> The original idea was that it rewards sharing innovation. You're supposed to publish in detail the way your invention works and in return get legally protect

> The original idea was that it rewards sharing innovation. You're supposed to publish in detail the way your invention works and in return get legally protected exclusive rights to it for a decade or so. Setting everything else aside, academia is the one place where patent issuance is matched with a non-legalese description of the invention. Nearly all software patents in academic CS are matched by a series of academic articles, which aren't always super well-written, but are almost always infinitely better than the inscrutable JD-drafted legalese found in patent applications. It's not perfect, but academia (and industrial labs that function similarly to academia) get much closer to the spirit of "sharing how your invention works" than patents that are issued to basically any other type of institution. > I'm surprised software patents don't require source code to be published, that would fit with the spirit of the thing. I think this view comes from a pretty common misunderstanding of patent law. You don't need to actually build a thing to get a patent. There are tons of patents for e.g. power-plant components that AFAIK have never actually been built. You don't need to write a single line of code to get a software patent.