I am confident the law will soon be changed to make it such that only people who seek to make a profit from software can be sued for patent infringement.
Who am I kidding? That would be sensible, and so of course will never happen.
I'm always a bit dubious about Bruce Perens, he seems to be on the frothing-at-the-mouth side of the debate.
The first example he cites is someone suing RedHat - big deal. RH are a commercial company, and they're big and ugly enough to fend for themselves - the fact that the alleged patent infringement happens to be on an open source technology (which has always been controlled by a for-profit company, let's not forget) is neither here nor there.
The other one is a shame, and if it goes through as Perens indicates he thinks it will, that'll be a clear problem. That said, the patent itself, unless I'm reading it wrongly, looks very obvious and non-novel - if the guy can put together a defence, he'll probably be able to defeat it, and with any luck the company will be told to pay his costs.
As usual, I trust in the free market to do what's best - if it turns out that open source authors need legal protection, either their customers or companies will come up with the money to do so.
Now the problem here is that software is inherently unpatentable, or rather it has no benefits for society when it is patented and is excluded from patentability for good reasons in the EPC and UK Law, not that an open source company has been sued.
Further to that problem is that software patents have particularly ill effect on free software as patent licensing and royalties do not allow for products that are distributed and replicated at zero cost like software, and especially where anybody is free to distribute and modify the software.
So we have 2 problems, software patents are bad, software patents exist nonetheless and are harming open source software and therefore the whole industry and society as a whole. Software patents already harm society as a whole by restricting innovation and shoring up monopolies.
I have no problem with patenting an actual invention - like the dyson (and if you look at what happened with regard to his patents you would see that patents are next to useless for any small business when push comes to shove), but patenting software is just plain wrong.
like the dyson (and if you look at what happened with regard to his patents you would see that patents are next to useless for any small business when push comes to shove)
This sounded intriguing, so I googled. It looks to me like he got 4 million pounds off of Hoover in the end, and currently has more than 50% of the UK vacuum cleaner market...
Ah, yes - that was one of several patent cases - if you read his autobiography you'll see that his patents were no use in america and he had to settle out of court after several years in court and despite having a clear-cut case.
A fair chunk of the history of the dyson involves him being fucked over by large corporations and risking his home and business with huge debts to pay for the lawyers, in fact it's almost certain the first dyson cleaner would have arrived about 3 to 5 years earlier if it wasn't for a handful court cases in the US.
Now the problem here is that software is inherently unpatentable
Is this the "algorithms aren't patentable" argument? I've never really bought that, even when I was a paid-up frothing-at-the-mouth free-software zealot. You can't patent an algorithm but you can patent machinery implementing that algorithm? Software *is* machinery patenting that algorithm!
it has no benefits for society when it is patented
I'm not even sure that's true. Consider a) we have no control groups, and b) that software isn't radically different to any other kind of machinery, and we *do* have knowledge of a system without patent law (albeit one that's reasonably far-removed from our system). I think your assertion needs a little bit more argument.
Further to that problem is that software patents have particularly ill effect on free software
Right, so this *is* about an open-source company being sued. :)
patent licensing and royalties do not allow for products that are distributed...at zero cost
(I'm leaving out the part about zero-cost replication because that's a red herring.) Patent licensing does allow for products with zero distribution cost because there's no requirement to charge royalties to grant a licence. Perhaps you mean that companies who own patents don't often do this, which is a different thing.
Software patents already harm society as a whole by restricting innovation and shoring up monopolies.
Nope, don't agree with this (in fact I consider at least the first part self-evidently untrue). Given that I don't see software as different to other kinds of machinery I can't agree with your last paragraph either.
No, I'm all for sensible patent law, handily it's specified quite clearly in the EPC and UK patent law, and it worked (sometimes) for dyson who invented actual stuff and built working prototypes that I could replicate.
It doesn't work for software - there isn't a single software patent that enables me to replicate the working prototype - there isn't even pseudocode, let alone source code, not to mention the fact that most software patents published by Microsoft, et al to protect their monopolies are impossible to replicate as they are part of software protected by trade secrets, it's like being provided with a hand wavey description of a single piece of a jigsaw.
Software patents are proscribed in the EPC for good reason, there are a lot of very compelling reasons for software patents not to be allowed, FFII provide them if you can be bothered to actually think about the subject.
Also we already have a control group - the software industry before software patents were allowed, it allowed most of the major names around today to flourish, google became successful not because of a patent, but because they innovated.
I'd really like a patent system that worked, and for that to happen we need to raise the bar on the quality and rule out software, business rules, etc.
That's rich coming from somebody who hasn't contributed anything sensible to this thread.
I provided a control group, and a source of reference so I don't have to repeat the pages of reasons why software patents are a dumb idea, contrary to the patents system in europe and the uk, don't work as patents should, and have done provable harm to the industry.
When you look at the actual facts, software patents go beyond not working but doing exactly what they aren't supposed to - i.e. stopping R&D and Innovation.
FYI, you're talking to someone who is an experienced software developer working in a particularly IPR-fraught part of the industry. He's fairly likely to have done extensive thinking on the subject of patents.
Even without knowing that, your assumption that someone who is having a fairly considered discussion with you on a particular subject hasn't actually given it any thought is fairly rude, imho.
From his posts he was either a newbie or a trolling so I gave him the benefit of the doubt and assumed ignorance.
I find the fact he was condescending, patronising and accused me of being anti-patents rude. Not to mention failing to address any of the points I raised.
Expand on the self-evident untruth of the notion that software patents are hurting innovation? I'm sure I've seen more than one story about bizarrely wide-scope patents causing problems for people with new products, when they belatedly find out that their 'new' idea is infringing on an old patent.
From my things I've read trying to keep up with this debate, it would seem that software patents are workable in theory, but our patent system isn't capable of only granting the valid ones, and so in practise I'm inclined to think that software patents are a bad idea. I assume the problem lies in interpretation of the patent documents in a fast-moving industry with no tangible products, but I'm really not sure... we just seem to see a lot of totally insane patents granted in software, which are either non-original, trivial, extremely wide-ranging, or sometimes all of the above.
I'm sure I've seen more than one story about bizarrely wide-scope patents causing problems for people with new products, when they belatedly find out that their 'new' idea is infringing on an old patent.
That's more or less the crux of my argument - by the point the infringement is sued for, the innovation has already occurred. Regardless of whether the people being sued win (and get to market their product) or lose (and are snapped up and employed by e.g. IBM, whom no-one would be stupid enough to sue for patent infringements[1] (#footnote1), if your sole concern is to drive innovation, you can't consider that patents don't stifle it. If your concern is that innovation isn't being rewarded correctly, you possibly have a better point, but that wasn't the argument that was stated.
In fact, I go further and suggest that several companies are at the moment driving towards a broader and deeper patent portfolio to discourage and reduce the risk of vexatious patent suits. Since it's obvious that the best patent is a strong one - i.e. one that springs from innovative research - then even in the current climate, patents encourage innovation.
There are a couple of reasons why we see lots of frivolous patents. The first is that it's traditional to apply for a broader patent than you expect to get, and be haggled down. The patent office doesn't have much experience of software patents, and so haven't haggled down any of the recent ones proportionately. I think it's obvious that that's a temporary thing.
The other reason is the American emphasis on "checks and balances" - the patent office was never intended to be the arbiter of patent correctness - they simply provide a filter for the most obviously ridiculous ones. The judicial branch has both the responsibility and the experience to decide these questions, and so arbitration is correctly delegated to them.
The problem at the moment is that patent suits are too expensive to fight, which has never been the case before - originally, patents were granted to single inventors and didn't cost much to fight - some inventors represented themselves. More recently, patents have tended to be awarded to people working in deep-pockets companies doing deep-pockets research, so it's not a problem that they're required deep-pockets lawyers to fight. The imbalance now is caused by the fact that software development is simple enough to do on one's own, but profitable enough to be done by big companies. I don't have a simple answer to this, but I suspect that one of the nascent trends of small patent owner protection will bear fruit.
To an extent, but it's a percent of a percent: most innovation is carried out by full-time employees of companies, and of the stuff that isn't, most is rewarded correctly. I just don't think that, "I was going to make something that would set the world alight, but there was a small chance that I'd run into patent issues, so I decided not to" is a common reason for young entrepreneurs not starting software companies.
No but it does result in cancelled projects, whether commercial or non-profit or open source.
I think you're also jumping to a false conclusion wrt full-time employees of companies carrying out the most innovation - patents yes, innovation no.
Edison, Dyson et al weren't working for big companies when they had their great ideas, and SME's generate more GDP (at least in the EU) than large companies and have greater need to innovate, as they can't just bully and buy their way out of problems.
Google innovated because it was small and it wasn't employees but the creators who innovated - if there were patents in that arena (as there are now) then they almost certainly wouldn't have got the backing and would have been sued before they could have made a difference.
You're also wrong about innovation happening even if companies lose patent cases - one of the major trends in the US and elsewhere now is to stop the innovative product or service - see the recent blackberry case, had the case gone to trial and the judge agreed then the blackberry would never have hit the streets.
The patent system isn't working particularly well any more, and was never appropriate for software. Something else might be appropriate for software but patents in a recognisable form quite simply arent.
You can find examples of innovation being stopped in two ways - research (you can find it at the FFII website) has shown that software patents increase as r&d spending decreases which is the exact opposite of what patents are supposed to achieve, also there are plenty of businesses involved in the anti-sw-patent lobby who have had to essentially ignore potential new ideas and methods because they can't afford the patent litigation which meant cancelling projects, development and research, I recently read about a company that is wasting thousands of manhours manually debugging microprocessors because software patents restrict them from writing a better debugger for that chip and software.
Has anyone ever considered creating a legal defense fund for small software developers? They all contribute a bit of cash to the fund, and then it can be used to pay for legal defense if any of the contributers gets sued. At the very least, the existance of such a fund should discourage companies from frivolously suing in the hope of getting an out-of-court settlement, because it would make sure the developers could afford a court battle.
Think of it as insurance against pointless lawsuits.
![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
![[identity profile]](https://www.dreamwidth.org/img/silk/identity/openid.png){kind=small}
(no subject)
Date: 2006-07-13 09:21 am (UTC)Who am I kidding? That would be sensible, and so of course will never happen.
(no subject)
Date: 2006-07-13 10:29 am (UTC)The first example he cites is someone suing RedHat - big deal. RH are a commercial company, and they're big and ugly enough to fend for themselves - the fact that the alleged patent infringement happens to be on an open source technology (which has always been controlled by a for-profit company, let's not forget) is neither here nor there.
The other one is a shame, and if it goes through as Perens indicates he thinks it will, that'll be a clear problem. That said, the patent itself, unless I'm reading it wrongly, looks very obvious and non-novel - if the guy can put together a defence, he'll probably be able to defeat it, and with any luck the company will be told to pay his costs.
As usual, I trust in the free market to do what's best - if it turns out that open source authors need legal protection, either their customers or companies will come up with the money to do so.
(I'm not anti-patents, btw.)
(no subject)
Date: 2006-07-13 12:13 pm (UTC)See, that's the problem. It takes superlative ill-faith for this to happen in the US, and a granted patent is presumed valid until proven otherwise.
software isn't patentable in any sensible patent law
Date: 2006-07-14 03:25 pm (UTC)Further to that problem is that software patents have particularly ill effect on free software as patent licensing and royalties do not allow for products that are distributed and replicated at zero cost like software, and especially where anybody is free to distribute and modify the software.
So we have 2 problems, software patents are bad, software patents exist nonetheless and are harming open source software and therefore the whole industry and society as a whole. Software patents already harm society as a whole by restricting innovation and shoring up monopolies.
I have no problem with patenting an actual invention - like the dyson (and if you look at what happened with regard to his patents you would see that patents are next to useless for any small business when push comes to shove), but patenting software is just plain wrong.
Re: software isn't patentable in any sensible patent law
Date: 2006-07-14 03:34 pm (UTC)This sounded intriguing, so I googled. It looks to me like he got 4 million pounds off of Hoover in the end, and currently has more than 50% of the UK vacuum cleaner market...
Re: software isn't patentable in any sensible patent law
Date: 2006-07-14 03:44 pm (UTC)A fair chunk of the history of the dyson involves him being fucked over by large corporations and risking his home and business with huge debts to pay for the lawyers, in fact it's almost certain the first dyson cleaner would have arrived about 3 to 5 years earlier if it wasn't for a handful court cases in the US.
And the definition of "sensible patent law" is that which doesn't allow patents, right?
Date: 2006-07-14 07:11 pm (UTC)Is this the "algorithms aren't patentable" argument? I've never really bought that, even when I was a paid-up frothing-at-the-mouth free-software zealot. You can't patent an algorithm but you can patent machinery implementing that algorithm? Software *is* machinery patenting that algorithm!
it has no benefits for society when it is patented
I'm not even sure that's true. Consider a) we have no control groups, and b) that software isn't radically different to any other kind of machinery, and we *do* have knowledge of a system without patent law (albeit one that's reasonably far-removed from our system). I think your assertion needs a little bit more argument.
Further to that problem is that software patents have particularly ill effect on free software
Right, so this *is* about an open-source company being sued. :)
patent licensing and royalties do not allow for products that are distributed...at zero cost
(I'm leaving out the part about zero-cost replication because that's a red herring.) Patent licensing does allow for products with zero distribution cost because there's no requirement to charge royalties to grant a licence. Perhaps you mean that companies who own patents don't often do this, which is a different thing.
Software patents already harm society as a whole by restricting innovation and shoring up monopolies.
Nope, don't agree with this (in fact I consider at least the first part self-evidently untrue). Given that I don't see software as different to other kinds of machinery I can't agree with your last paragraph either.
Re: And the definition of "sensible patent law" is that which doesn't allow patents, right?
Date: 2006-07-14 08:38 pm (UTC)It doesn't work for software - there isn't a single software patent that enables me to replicate the working prototype - there isn't even pseudocode, let alone source code, not to mention the fact that most software patents published by Microsoft, et al to protect their monopolies are impossible to replicate as they are part of software protected by trade secrets, it's like being provided with a hand wavey description of a single piece of a jigsaw.
Software patents are proscribed in the EPC for good reason, there are a lot of very compelling reasons for software patents not to be allowed, FFII provide them if you can be bothered to actually think about the subject.
Also we already have a control group - the software industry before software patents were allowed, it allowed most of the major names around today to flourish, google became successful not because of a patent, but because they innovated.
I'd really like a patent system that worked, and for that to happen we need to raise the bar on the quality and rule out software, business rules, etc.
Re: And the definition of "sensible patent law" is that which doesn't allow patents, right?
Date: 2006-07-15 03:16 pm (UTC)Funny, when you say that I lose all interest in talking to you. Thanks for playing. :)
Re: And the definition of "sensible patent law" is that which doesn't allow patents, right?
Date: 2006-07-15 03:32 pm (UTC)I provided a control group, and a source of reference so I don't have to repeat the pages of reasons why software patents are a dumb idea, contrary to the patents system in europe and the uk, don't work as patents should, and have done provable harm to the industry.
When you look at the actual facts, software patents go beyond not working but doing exactly what they aren't supposed to - i.e. stopping R&D and Innovation.
Re: And the definition of "sensible patent law" is that which doesn't allow patents, right?
Date: 2006-07-15 03:49 pm (UTC)Even without knowing that, your assumption that someone who is having a fairly considered discussion with you on a particular subject hasn't actually given it any thought is fairly rude, imho.
Re: And the definition of "sensible patent law" is that which doesn't allow patents, right?
Date: 2006-07-15 05:59 pm (UTC)From his posts he was either a newbie or a trolling so I gave him the benefit of the doubt and assumed ignorance.
I find the fact he was condescending, patronising and accused me of being anti-patents rude. Not to mention failing to address any of the points I raised.
Re: And the definition of "sensible patent law" is that which doesn't allow patents, right?
Date: 2006-07-15 03:55 pm (UTC)From my things I've read trying to keep up with this debate, it would seem that software patents are workable in theory, but our patent system isn't capable of only granting the valid ones, and so in practise I'm inclined to think that software patents are a bad idea. I assume the problem lies in interpretation of the patent documents in a fast-moving industry with no tangible products, but I'm really not sure... we just seem to see a lot of totally insane patents granted in software, which are either non-original, trivial, extremely wide-ranging, or sometimes all of the above.
Re: And the definition of "sensible patent law" is that which doesn't allow patents, right?
Date: 2006-07-15 04:33 pm (UTC)That's more or less the crux of my argument - by the point the infringement is sued for, the innovation has already occurred. Regardless of whether the people being sued win (and get to market their product) or lose (and are snapped up and employed by e.g. IBM, whom no-one would be stupid enough to sue for patent infringements[1] (#footnote1), if your sole concern is to drive innovation, you can't consider that patents don't stifle it. If your concern is that innovation isn't being rewarded correctly, you possibly have a better point, but that wasn't the argument that was stated.
In fact, I go further and suggest that several companies are at the moment driving towards a broader and deeper patent portfolio to discourage and reduce the risk of vexatious patent suits. Since it's obvious that the best patent is a strong one - i.e. one that springs from innovative research - then even in the current climate, patents encourage innovation.
There are a couple of reasons why we see lots of frivolous patents. The first is that it's traditional to apply for a broader patent than you expect to get, and be haggled down. The patent office doesn't have much experience of software patents, and so haven't haggled down any of the recent ones proportionately. I think it's obvious that that's a temporary thing.
The other reason is the American emphasis on "checks and balances" - the patent office was never intended to be the arbiter of patent correctness - they simply provide a filter for the most obviously ridiculous ones. The judicial branch has both the responsibility and the experience to decide these questions, and so arbitration is correctly delegated to them.
The problem at the moment is that patent suits are too expensive to fight, which has never been the case before - originally, patents were granted to single inventors and didn't cost much to fight - some inventors represented themselves. More recently, patents have tended to be awarded to people working in deep-pockets companies doing deep-pockets research, so it's not a problem that they're required deep-pockets lawyers to fight. The imbalance now is caused by the fact that software development is simple enough to do on one's own, but profitable enough to be done by big companies. I don't have a simple answer to this, but I suspect that one of the nascent trends of small patent owner protection will bear fruit.
1. ;) ()
Re: And the definition of "sensible patent law" is that which doesn't allow patents, right?
Date: 2006-07-15 04:46 pm (UTC)You don't think that a lack of proper reward for innovation will tend to lead to less people trying to innovate?
Re: And the definition of "sensible patent law" is that which doesn't allow patents, right?
Date: 2006-07-15 05:07 pm (UTC)Re: And the definition of "sensible patent law" is that which doesn't allow patents, right?
Date: 2006-07-15 06:09 pm (UTC)I think you're also jumping to a false conclusion wrt full-time employees of companies carrying out the most innovation - patents yes, innovation no.
Edison, Dyson et al weren't working for big companies when they had their great ideas, and SME's generate more GDP (at least in the EU) than large companies and have greater need to innovate, as they can't just bully and buy their way out of problems.
Google innovated because it was small and it wasn't employees but the creators who innovated - if there were patents in that arena (as there are now) then they almost certainly wouldn't have got the backing and would have been sued before they could have made a difference.
You're also wrong about innovation happening even if companies lose patent cases - one of the major trends in the US and elsewhere now is to stop the innovative product or service - see the recent blackberry case, had the case gone to trial and the judge agreed then the blackberry would never have hit the streets.
The patent system isn't working particularly well any more, and was never appropriate for software. Something else might be appropriate for software but patents in a recognisable form quite simply arent.
Re: And the definition of "sensible patent law" is that which doesn't allow patents, right?
Date: 2006-07-15 05:57 pm (UTC)(no subject)
Date: 2006-07-13 05:32 pm (UTC)Think of it as insurance against pointless lawsuits.