I agree that these were historical arguments for the patent system. But I don't think those arguments were entirely valid then, and they're even more suspect to

I agree that these were historical arguments for the patent system. But I don't think those arguments were entirely valid then, and they're even more suspect today. The end of trade guild secrecy had far more to do with the advent of mass production than with the patent system. Highly trained craftsman could be organized into a guild. But large numbers of easily replaceable workers necessarily couldn't. As for the elimination of generic drugs, the drug makers themselves argue that figuring out how to copy most new molecules is so cheap and easy that they need strong patent protection to recoup their research costs. Reverse engineering drugs is cheap and getting cheaper. Software is not going to get more secretive without patents. It's hardly possible that it could -- almost all commercial source code is already treated like trade secrets. The implication that software developers actually utilize techniques gleaned from patent applications is pretty far-fetched. Security-by-obscurity is orthogonal to the issue of patents. All the peer-reviewed security algorithms that actually get used are unencumbered by patents, which is precisely why they actually get used. (The slow adoption of elliptic curve cryptography has been blamed on the existence of certain patents.) There's no incentive to design your own crypto algorithm and keep it secret -- that's unnecessarily expensive and stupidly risky. People share these techniques out of self interest, not because they enjoy patent protection. The cost of spreading information continues to plummet. Conversely, the cost of keeping secrets continues to increase. I don't think the premise of a permanent monopoly of secrecy is remotely plausible. Even military technology (which of course relies only on secrecy, not patent protection) seems to proliferate on time scales shorter than patent lifetimes, though this is a hard contention to prove.