From Robert O’Callahan, via Brenda, one of the most insightful things I’ve ever read on the subject of software patents:
In software, especially cutting-edge software like Firefox, every developer is an inventor; coming up with new ways of doing things is not exceptional, it’s what our developers do every single day. Invention created at such a rate does not deserve or benefit from years of monopoly protection. Indeed, it will be crippled if we are forced to play the patent system “to the hilt”, to acquire vast numbers of our own software patents and to navigate the minefield of other people’s patents.
This echoes my instinctual feelings about software patents – where do you even start? So much innovation happens so constantly in software development (ok, and so much reinvention of various wheels not invented here, but that’s another story…) that the patent system as is just doesn’t make sense.
hypatia dot ca 
I think there’s supposed to already be a protection against that integral to the patent system – the test for obviousness (or rather lack thereof) where a “person having ordinary skill in the art” wouldn’t have been able to come up with it. Obviously that definition is more than a little subjective, but I don’t think that’s avoidable.
What IS avoidable is that this test seldom seems to be applied, which is where I think the real problem lies – people are granted patents on things that really aren’t all that innovative, and just about any practitioner in the field could have come up with quite easily. As an example, last I checked Philips has a patent on using pulse-width modulation and RGB LEDs to produce polychromatic lighting – a technique that should be blindingly obvious to anyone who has an Arduino and took a Grade 1 art class at some point.
I do think that software should be patentable, just like every other field of invention, but the approval process has to be aware of what really counts as innovative. It also might be a good idea to give patents for a shorter term on such quickly-moving fields. The net benefit to society that patents are supposed to afford – well-documented, publicly-available records of how and why inventions work in exchange for a brief monopoly over commercial uses of that invention – is still a useful one, I think, so long as the barriers to entry are appropriate.
There’s also a general shift in tone away from patents (and copyright, for that matter) as a reward and inducement to continue inventing / creating, and towards a right of anyone who put pen to paper (or what not) to be able to control their “intellectual property”, which is at the very least ludicrous. Even the term is sheer insanity – it’s almost literally ownership of an idea, and the second anyone ever said they wanted to own an idea, they should have been laughed out of the room – but that’s another topic entirely.
It sounds a lot like the Tragedy of the Anti-Commons, which I first heard proposed concerning research in biology. If the Tragedy of the Commons is that because no one owns the resource, everyone overuses it, then the Tragedy of the Anti-Commons is that because so many people own the resource, no one can use it.
It seems software is even a worse state with this idea that a thicket of competing claims makes actually doing anything extremely difficult.
As a few people have pointed out, the idea of patents/copyright/trademarks, etc, was originally to provide enough protection for the investment to make it worthwhile, and then to free it up so others could modify it. The whole point was to encourage people to create and for those creations to enter the public domain and make people’s lives better. From what I know, the research on whether that ever worked is not so clear. (I would be happy to be pointed to more complete and recent data on that, though.)
My gut is much like yours, in that fields have differing needs and that software is just a terrible fit for patent law as it stands (and possibly even in general).
I thought the second half of the US Patent Office was to record these inventions, so they wouldn’t be lost in the mists of time. A crystalization of the Public Domain, it would be like a Smithsonian Museum or a US Library of Congress of ideas, which could be used when new innovators need to “stand on the shoulders of giants.”
But… if the ideas are under lock and key for so long that people need to re-invent them (or worse, invent workarounds), then the records of these patents will be of little or no interest once they enter the Public Domain — the records are not crystals being preserved, but calcified weights to be endured until they can finally be gotten rid of once the protection expires. Yes, gotten rid of. After 50 or 85 years anyone who needs that idea will have been using the unprotected work-around for at least the duration of one human generation, possibly two.