Ridiculous Software Patents: a Developer's Nemesis
StormDriver writes "Have you ever thought about patenting a pop up note, an online poll, a leaderboard in an online game, or a system where you open apps by clicking icons? I have some bad news for you – it's impossible. Not because the claim is stupid, it's just that all of those things are already patented. And it's all fun and factoids, until one day you find yourself in the role of a software start-up."
"Software patents serve no one but giant companies, and only to stifle innovation. Exactly the opposite of their stated purpose."
However, that was always their purpose. It's always been the largest firms (particularly IBM pre-2000 and Microsoft post-2000) that expanded patent law into software, and that bought the most patents. Small firms have never mixed innovation with patenting, it's contradictory. You can either hold patents and sue others, or you can make products and avoid patents. Trying both means you are sued systematically unless you are in a field with practically no patents, e.g. a technical patent on a fashion item like a shoe.
Patents are inherently anti-competitive, this is well known in the patent industry, where people who actually make products and innovate are considered as a kind of food source for higher level patenting predators.
In software, the food source seems endless, which is why no-one's worried. But in other industries, it's already tipped so far that basic research is throttled in the US, EU, and Japan, and other countries are easily taking a lead.
All patents are bad, they all allow one entity to control the use of an idea in the market, they all act to restrict competition and they are all inherently anti-free market. There are no good patents.
My blog
Fortunately, patent terms haven't ballooned the way copyright terms have. Patents now cover up to 20 years from the first filing date (which can be many years before the patent is ultimately issued). In most industries that's pretty reasonable, but in software 10-20 years can be an eternity.
It seems like the best approach would be to change the patent term to whatever the length of a "generation" is for a particular industry, consulting experts in a given field to determine what that epoch may be. In automobiles, it might be twenty years. In software development, it might be two.
How can I believe you when you tell me what I don't want to hear?