>> Which for software would mean an absolute maximum of 10 years, with 3 years much more reasonable. This still devalues the creation of software that is genui

>> Which for software would mean an absolute maximum of 10 years, with 3 years much more reasonable. This still devalues the creation of software that is genuine innovation. Think about mainframe software that controls factories, for instance, which can be bought over 10 years after creation. Also consider the prevalence today of Windows XP, which shipped in 2001 (and where any relevant patents would date to the 1990s when it was developed). Microsoft has reluctantly stopped selling it, but I'd argue that we still don't know how long the useful lifespan of WinXP will be. And we're probably ~15 years in from when patents in XP would have been filed. Before artificially segregating software into a ghetto, ask this: would this innovation be worthy of protection if there were a purpose-built device that performs the same function? If the device would be worthy of protection, then it's going to be hard to argue that the software embodiment is somehow worth less protection. (At least it will be hard to argue this with me, but I'm a chauvinist for programmers.) Finally, the argument that the old software becomes irrelevant is congruent with equal protection for software. If old software loses relevance, then so do the patents (and correspondingly their utility to patent trolls). I think you mean precisely the opposite: old innovations in software actually become more relevant, so that everyone wants to use them in their own wares. Nobody rushed to shorten the lifespan of vacuum tube patents because they became irrelevant. I don't think it's reasonable to reduce software patent lifespans to shorter durations than that of the hardware the software runs on, or the cables used to transmit it.