In software, all of them are. It's like patenting a mathematical formula.

It's bad enough that the USPTO will rubberstamp everything with the right fees, but that they can issue patents on concepts that are far outside the original intent what patents are for.

Yes, developing software involves "invention" in every sense of the word, but if you can patent algorithms, why can't you patent mathematical formulae or proofs?

" It's like patenting a mathematical formula." More like an Mathematical Algorithm but even then, most devices are physical representation of a algorithm in some way. I don't see how this is a valid attack on software patent or patents in general.

I don't support software patents but I have a problem with the whole "it's patenting a mathematical formula" argument.

The US courts do place weight on the "it's patenting a mathematical formula" argument. There's some nuance and murk, but an application will probably have to pass the "Machine or Transformation Test" to be patentable: http://en.wikipedia.org/wiki/Machine-or-transformation_test

Essentially, software implementations of algorithms cannot generally be patented either because they are simply an expression of the algorithm in a particular language. The fact that this language can be interpreted by a particular hardware device does not change the situation. The hardware is a pre-existing piece of general-purpose commodity hardware that is easily interchangeable with any number of other pre-existing pieces of general-purpose commodity hardware, up to and including a human being equipped with a pen and paper, and therefore cannot really be claimed as part of the patent.

A physical device which implements the algorithm can, of course, be patented. And a hardware/software combination where the computer is an integral part of the mechanism (e.g., it cannot in principle be replicated by a human with a pen and paper) can also be patented, so Amazon's 'One-Click' patent is safe. However, simply tacking on the phrase, "Written in a programming language," does not take an unpatentable idea and make it patentable.

> It's bad enough that the USPTO will rubberstamp everything with the right fees

Clearly you have never prosecuted a patent application. Yes, examiners make mistakes sometimes, and patents issue that shouldn't. But your statement is just nowhere near being true.

I agree that GP overstated the problem, but I think you understate it. The USPTO systematically and recklessly grants patents that should not be issued.

Do you not need to actively defend patents like you do with trademarks?

There is a defense against patent infringement called "laches", but it's not very broad. It only applies to specific cases where the patent holder knew of the alleged infringement and did nothing for a long time. It doesn't lead to a general invalidation of the patent.

You need to claim damages withing 6 years of the infringement:

35 U.S.C. § 286 Time limitation on damages

Except as otherwise provided by law, no recovery shall be had for any infringement committed more than six years prior to the filing of the complaint or counterclaim for infringement in the action.

In the case of claims against the United States Government for use of a patented invention, the period before bringing suit, up to six years, between the date of receipt of a written claim for compensation by the department or agency of the Government having authority to settle such claim, and the date of mailing by the Government of a notice to the claimant that his claim has been denied shall not be counted as a part of the period referred to in the preceding paragraph.

Wait... why wouldn't you release it to the world?

Because there's no gain in doing so.

Look, as soon as it's released, it will be copied by those who already have market penetration in that space, leaving the big USP of "cool software feature" as something useless, and no longer a USP.

Instead of releasing them out, which is what I'd like to do, I just keep them to myself. Something which I would rather not do, but I see no other way of gaining value from them.

Except the software industry is full of stories opposite of your expectations--startups with better ideas who unseat or out maneuver giants. Do you think Mint would have been better off keeping their ideas to themselves? How about Facebook? There are tons of these: Dropbox, Square, Heroku, Twilio, etc etc.

Good software is rewarded (even more so in today when there's instant distribution), don't worry about your competitors.

It might be instructive to ask yourself why the software industry is full of stories opposite brador's expectations. For example, consider that by definition, there can be no stories about technologies that have been kept secret.

(I dislike software patents myself and agree with you, though, that, for a variety of reasons, we'd still see plenty of innovation without them.)

I'm not one of the downvoters, but I can speculate as to the reason:

Perhaps it's because you're using most likely a non-patented browser that fetches a non-patented hypertext format (HTML) via a non-patented application protocol (HTTP) built on a non-patented network stack (TCP) by connecting to a non-patented web server (HN) written in a non-patented programming language (Arc), and you're using this stack to argue that people have no motivation to innovate in software if patent protection is unavailable.

Or perhaps it's because you're trotting out that old argument that software patents can help the little guy compete against big players. But so far the only "little guy" success anyone can point to was Stac Electronics, a publicly traded company for whom a $9.7 million sales quarter was evidence of declining business. That's not little: http://articles.latimes.com/1994-06-22/business/fi-7159_1_pa...

>"Looks like I'm getting downvoted for no reason. "

Yeah, there seems to be a lot of that these days. People down-vote for disagreement, as opposed to just spammy posts. It's really kinda sad for a community that should be open to discussion and critical thought.

This underscores an oft-ignored important part of the patent system, which is that it encourages people to make their discoveries known to others by patenting them in exchange for temporary monopoly. If the patent were completely unavailable as an option, many would be tempted to hide or obscure their discoveries and methodology for as long as possible to make a profit, so there is at least an argument to be made for patents serving the public interest in the long term.

Unfortunately, temporary often ends up being effectively forever when it comes to software, and the system has been gamed by lawyers. Sigh.

Not only in software: http://en.m.wikipedia.org/wiki/Method_of_exercising_a_cat

The documentary by This American Life about patents had an IP researcher claiming that over 1/3 of all patents ignore prior art, just by comparing identical text between them: http://www.thisamericanlife.org/android-app?destination=radi...

I'm curious how many patents there are for devices that are impossible to actually build. You know, fun stuff like warp drives and time machines.

Presumably zero. Patents are only permitted for "useful" subject matter. Useful is a term of art: you can still patent silly things like Slinkies: http://www.google.com/patents/US2415012?printsec=drawing – but whatever you patent has to actually function as advertised. So while the Slinky is objectively useless, it does what the patent says it does. A warp drive would be objectively useful, but when you build it, it won't do what the patent says it does; therefore, it's not patentable.

Though not quite in practice of course. I cannot substantiate my claim with data, but have come across several (approved) patents which are actually not even correct -- the described does not function as advertized.