Everyone and the birds in the trees have been expressing an opinion on software patents, but no-one that I have seen is making, what seems to me to be the obvious argument - the inevitability argument. I was moved to make this argument in a Blog comment recently, and I amplify those remarks here.
The argument is this: software patents, on general purpose computers (networked or not, devices, servers or desktop), are "obvious to one skilled in the art." This is one of the phrases that the USPTO uses to reject patent applications and its application in software patents is insufficiently broad. Here's why. Solutions to new applications are inevitable. They are the product of the general purpose computer in the hands of an individual using well-know principles that have inevitable consequences. Just because you were the first to consider the problem does not justify a patent award. The solution was the inevitable product (one of an inevitable few) of a general purpose computer architecture and a priori computer science. I hold such a patent, US Patent 6,519,771. The Amazon one-click and Gemstar one-touch patents also come readily to mind.
Let's consider some patent examples to explain. Was the electric light bulb patentable? Yes, because Edison had to invent fundamental new technology - he invented electrical engineering associated with his inventions, he was not a trained electronics engineer, no-one was. The electric light bulb was not the inevitable product of existing training and technology. Was VisiCalc patentable? No, because spreadsheets already existed in the world and their implementation was inevitable on general purpose computers.
If general purpose computers did not exist and had Briklin and co. built a special purpose machine that was a VisiCalc machine, then that would be patentable in my view.
The software patent gold-rush has served only multinational corporations; it has not served individual inventors or the general public. Corporations are hijacking the intellectual property of employees whose "work-for-hire" contracts deny them any reward for their so-called invention, the bounty paid at companies like Microsoft for a patent award (as I recall , a few years ago) is an insulting few hundred dollars. Meanwhile Corporations are raping computer science, and reaping unjust rewards by laying claim to the inevitable consequence of established engineering practice and technology.
Are any software patents justified? I don't think so. I can see how software can be a part of a patent, or a component of the programming in a special device, but I simply cannot justify patents on general purpose computers.
To summarize, I argue against a particular type of patent on general purpose computers - generally called "software patents." Patents that are the inevitable product of existing knowledge in computer science and a general purpose computing platform; such that anyone skilled in the art (with a conventional CS education) would have produced a comparable solution when faced with the application. I have given three examples of this type of patent - including one that I hold personally.
If you have seen this argument before, excuse my rant, I have simply not seen it.