Also, the patent shouldn't be on the idea, but on the exploitation of an idea. If a given entity makes more money from a patent by licensing it compared to actually using it, it's a pretty good sign that the society didn't benefit from the patent but actually granted a temporary monopoly on something and that's actually slowing down progress.
For example, having a touchscreen that makes things zoom by pinching. How else would anyone do that? It's trivial, obvious, and there actually isn't any other useful way you could do it.
I'm all for bounties on things that are actually difficult, but having ideas, and making software, isn't hard.
The problem with implementing stuff is that it's a lot harder to build something the first time than the second. It's easy to sit around in a chair and imagine the UI from Minority Report. It is a ton of work to actually design and build that interface, because for every feature and tweak you include in the final product you must test and discard dozens of others. (Unless your business is willing to produce poorly-designed crap, of course.)
Unfortunately, a designed product is much easier to copy than to produce in the first place, so every piece of R&D you do is also R&D for your competitors. The patent system tries to rectify this.
You complain that patents let people have an idea and then sit around charging others money. And this is true, but - importantly! - it is only true for the life of the patent, which is something like 14 or 17 years. That's the whole point. Once someone has has a brilliant idea they must either build it, or license someone else to build it, or within less than two decades the idea will be worthless. (You can also pray that nobody else has the same idea. This is usually a bad bet: ideas arise when their time is right, and lots of things are invented simultaneously by half a dozen people at once.)
In a world without patents, nobody has any incentive to be the first to implement anything bold - your competitors will freeload on your R&D and outcompete you. So you're reduced to tiny incremental improvements with low risk and obvious value (so that they can be sold immediately in high volume, to take advantage of the tiny time window when the innovation is exclusive to your product.)
The problem with software patents is not that patents are useless or don't work in general. It's that they don't work well for software, which has many degrees of freedom and is hard to understand. It's that the standards of review are broken; far too many obvious patents are issued. The adjudication is broken: by rubber-stamping patent applications and leaving it to courts to adjudicate them, the Patent Office turns IP into a shakedown racket where the team with the largest legal budget wins. And the whole idea is just too new. If software patents had existed in 1968 Doug Engelbart would have patented everything we use today, those patents would have expired in the eighties and we wouldn't be worried about them anymore.
I never suggested a world without patents, not at all. I certainly see the usefulness of patents for areas where you genuinely have to put in years of R&D, where problems are hard, etc. But software/UI certainly isn't an area like that.
If someone else thinks it's worth doing that work, that suggests there is a payoff to implementing it. If you are in the position to implement it, you probably stand more to gain from doing so yourself than hoping somebody stumbles into it.
If you're not in the position to implement, or perhaps not on the right scale, it's in your interests to license it to somebody who does or for them to buy you out. It would be pointless to just leave it unimplemented and unlicensed.
That isn't to say patent trolling hasn't worked for some--only that it's a bug and not a feature. It has more to do with bad patent grants than a lack of incentive.
...where problems are hard, etc. But software/UI certainly isn't an area like that.
Then why do so many do it so poorly?
Because IMHO you have to have a talent/practice at it. It's trivial for anyone skilled in the art.
Apple doesn't actually have a patent on pinch-to-zoom. But I'll play along anyway...
It's trivial, obvious, and there actually isn't any other useful way you could do it.
Bullshit. You're playing dumb--probably without even realizing it--but I think if you actually had to apply yourself to the problem you could think up dozens of alternative approaches. These are the ones you're pre-filtering with "useful" because you assume that they are all inherently worse, but a huge amount of that is left up to the actual execution. Is this hard? Yes. But that's why it's competition. If it were easy to think of better ideas and create better implementations than the competition, the competition would have already done it!
So let's try to think of some other ways to zoom with a touch screen other than pinch-to-zoom. For starters, the easy one: virtual button presses. Tap the plus to zoom in, the minus to zoom out. Ok, that's pretty lame, how about a zoom track along one side of the screen, and you just move your finger up and down it, or perform some action to select and move a marker object along it.
Ok, maybe that's too much motion...lets make it a kind of rocker switch where you can just swipe your finger a small distance from a center point to activate the zoom one way, or the opposite to zoom another way. That might be great if meant for thumbs, kind of like a hat switch on a joystick, but let's say we want something better for a larger surface, like a table. How about tap and hold to get a menu with a zoom option. Hmm, menus are lame. Let's make it a dial instead.
Ah, but that requires two motions, which might not be good, depending on the rest of our interface. Instead, how about we can just select and drag from an object corner, like any windowing interface, binding the zoom level to the scale of the object in a very natural fashion. That could work for some things, though maybe not a map unless you've got a very large screen.
Or maybe rather than treating our interface as simply a flat surface we lend it some depth and treat the scale of the object as distance from the screen. Then we can use any gesture or button we like to suggest getting nearer or farther from the object we're looking it. Say, a two finger swipe downward, to suggest "come here" or "pulling towards" and upwards to suggest "go away" or "pushing away". When the object gets so near that it fills the screen we can keep using that gesture to get "nearer" to the object, effectively zooming in on it without the user consciously even thinking of it as zooming.
Or hey, maybe we're overthinking this. What's the simplest thing you can do with a touch screen? Tap it, right? Maybe we could...tap once to zoom in, tap twice to zoom out?
I didn't say there weren't other ways to control zoom on a touch screen, but IMHO all of the ideas you suggest are inferior, and less obvious - less useful.
The obvious, intuitive one, is clearly pinch to zoom.
I suppose you could have a bureaucratic body which tries to determine the value of that idea and then pays a bounty in cash, but that sounds like a terrible idea in all sorts of ways.
Almost everyone is a "modest libertarian". They all want the govt to intervene in some areas, but not in some others.
I think people who are opposed to patents could benefit from thinking about them in a different way. Instead of seeing them as "ownership of ideas" think of them as a bounty on ideas. The generation of new ideas benefits us all, so when somebody thinks up a new idea the government pays the inventor a bounty in the form of an exclusive right to make commercial use of that idea for a certain number of years. The great thing about paying it in the form of an exclusive right is that the value of the bounty automatically winds up being proportional to the value of the idea.
Now if you're a strict libertarian you can quite sensibly oppose the payment of bounties by the government, but if you're a more modest libertarian like myself or a non-libertarian then I think it starts to look pretty sensible.