Talk:Anti-lock braking example

significant

Some people misinterpreted this as a software invention, but this is incorrect. ABS is an innovation in ways of using brakes. Brakes use "controllable forces of nature", and are an activity of "applied natural science", and they involve "significant physical activity". So it doesn't fail any of the tests we propose.

This implies that if your software controls or is controlled by something physical, it's patentable... including printers, scanners, mice, keyboard, knobs, switches, etc. If the distinction hinges on "significant physical activity", then what's the definition of "significant"? --38.97.97.99 15:40, 11 May 2010 (UTC)

You're correct - our definitions are still flawed. This is developed a little on choose your words. To prevent a printer or mouse from being able to confer patentability, Google's suggestion is that "conventional use of a machine" should be ignored. So, conventional use of a mouse or printer wouldn't make something patentable. Not perfect, but it's a starting point. Got an improvement?

Definition of "significant"? We don't have one. If you have a suggestion, that would be very welcome, but there's also some value in leaving it for the judges to decide. If we define it, then the patent lawyers have a rigid thing to try to get around, and judges will be required to give the patent lawyers a victory if they do find a way around that wording. If we leave it vague, the judges can adapt it to the circumstances of the day and apply common sense regarding the intention of the law. If the law says that the intention is to ensure that computer users can program and control that devices they've bought, then the judge can define "significant" as he/she sees necessary in order to fulfil the intention. Ciaran 14:25, 12 May 2010 (UTC)

Bad: Can be done manually, can be implemented with a single chip

There are people who build cars at home. There are people who build robots at home. There people who build almost anything at home. And don't forget small companies who want to build and sell safe custom cars!

In reality, this is just another software function in the braking system chip. I could build a car, control the brakes with a chip that use power linearly based on how much I push down the break pedal (old, default way). And then I can add a function that just don't apply full braking power to not lock the weels at certain high speeds, so that I can keep control. In fact, almost anybody could do it. A few lines of Python and an Arduino board is enough. Just read some documentation on it and you could be done in a few days, really.

Was the idea of being able to patent it a major incentive to inventing it? I doubt it. Is it even "inventive" enough? Making the brakes do something automatically that professional drivers has been doing for ages shouldn't really be patentable. If it's about "nobody knew how to implement it before", then I can safely assume that any existing patent are narrow enough to hardly even apply to 1% of the ABS brakes today (if I write one in Python, you can be sure it's not covered).

Or as a judge has said: Just adding a computer to it is not enough to get a patent. Natanael L 13:49, 12 May 2010 (UTC)

Hi. Yes, there are people that build cars at home, but their quantity of production is so small that the car patent holders have no motivation to sue them, so even if the law is bad, the negative effects can be ignored because they never materialise. What you do with an Arduino board will probably never draw the wrath of patent holders - at least, not until you get to the stage where you have the cash and lawyers required to go into the mass production of you idea. At that stage, you'll have plenty of regulation to deal with. Patent suits will be one of many.

In those fields where the players can be assumed to have stacks of cash and lawyers, what remains is to do an economics analysis. (maybe car patents would fail that analysis - but that's not my area - but it can be developed here: Should the whole patent system be axed?)

In contrast, software patents require this economic analysis plus the social analysis. (see: Why focus only on software) The right to develop software is one that's used by ordinary members of society - ones that aren't organised into structures with war chests and legal departments. In this situation, patents are plain unjust.

So, the line we're looking to draw is not "where are patents good and where are they bad". Instead, we're looking at where patents can't even be considered. If the act is performed only by people who have the facility to manufacture cars, then we say patents can be considered there - it's just an economic issue. If the act is performed by ordinary folk and non-commercial communities (as well as mega corps etc.), then we say that patents in that field should be a non-starter. Did I reply to the right question? Ciaran 14:17, 12 May 2010 (UTC)

Almost. What's the lowest requirement to get a patent? What's the definition of non-trivial? Inventive height? Are you sure that individuals never will be sued? It *has* happened before, and will probably happen again. That last line on my comment above says the most about what I am interested in.

As the wiki says in several places, it's not about fixing the system - it's just not worth it. Getting rid of it is better. But where should the line be drawn? How much of the system should go? Only plain software patents, or anything that a hobbyist can build at home?

Consider RepRaps, Arduino boards, the scripting language Python, and what it means for the future. What's difficult for large companies today could be a piece of cake for hobbyists tomorrow. Then what? Should the patent system of today not consider the situation in 10 years from now and stall development for another 10 years from there?

And as the other person said above, and what I partially said before, is that just using software to apply a method to some physical process just shouldn't be enough for a patent. Natanael L 10:12, 14 May 2010 (UTC)

Ah yes, there are two points worth addressing more clearly there: perfection, and the future.

We'll never get to perfect. And since legal contexts, technological contexts, and the science of twisting laws are all in motion, anything that's perfect one day, won't stay perfect forever. When we "win", our struggle won't stop. So for now, rather than saying that we want to go to "number 7, Bayani street, North Manilla District, Philippines", we're still at the stage where we have to convince the legislators/judges to go in anything resembling a South-South-Easterly direction.

I'm not trying to dodge the question. It's something we have to work on, but those are my thoughts on why it's not always a top priority (and we have sooo much work to do, so it doesn't get looked at much - if you want to jot down some notes on it, just start a page on the wiki). One of the best attempts so far to answer the question you've asked is here: EU 2005 proposed amendments#What is patentable

But if we had a chance to propose legislation in the EU next week, there'd surely be a mad scramble to improve on that.

If we can get any type of software removed from the patent system, then that sets the scene for the future. That says that some new fields enter the scope of patentability, and some fields leave. Today, software patents are a problem and they should make an exit from the patent system. If the proliferation of RepRaps causes patents on another field of ideas to become problematic, then that field should make an exit from the patent system too - and if we can set a precedent via software, then the exit should be clearer for this new field too when the time comes.

Bilski requires us to do less drafting work than the EU directive did because the US Supreme Court is much more skilled at legal drafting than the European Parliament is. So the work on Bilski was even more focussed on the "Please head in a generally South-South-Easterly direction" approach, unlike the EU directive, for which we had to nail down an address. Ciaran 10:43, 14 May 2010 (UTC)