In theory they should make sense if * copycats would have significantly lower costs without any offsetting disadvantages, and * what's being pate

In theory they should make sense if * copycats would have significantly lower costs without any offsetting disadvantages, and * what's being patented would not be independently re-discovered before the patent's expiration In practice neither happens: copycats tend to be comparatively incompetent, and simultaneous inventions is more the rule than the exception. I think the core disconnect is whether those two points are taken as true or false: you get the imagery of some lone heroic inventor being always crushed by some conglomerate that patents could save him from; a more accurate version would be a bunch of small businesses that keep doing cool things at more-or-less the same time which the conglomerates pick up on after a year or two, and nobody has the concept of going broke on legal fees arguing over who's allowed to do what. > Yeah, these monopolies make sense if I have to set up a plant, or do years of research, but that's not true in software anymore. The hard part of software is figuring out what to do, reimplementing a program is easy (thus the prevalence of programs that are "clones" of other programs). If patents actually worked, I think software would probably be a nearly-ideal example.