Software Patents

[melandron]

As you may (or may not) be aware, the issue of software patents is still burning in the EU.

In order that i can gather a nice balenced view, do you think software patenting (sp?) is a good thing or a bad thing, are you for or against the bill tightening rule description.

At the moment i seem to be swaying towards the idea that software patenting is just generally bad for inovation, with some stupid patents being awarded.

[JonnyBGood]

I blame the government mandate. Isn't it great how you can always avoid the blame when your chosen economic ideology starts being shit?

[queball]

Originally Posted by Toccata & Fugue

Paging Dante.

He has a rant here

That's not actually about software patents. Though I do agree with the general pro-Dante ideology of this post.

[meglamaniac]

No, they're a shit idea.
As has been seen in the US, it leads to companies applying for patents for bloody obvious things that a school child could think up just so they can then wave their patent around and sue anyone who doesn't licence the "technology" from them.
As usual, small companies and individuals are the ones who lose out here which is stupid, because historically that's where the innovation has come from in software.
A relatively recent example of a company set up purely to try and shaft other companies with software patents can be see here. In this case the company getting shafted was Microsoft so sympathy wains a bit, but you can see where this is going.

Basicly, it'll make the cost of legitimate software development skyrocket, as the legal teams required to make sure your slightly improved version of the Windows Calculator doesn't infringe any patents will be huge. Never mind for real software projects.

[JC]

Originally Posted by meglamaniac

As has been seen in the US, it leads to companies applying for patents for bloody obvious things that a school child could think up just so they can then wave their patent around and sue anyone who doesn't licence the "technology" from them.

In the Biotech industry you can only get a patent for your idea/product/whatever if it is truely innovative. By that they mean a team of post doc researchers who specialise in that area couldnt have come up with that undre normal circumstances. Basically that means your idea has to be extremely innovative if you are going to get a patent.

If what you say about these cases in the US is true then they could do with a similar law to solve that problem.

[MrL_JaKiri]

Originally Posted by meglamaniac

As has been seen in the US, it leads to companies applying for patents for bloody obvious things that a school child could think up just so they can then wave their patent around and sue anyone who doesn't licence the "technology" from them.

That's not really the problem. The problem is that patents can only be challenged in a legal challenge by the owner of the patent, or something to that effect. This means that, in order to challenge a patent, people first have to break the patent and be sued by the owner of the patent, which, as these mostly affect small companies, never happens because they could never afford the legal fees, or the damages if they lost.

See: iD (and John Carmack, specifically)'s little spat with Creative.

[meglamaniac]

I missed that one, what did our beloved John do to get up Creative's nose?

[MrL_JaKiri]

Originally Posted by meglamaniac

I missed that one, what did our beloved John do to get up Creative's nose?

http://www.techspot.com/story14454.html

[queball]

Originally Posted by JC

In the Biotech industry you can only get a patent for your idea/product/whatever if it is truely innovative. By that they mean a team of post doc researchers who specialise in that area couldnt have come up with that undre normal circumstances. Basically that means your idea has to be extremely innovative if you are going to get a patent.

If what you say about these cases in the US is true then they could do with a similar law to solve that problem.

Of course patents only cover non-obvious innovations. The US patent office seems to have some trouble applying that rule.

[Phil^]

if getting a patent required the full source code to be opened up, that might cut the clear abuses of the system which will result, but it wont stop it.
Copyright is a perfectly acceptable alternative for code, theres no need to patent it since code is in essence a mathematical operation and you cant patent maths

[Dante Hicks]

Originally Posted by queball

Of course patents only cover non-obvious innovations. The US patent office seems to have some trouble applying that rule.

I think this is the clear problem (even if you supported the notion of patents more generally). It seems rather difficult to identify how "obvious" something is, even if you're an engineer working in the field. Different approaches are going to find certain things obvious (is a double click a natural extension to single click or something new?). On a very small scale whenever I ask anyone else for help on a software project / problem I am always amazed that they often approach from a completely different angle from me (that I wouldn't necessarily ever have been able to come up with). What does this mean, ultimately?

But anyway even if you support copyright laws, patent laws seem totally morally unjustifiable. The idea that I could work on something, in total isolation from the rest of the world, invent something, and then be forced to pay money to someone else simply because I happened on the same solution as them, is well, horrible.

The consequences of the current patent system (especially in regards to software and "business models") are both obvious and evident : a land-grab where major companies will simply try to patent any idea ever in the hope it proves useful to someone, eventually. The United States (with a very corporate dominated Congress) will be quagmired with this nonsense for some time, but Europe still has a chance to avoid it. Partially at least.

[Yahwe]

Originally Posted by Dante Hicks

I think this is the clear problem (even if you supported the notion of patents more generally). It seems rather difficult to identify how "obvious" something is, even if you're an engineer working in the field. Different approaches are going to find certain things obvious (is a double click a natural extension to single click or something new?). On a very small scale whenever I ask anyone else for help on a software project / problem I am always amazed that they often approach from a completely different angle from me (that I wouldn't necessarily ever have been able to come up with). What does this mean, ultimately?

But anyway even if you support copyright laws, patent laws seem totally morally unjustifiable. The idea that I could work on something, in total isolation from the rest of the world, invent something, and then be forced to pay money to someone else simply because I happened on the same solution as them, is well, horrible.

The consequences of the current patent system (especially in regards to software and "business models") are both obvious and evident : a land-grab where major companies will simply try to patent any idea ever in the hope it proves useful to someone, eventually. The United States (with a very corporate dominated Congress) will be quagmired with this nonsense for some time, but Europe still has a chance to avoid it. Partially at least.

i defer to your knowledge in re software.

all i can say is that the logic of patents is clear.
patents derrive from a far simpler world
the logic of patentents is pure

"if you don't allow the invetor to gain, no one wil bother to invent".

as i understand it however the modern question has grown beyond that pure logic. in several ways modern patents have grown beyond pure logic. global economics undermine national rules. regarding software (and i confess i do not know enough to say yay or nay) but if the US has them and the EC decides not to, there is still not an issue. software transends national boundaries and the people in the EC, should the EC decide not to allow such patents, will in practice still be sued by in the US

similarly if the US decided not to and the EC decided to, they'd sue in the EC. patents are a demonstration of what we call the global market place.

the reason they are silly, yet still here, is the pure logic of their necessity.

[Dante Hicks]

Originally Posted by Yahwe

"if you don't allow the invetor to gain, no one wil bother to invent".

I think the issue here is that people are willing to invent (again, perhaps only with regards to software, I'm not sure) without personal gain*. Or that gain could be defended without recourse to patents (trade secrets, copyrighted code, etc).

* = I am putting this very poorly. 90% of people who contribute code to free software projects are doing so for personal gain in some form another, albeit not direct financial gain. Perhaps they crave respect from their peers, or they wish for their CV to be bolstered by additional experience. Or they want code that does specifically what they want. Other people write code so they can make money through support contracts. EIther way it is not a direct production / sale arrangement in any case.

[Yahwe]

Originally Posted by Dante Hicks

I think the issue here is that people are willing to invent (again, perhaps only with regards to software, I'm not sure) without personal gain*. Or that gain could be defended without recourse to patents (trade secrets, copyrighted code, etc).

not poorly at all.
a patent is just one of many legal tools.
in law it is the equivalent of putting a name tag on your property.
patents, as i understand but but but i might be wrong because i don't really know the software industry, should just stop someone taking your new invention and using it themselves.

they've always been good and bad. they stop growth by copying but they stop copying.
and at least patents don't last forever
tho i suppose computers move so fast there is an argument they should be even briefer there

[mist]

software, in many respects, is different to things such as drug licencing in that a lot of software patents require little or no 'origonal' research.

for example, take semaphores (if you don't know what these are this isn't going to make a lot of sense, but they're one of two ways i'm aware of to achieve an end, and the other way really really sucks). they're blatantly obvious and require about 5 minutes of thought to come up with. had someone patented them then real time computing would be rather screwed.

another area that worries me is that of prime numbers. tbh, i'm not quite sure why majino (sp? those primes that are (n^2)-1) are important, but opening up software patents would seem to indicate that you can start patenting this kind of thing as well, and realistically it's a number that you just happened to stumble accross first. would/should this be acceptable?

if the EU were to be sued by the us for not having software patent laws (why hasn't it been already? it's not like an american to miss a chance to sue someone...) would this apply to software soley launched within the EU?

[Yahwe]

Originally Posted by mist

if the EU were to be sued by the us for not having software patent laws (why hasn't it been already? it's not like an american to miss a chance to sue someone...) would this apply to software soley launched within the EU?

first to say i don't understand the rest.
but to clarify. what i meant was that in software as in most big companies, they 'exist' in a legal sense in most countries

it wouldn't be the US 'suing' the EC it is just that if either the US or the EC makes it illegal then you can sue under their law, most companies and distrubotors exist in both.
to make it clear, with software:

A patents, B breaks that patent,

A is a multinational company, B is likely a multinational company.

it makes no real difference if A patents in the US or the EC. if he can patent in either then he can still sue B in either the EC or the US.

[mist]

i was thinking from the perspective of web based companies

for example, if pa were to take out a us patent on the idea of roids in online games, would plannetia, also hosted in the uk i believe, be in any danger of being sued?

[melandron]

Originally Posted by mist

i was thinking from the perspective of web based companies

for example, if pa were to take out a us patent on the idea of roids in online games, would plannetia, also hosted in the uk i believe, be in any danger of being sued?

think the fact that they've operated before patent was filed is evidence of prior use, and thus the patent would be invalid

[melandron]

Also code in the UK is currently covered by a certain amount of copyright and IP laws, so would a patent offer any further benifit?

[meglamaniac]

The problem seems to be that judging from court cases so far (in the US at least), prior art has to come from before the date given for the invention, not the date the patent was filed. Thus companies file patents for things that've been around for years which they think they might have a reasonable claim on, and then try to backdate the patent.

If that sort of thing starts happening over here, we're all in big shit.

[hyfe]

Originally Posted by Yahwe

not poorly at all.
a patent is just one of many legal tools.
in law it is the equivalent of putting a name tag on your property.
patents, as i understand but but but i might be wrong because i don't really know the software industry, should just stop someone taking your new invention and using it themselves.

Yes, but here are you defending the good side of patents, which supposedly exists.

The bad side of software patents are the obvious ones, who merely are food for lawyers and scumbags and stiffle innovation and the 'freedom of the market'. And plenty of them does exists, both held by "legitimate" companies and patent-holding scumbag companies.

I think Dante's point goes that regardless of the morality, poor implementation of the current system, a good implementation is next to impossible due to difficulty involved in detemerning if something actually is obvious or not. If we have trouble deciding on a micro level, how on earth are we going to decide general guidelines for it?

Besides, the difficulties in software lies in the implentation, not the ideas. Plenty of people with good ideas around, but a severe lack of people with broad skillset required to make something workable.

[roadrunner_0]

wasn't it the polish guy in the EP who voted against this?

[meglamaniac]

Yeah.
Poland in fact blocked it again just recently, as described here.

Hurrah for poland etc!

[mist]

they may be buying all our lard out from under us, but for that they get respect \o/

[Nodrog]

Originally Posted by meglamaniac

No, they're a shit idea.
As has been seen in the US, it leads to companies applying for patents for bloody obvious things that a school child could think up just so they can then wave their patent around and sue anyone who doesn't licence the "technology" from them.

In Iran the courts often jail people for stupid reasons therefore having courts is a bad idea.

[Nodrog]

Originally Posted by Dante Hicks

But anyway even if you support copyright laws, patent laws seem totally morally unjustifiable. The idea that I could work on something, in total isolation from the rest of the world, invent something, and then be forced to pay money to someone else simply because I happened on the same solution as them, is well, horrible..

That's pretty unlikely to happen. Probably a lot less likely than an innocent person being jailed for a crime he didn't commit (since any legal system is going to make mistakes), and yet the fear of punishing innocent people is generally not thought to be severe enough to justify scrapping the legal system altogether. Rather, the legal system is designed to try and minimise the liklihood of this occurring, although it is still recognised as an unfortunate possibility.

I'm not sure why this is fundamentally any different.

[Nodrog]

Originally Posted by Dante Hicks

I think the issue here is that people are willing to invent (again, perhaps only with regards to software, I'm not sure) without personal gain*. Or that gain could be defended without recourse to patents (trade secrets, copyrighted code, etc).

* = I am putting this very poorly. 90% of people who contribute code to free software projects are doing so for personal gain in some form another, albeit not direct financial gain. Perhaps they crave respect from their peers, or they wish for their CV to be bolstered by additional experience. Or they want code that does specifically what they want. Other people write code so they can make money through support contracts. EIther way it is not a direct production / sale arrangement in any case.

This is naive. Making 'free' software has very little overheads, whereas developing a cure for cancer requires an investment totalling millions of pounds. While I'm sure you would have very little trouble persuading some unpaid linux nerds to develop a kernel patch, you may well have a harder time motivating large corporations to invest money into R&D when you are happily allowing others to steal whatever they eventually come up with.

Not that I think that the pragmatic justification is the primary issue.

[Nodrog]

Originally Posted by hyfe

Besides, the difficulties in software lies in the implentation, not the ideas. Plenty of people with good ideas around, but a severe lack of people with broad skillset required to make something workable.

I'm unsure how the current patent system works, but a properly functioning one would only grant patents to people who intend to implement their ideas, and would expire after a given time if they failed to do so. It would be absurd to simply grant patents to whoever asked, regardless of whether they had any intention of putting them to use. Even if the original thinker intented to build something but proved incapable, someone else would be able to pick it up after the patent expired (I'd personally say patents should last around 10 years in the longest cases).

[Nodrog]

Originally Posted by mist

another area that worries me is that of prime numbers. tbh, i'm not quite sure why majino (sp? those primes that are (n^2)-1) are important, but opening up software patents would seem to indicate that you can start patenting this kind of thing as well, and realistically it's a number that you just happened to stumble accross first. would/should this be acceptable?

Mersenne primes.

I'm not sure why a number would be patentable - that isnt how the law works now afaik (leaving aside "hurr hurr a software program is just a long number!" arguments) and there's reason why it should work that way. However if someone invented a previously unknown algorithm for discovering Mersenne primes, then granting him a temporary patent to enable him to profit from this (if possible) wouldn't be unreasonable.

for example, take semaphores (if you don't know what these are this isn't going to make a lot of sense, but they're one of two ways i'm aware of to achieve an end, and the other way really really sucks). they're blatantly obvious and require about 5 minutes of thought to come up with. had someone patented them then real time computing would be rather screwed.

All you're really saying is that obvious things shouldnt be patentable. I agree. Having sex with drunk people shouldnt be classed as rape either, but this isn't a problem with rape laws in an abstract sense, simply particular legal systems.

[MrL_JaKiri]

Originally Posted by Nodrog

leaving aside "hurr hurr a software program is just a long number!" arguments

http://home.att.net/~jbcole/humor/Microsoft_patents.htm

[Super]

I think if these patents were to exist, then they should expire very quickly - say after 2 years or something. This way, the person who came up with the original idea can get the monetary benefit from it, and then when demand dies down it can be applied by others.

[mist]

Originally Posted by Nodrog

I'm not sure why a number would be patentable - that isnt how the law works now afaik (leaving aside "hurr hurr a software program is just a long number!" arguments) and there's reason why it should work that way. However if someone invented a previously unknown algorithm for discovering Mersenne primes, then granting him a temporary patent to enable him to profit from this (if possible) wouldn't be unreasonable.

i'm unsure as to the specifics, so someone will probably correct me, but i was under the impression that the xbox uses specific hardware keys to unlock the box's hardware, hence making it very difficult to 'chip' it. iirc, the chippers had to come up with a way around using these numbers, rather than just using the numbers. therefore, i'd have thought the numbers were somehow legally protected?

likewise, don't digital signatures have something to do with this?

[queball]

Originally Posted by mist

i'm unsure as to the specifics, so someone will probably correct me, but i was under the impression that the xbox uses specific hardware keys to unlock the box's hardware, hence making it very difficult to 'chip' it. iirc, the chippers had to come up with a way around using these numbers, rather than just using the numbers. therefore, i'd have thought the numbers were somehow legally protected?

likewise, don't digital signatures have something to do with this?

Digital keys are covered by trade secret law, and more importantly physical security.

[JonnyBGood]

I wish I could patent queball

[Marilyn Manson]

Originally Posted by JonnyBGood

I wish I could patent queball

He wouldn't be very good sex slave material.

[queball]

That's not very nice.

[melandron]

yey, just when i was looking at the problem another stupid example comes hot off the press
http://www.betanews.com/article/McAf...ent/1106703454

apparently your firewall will no longer be able to log where attacks come from, or atleast not without McAfee getting a licence fee.

[Cyp]

Originally Posted by mist

... (sp? those primes that are (n^2)-1) ...

You probably meant 2^n - 1, since the only n^2 - 1 prime for integer values of n is 3.

[mist]

i did, yes.

[Dante Hicks]

Originally Posted by Nodrog

That's pretty unlikely to happen. Probably a lot less likely than an innocent person being jailed for a crime he didn't commit (since any legal system is going to make mistakes), and yet the fear of punishing innocent people is generally not thought to be severe enough to justify scrapping the legal system altogether. Rather, the legal system is designed to try and minimise the liklihood of this occurring, although it is still recognised as an unfortunate possibility.

I'm not sure why this is fundamentally any different.

Nod, this is a terrible analogy. If the legal system identifies that they have imprisoned someone incorrectly that person is released, and in most countries the person is compensated. Under the patent system, how you come across an idea is irrelevent. I'll say this again : the patent system does not just stop people "stealing" (lol) your idea, it also prevents people from arriving at the same idea independently. I don't see how you can dismiss this as an unlikely event (empirical evidence points the other way) - even if that were true it'd be akin to saying "Well, we'll ban people from performing x activity on the grounds not many people do it".

All you're really saying is that obvious things shouldnt be patentable. I agree.

How would you determine what is "obvious" and what isn't? Obvious to who exactly?

I'm unsure how the current patent system works, but a properly functioning one would only grant patents to people who intend to implement their ideas, and would expire after a given time if they failed to do so.

I thought you said pragmatics weren't the primary justification? Since you are a believer in intellectual property rights, I don't see what your justification here is. If I own a piece of land and I refuse to develop it, do you think other people should be able to take it away from me?