, or came up with a new use of something? In medical research there is never one single company that has done all the research needed to develop a medicine. Research on how the disease works are done over a long time by many companies and universities. Yet one company get the patent.

Another problem with patents it means nobody is researching things that are unpatentable or hard to sell. Medicine for malaria is one example, diets and losing weight is another. You can’t take out a patent on what food to eat, or make money from it. With an open source type medicine, your selling point is no longer a patented drug, but your knowledge. And that makes all knowledge profitable. Open source medicine might actually increase the development of medicines, and at the very least give access to medicines to more people. And that’s not a bad thing.

On to Part 3.

This is my notes from the talk I did at the 2009 Plone Conference in Budapest, reworked into a series of blog posts. This is part 1.