European Parliament rejects software patent
July 6, 2005 6:47 AM Subscribe
European Parliament rejects software patenting..and by a large politically crushing majority of 648 votes vs 14. This is a great measurable success for organization like Foundation for a Free Information Infrastructure whose members, supporters and sympathizers have fought against the insane idea of software patent for more then four years (more info on euro software patents on the organization site).
So let me understand this. You are a small time software designer and you make a new application, how exactly are you supposed to make money on your work? How are you supposed to be able to prevent some mega-corporation from making your money on your work? I think I'm missing something here.
posted by Pollomacho at 7:21 AM on July 6, 2005
posted by Pollomacho at 7:21 AM on July 6, 2005
It's worth mentioning that this is the first time that the EU parliament has rejected a common position of the 25 member states. Considering that this bill also had the support of heavywights like Microsoft, Nokia, Siemens, and others, it's quite surprising that it was so resoundingly rejected.
Pollomacho: It just means that there won't be a common EU-wide policy. Individual countries will still maintain their own patent system and laws.
posted by Ljubljana at 7:27 AM on July 6, 2005
Pollomacho: It just means that there won't be a common EU-wide policy. Individual countries will still maintain their own patent system and laws.
posted by Ljubljana at 7:27 AM on July 6, 2005
Patenting software is like patenting a recipe.
There are very few things in software that are truly, genuinely unique. Most of the success of a company comes from good customer service, excellent usability and establishing a niche. Software patents are mostly used to keep out small, unestablished companies / developers who don't have the cash to pay out a settlement when they inadvertantly step in someone's "IP".
posted by bshort at 7:29 AM on July 6, 2005
There are very few things in software that are truly, genuinely unique. Most of the success of a company comes from good customer service, excellent usability and establishing a niche. Software patents are mostly used to keep out small, unestablished companies / developers who don't have the cash to pay out a settlement when they inadvertantly step in someone's "IP".
posted by bshort at 7:29 AM on July 6, 2005
Pollomacho,
This site seems to have a good overview:
Copyright protects authors but doesn't hurt any honest person. Patents, in contrast, are 20-year monopolies that the government grants on broad and general ideas. Patents are potential weapons against all of us.
posted by jikel_morten at 7:32 AM on July 6, 2005
This site seems to have a good overview:
Copyright protects authors but doesn't hurt any honest person. Patents, in contrast, are 20-year monopolies that the government grants on broad and general ideas. Patents are potential weapons against all of us.
posted by jikel_morten at 7:32 AM on July 6, 2005
I have not read the article, but on the surface, my initial thoughts are the same as Pollomacho: how do I protect something that I created? Can I protect something that I created? What is there to stop someone from making a copy of, say, iTunes, re-badging it and selling it as something else?
posted by tgrundke at 7:32 AM on July 6, 2005
posted by tgrundke at 7:32 AM on July 6, 2005
The above site says that you automatically own the copyright upon writing code.
posted by jikel_morten at 7:35 AM on July 6, 2005
posted by jikel_morten at 7:35 AM on July 6, 2005
What is there to stop someone from making a copy of, say, iTunes, re-badging it and selling it as something else?
That hasn't got much to do with software patents. The fact is there are iTunes type clones from Real, Napster etc.
posted by PenDevil at 7:38 AM on July 6, 2005
That hasn't got much to do with software patents. The fact is there are iTunes type clones from Real, Napster etc.
posted by PenDevil at 7:38 AM on July 6, 2005
pollomacho: individual small companies, the small guy with an idea..hasn't got the money to patent anything to begin with (patent lawyers are deadly expensive) and if he goes asking for financing he basically leaves all the rights to whoever is financing...sometimes the patent lawyers claim he did the patenting work but he just stole the idea, giving it to some associate..such is the misery of these parasites.
In other words, the real creators (unless they're big money and resources rich companies) don't see but a fraction of a dime..if they get to see some, it's just to distract them from the enormous log they're being beaten with.
Additionally in a best-case scenario software patenting is likely to create primarily (if not exclusively) enormous amount of useless litigations..which would make a lot of lawyers happy, but the rest of the world would suffer the cost of litigations that wouldn't register as frivolous in the world of law but would be frivolous in the world of tangible results.
You can also see a potential effect of software patenting here.
posted by elpapacito at 7:38 AM on July 6, 2005
In other words, the real creators (unless they're big money and resources rich companies) don't see but a fraction of a dime..if they get to see some, it's just to distract them from the enormous log they're being beaten with.
Additionally in a best-case scenario software patenting is likely to create primarily (if not exclusively) enormous amount of useless litigations..which would make a lot of lawyers happy, but the rest of the world would suffer the cost of litigations that wouldn't register as frivolous in the world of law but would be frivolous in the world of tangible results.
You can also see a potential effect of software patenting here.
posted by elpapacito at 7:38 AM on July 6, 2005
Pollomacho writes "I think I'm missing something here."
Yes, Pollomacho, you're missing the fact that copyright laws offer you enough protection without any need for your software "invention" to be patentable - the software industry and many other "pure ideas" industries survived and thrived without patents for centuries. This is a great step towards bringing sanity back to a system that has become more and more stupid along ther years.
A transcript/report of the EP debate is available here .
On preview, tgrundke, you seem to be confusing the lack of patents with the lack of protection. It is not like Europe Parliament has decided everything is now free software. Try to do that with iTunes and Apple will throw the copyright laws in your head so hard that you will actually need the jail time to recover.
posted by nkyad at 7:39 AM on July 6, 2005
Yes, Pollomacho, you're missing the fact that copyright laws offer you enough protection without any need for your software "invention" to be patentable - the software industry and many other "pure ideas" industries survived and thrived without patents for centuries. This is a great step towards bringing sanity back to a system that has become more and more stupid along ther years.
A transcript/report of the EP debate is available here .
On preview, tgrundke, you seem to be confusing the lack of patents with the lack of protection. It is not like Europe Parliament has decided everything is now free software. Try to do that with iTunes and Apple will throw the copyright laws in your head so hard that you will actually need the jail time to recover.
posted by nkyad at 7:39 AM on July 6, 2005
The war is not yet over, but this is a major battle that's won. *happydances*
Pollomacho: the problem with software patents is people who register ideas that are so generic, they stifle any further development in those areas. A company may not even be interested in developing a product, merely enforcing the patent and making money by suing for damages. To further complicate matters, patents can be kept secret for a number of years, so if you independently develop software based on a patented idea, you could be in trouble and not even know it.
Software is already protected by existing copyright laws, which, despite their many flaws, are fair enough. Software patents are just tyranny.
posted by Goblindegook at 8:02 AM on July 6, 2005
Pollomacho: the problem with software patents is people who register ideas that are so generic, they stifle any further development in those areas. A company may not even be interested in developing a product, merely enforcing the patent and making money by suing for damages. To further complicate matters, patents can be kept secret for a number of years, so if you independently develop software based on a patented idea, you could be in trouble and not even know it.
Software is already protected by existing copyright laws, which, despite their many flaws, are fair enough. Software patents are just tyranny.
posted by Goblindegook at 8:02 AM on July 6, 2005
Software patents are not just a theoretical bogeyman, they already have an impressive track record of abuse.
One of the more infamous cases of patent abuse was the case of Amazon's "One-Click" patent. In this case, the patent wasn't protecting a novel idea, it was used to hurt competitors by preventing them from using a very obvious idea. To his credit, Amazon CEO Bezos has admitted that this patent is rediculous and that the patent system should be changed to prevent exactly this sort of abuse. But he felt it was better to be the abuser than the abused.
Another infamous abuser is a company called Forgent, which has made suing over insane patent claims into a business model. These guys aren't "protecting" anything, they're digging for popular bits of intellectual property they can lay some tenuous legal claim to, not in the name of protecting their business, but in the name of lawsuit profiteering alone.
Patents were invented to promote invention and innovation. Today, especially in the software field, they are increasingly used as a weapon, weilded by large, rich corporations, used to prevent innovation.
posted by Western Infidels at 8:10 AM on July 6, 2005
One of the more infamous cases of patent abuse was the case of Amazon's "One-Click" patent. In this case, the patent wasn't protecting a novel idea, it was used to hurt competitors by preventing them from using a very obvious idea. To his credit, Amazon CEO Bezos has admitted that this patent is rediculous and that the patent system should be changed to prevent exactly this sort of abuse. But he felt it was better to be the abuser than the abused.
Another infamous abuser is a company called Forgent, which has made suing over insane patent claims into a business model. These guys aren't "protecting" anything, they're digging for popular bits of intellectual property they can lay some tenuous legal claim to, not in the name of protecting their business, but in the name of lawsuit profiteering alone.
Patents were invented to promote invention and innovation. Today, especially in the software field, they are increasingly used as a weapon, weilded by large, rich corporations, used to prevent innovation.
posted by Western Infidels at 8:10 AM on July 6, 2005
Patents enable mafia-like tactics in the tech industry.
A friend of mine tells the story where IBM walked into Sun's offices with about 10 patents and asked for a million dollars. The engineers there said "Hey, we don't infringe any of these 10 patents!"
IBM replied "Yeah, but we've got 200,000 patents, and I'm sure that if we show up here with them and go through each one thoroughly, we'll find some that you DO infringe. Let's make it easy on all of us and cut us a check."
Apparently, there was no more discussion. They got their money.
posted by Kickstart70 at 8:28 AM on July 6, 2005
A friend of mine tells the story where IBM walked into Sun's offices with about 10 patents and asked for a million dollars. The engineers there said "Hey, we don't infringe any of these 10 patents!"
IBM replied "Yeah, but we've got 200,000 patents, and I'm sure that if we show up here with them and go through each one thoroughly, we'll find some that you DO infringe. Let's make it easy on all of us and cut us a check."
Apparently, there was no more discussion. They got their money.
posted by Kickstart70 at 8:28 AM on July 6, 2005
Goblinedegook: the problem with software patents is people who register ideas that are so generic, they stifle any further development in those areas.
Therefore, the patent office is the flaw, not the patents themselves, intrinsically. If the patent office was doing its job and following the rules set out for it, this wouldn't be an issue. A full investigation into corruption there would probably find some -amazing- things. I wonder how fancy the houses of patent office staff are.
posted by Kickstart70 at 8:31 AM on July 6, 2005
Therefore, the patent office is the flaw, not the patents themselves, intrinsically. If the patent office was doing its job and following the rules set out for it, this wouldn't be an issue. A full investigation into corruption there would probably find some -amazing- things. I wonder how fancy the houses of patent office staff are.
posted by Kickstart70 at 8:31 AM on July 6, 2005
Kickstart70: I wonder how fancy the houses of patent office staff are.
I think (well, I'm guessing) that it's more a matter of incompetence than malice or corruption.
The patent office supposedly has a policy against issuing patents for perpetual-motion machines, for example; yet there are hundreds of such patents, and a few more every year. These patents are granted to lone crackpot garage tinkerers, not to powerful or well-funded interests. Dressed up in dense, hard-to-read wording and layers of obfuscation, a patent on a perpetual-motion device looks like any of a zillion other mechanical patents of little importance.
I think the patent office is woefully overburdened with work and is attempting to do a nearly impossible task besides. To be effective, the patent office would need technical expertise in basically everything, and literally no one has that kind of resources.
posted by Western Infidels at 9:33 AM on July 6, 2005
I think (well, I'm guessing) that it's more a matter of incompetence than malice or corruption.
The patent office supposedly has a policy against issuing patents for perpetual-motion machines, for example; yet there are hundreds of such patents, and a few more every year. These patents are granted to lone crackpot garage tinkerers, not to powerful or well-funded interests. Dressed up in dense, hard-to-read wording and layers of obfuscation, a patent on a perpetual-motion device looks like any of a zillion other mechanical patents of little importance.
I think the patent office is woefully overburdened with work and is attempting to do a nearly impossible task besides. To be effective, the patent office would need technical expertise in basically everything, and literally no one has that kind of resources.
posted by Western Infidels at 9:33 AM on July 6, 2005
Why are software patents such a stupid idea ? Patenting software is basically like a land grab - those with the biggest army of lawyers rush in and try to grab as much territory as possible. This has resulted in extraordinary patents being awarded such as the infamous Amazon "one click" patent (they own - and vigourously defend - the idea that you can buy something on the web by just clicking a button). If you want to know why the majority of online shopping websites are so awkward, now you know who to blame. Here's a quick list of a few more (found by just going to the top google link - there are thousand more complete lists if you want to look). Software patents rarely have anything to do with "inventions" and much more to do with legal firepower. And it's not *just* companies like Microsoft that have been pushing the EU to adopt these ridiculous patents, the US government has been putting on some pretty serious pressure too. The US wants the EU to copy their (almost universally derided) patent system so that the huge (mainly US) companies who own these patents can enforce them worldwide and extend their various monopolies. They have already succeeded in using their weight to getting many smaller countries to comply, so how is a little software developer in, say, Jordan, going to take on the might of Microsoft ? Microsoft already own patents on absolutely insane things like the "not equals" programming operation, which are so wide ranging that (if enforced) they could be used against almost any piece of software - a small developer would never have the ability to fight them in court.
This decision by the EU is astonishingly good news. These days it almost seems unbelievable when the voice of reason wins.
posted by silence at 9:37 AM on July 6, 2005
This decision by the EU is astonishingly good news. These days it almost seems unbelievable when the voice of reason wins.
posted by silence at 9:37 AM on July 6, 2005
OK, I think I've got a better understanding of the issue here, so now here's the question that this raises next:
This law was to unify software patents over the entire EU, yes? So now if you were a small company who wanted to lay claim to your software through a patent, you'd have to apply in each and every nation of the EU, right? Doesn't this seem to make it worse for the small guy who doesn't have the money, connections or lawyers in every one of those nations?
Again, I'm asking here, you gave me great explanations for my first questions! Thanks!
posted by Pollomacho at 10:57 AM on July 6, 2005
This law was to unify software patents over the entire EU, yes? So now if you were a small company who wanted to lay claim to your software through a patent, you'd have to apply in each and every nation of the EU, right? Doesn't this seem to make it worse for the small guy who doesn't have the money, connections or lawyers in every one of those nations?
Again, I'm asking here, you gave me great explanations for my first questions! Thanks!
posted by Pollomacho at 10:57 AM on July 6, 2005
damn. I knew it was too good to be true. I just read some more about the voting and apparently BOTH sides of the argument (pro and anti patent) voted it down. The Pro-patent lobby voted against it because there were proposed amendments which if voted into law would have watered down what they wanted. So, the end result is that we're back to square one. On the one hand we don't have an EU bill which mandates insane patent laws, but on the other hand we don't have one which expressly forbids them, so it's up to each country to set their own guidelines. It seems likely that the pro-patent lobbying will continue on a per-country level and will probably return to have another crack at an EU wide directive at some point. The only real defence would be to have a proper EU ruling against software patents - which at this point means starting the whole process again.
posted by silence at 11:03 AM on July 6, 2005
posted by silence at 11:03 AM on July 6, 2005
Pollomacho, you're right that now the situation continues to be confusing, but at least it means that some countries within the EU can reject software patents. It's not just a question of the "big guy" versus the "little guy" (though that's an important part of the issue).
The problem is that software patents are inherently a stupid idea. Software is a bit like writing. Software is already covered by copyright just like writing - if I copy and paste a load of your code and you don't want me to, then you can assert your copyright to that code. But patenting the structures and ideas behind the code is mad. Imagine if the EU was debating "literary patents" or "cinema patents" or even "cookery patants" and large corporations were patenting the idea of the "paragraph", or claimed rights to the idea of the "flawed hero", or forbid anyone from making a bechamel based sauce - you'd be outraged. That exactly what's happening here.
posted by silence at 11:18 AM on July 6, 2005
The problem is that software patents are inherently a stupid idea. Software is a bit like writing. Software is already covered by copyright just like writing - if I copy and paste a load of your code and you don't want me to, then you can assert your copyright to that code. But patenting the structures and ideas behind the code is mad. Imagine if the EU was debating "literary patents" or "cinema patents" or even "cookery patants" and large corporations were patenting the idea of the "paragraph", or claimed rights to the idea of the "flawed hero", or forbid anyone from making a bechamel based sauce - you'd be outraged. That exactly what's happening here.
posted by silence at 11:18 AM on July 6, 2005
Well, I'll play Devil's Advocate and argue that this is extraordinarily bad news. First of all, the EP hasn't "rejected software patents". It has rejected a directive that would have delimited which patents involving computers would have been allowable and which ones would not. This, considering the current legal chaos in Europe on that matter would have been, by itself, a good thing. After all, in a situation without legal certainty and plenty of grey areas, the companies with the best legal teams are more likely to get it their way. Which is why the business software lobby, for instance, preferred no directive at all that a directive that would set clear boundaries.
Now, the charge is that the directive proposal did little to increase that legal certainty. That is not an entirely groundless charge. However, the amendment proposals of the (not at all software-patent-friendly) French Socialist MEP, Rocard, would have gone far (some of them too far, IMHO) towards tackling those problems. By rejecting the whole directive, the EP made so that the amendments weren't even considered. This may explain why the conservatives voted against the directive...
The trouble is, in this debate there has been a great deal of ignorance and disingenousness. I notice that many of those that give an opinion know very little about patents in general and the directive in particular.
OK, a few words about patents. What differentiates patents from copyright? Well, what a patent protects is the basic concept behind a new creation, not just its external form, like the copyright. So, in the software field, a patent would protect against somebody using the same algorithm, not just the same code. In exchange, the validity of patents is more short-term (only twenty years), and patents should only be granted to ideas that are not just new, but also inventive and (about everywhere but in the US) technical. Patents have basically two parts:
a) The description, which discloses the invention in a way that a skilled person can reproduce it.
b) The claims, which define exactly what is protected by the patent. There are "device" or "product" claims and "process" or "method" claims. That way one patent can protect, say, a new chemical compound, as well as the process to produce it and the device realising that process. (And bshort, "recipes" are certainly patentable: the chemical industry is constantly patenting recipes for new compounds). Patent attorneys don't like process claims much, though, because they can only prove infringement if they catch the infringer actually carrying out that process.
Now for a surprise: software is patentable. At least if you define it as an algorithm or a series of instructions governing a process. It must be a technical process, of course (more to it later). What is not patentable, according to most European legislation, are "computer programs". This means you can get a patent with a claim concerning a process to be executed by a computer (so, an algorithm), but not for the computer program as a device. Which means the infringement takes place only when the program is actually executed, but that a support carrying the program does not infringe the patent.
This is a big problem for the industry, and I don't mean by any means only the software industry. The "computer program" exclusion was written into law in the early '70s, when computers were still big unwieldy devices and nobody could dream that we would have programmable computers at the core of almost every device. When engines were controlled with carburettors, you could get a patent covering a carburettor and the control process it executed. When we came to injection engines controlled by purpose-built electronic circuits, it was still possible to protect the process and the circuit executing it. But now that we have engines controlled by programmable, generic CPUs, the device actually behind the process, the computer program, can no longer be patented.
The obvious solution, attempted by the directive, is to allow claims for computer programs, as long as they have a technical character. Thing is, it is damn difficult to define technical character. A program for controlling a car engine, a machine tool or say a camera focussing mechanism has technical character. A program for comparing investments clearly not. But would a program for, say, enhancing the image in that same camera, or for controlling wireless networks have a technical character or not. This is what the serious debate has been all about. To further muddy the waters, a few years ago the US Supreme Court in a rather surprising decision almost entirely did away with the technical requirement by declaring patentable "everything under the sun", thus opening the way to Amazon's much-mentioned "One-Click" patent and more.
Finally, to the issue of patent offices (notice the plural: there is not "one" patent office, but one per country, and a European Patent Office too) and trivial patents. As said, the subject of a patent should be inventive. While it is quite easy to say whether something is new or not, to determine that something is "not inventive" or "trivial", particularly in the face of a determined attorney, is less than straightforward. And as patent examiners are human, and therefore fallible, they may also blunder: hundreds of thousands of patents are applied for every year, it is unsurprising if a few howlers get through. At least in Europe the process of opposing a patent grant is far easier than in the US. (And regarding those "perpetuum mobile" patents: do they harm anybody? No. A patent only serves to prevent others from copying your idea. If your idea is stupid, who will want to!? Few patent examiners will waste much ink trying to convince somebody that his invention does not work...)
posted by Skeptic at 11:54 AM on July 6, 2005
Now, the charge is that the directive proposal did little to increase that legal certainty. That is not an entirely groundless charge. However, the amendment proposals of the (not at all software-patent-friendly) French Socialist MEP, Rocard, would have gone far (some of them too far, IMHO) towards tackling those problems. By rejecting the whole directive, the EP made so that the amendments weren't even considered. This may explain why the conservatives voted against the directive...
The trouble is, in this debate there has been a great deal of ignorance and disingenousness. I notice that many of those that give an opinion know very little about patents in general and the directive in particular.
OK, a few words about patents. What differentiates patents from copyright? Well, what a patent protects is the basic concept behind a new creation, not just its external form, like the copyright. So, in the software field, a patent would protect against somebody using the same algorithm, not just the same code. In exchange, the validity of patents is more short-term (only twenty years), and patents should only be granted to ideas that are not just new, but also inventive and (about everywhere but in the US) technical. Patents have basically two parts:
a) The description, which discloses the invention in a way that a skilled person can reproduce it.
b) The claims, which define exactly what is protected by the patent. There are "device" or "product" claims and "process" or "method" claims. That way one patent can protect, say, a new chemical compound, as well as the process to produce it and the device realising that process. (And bshort, "recipes" are certainly patentable: the chemical industry is constantly patenting recipes for new compounds). Patent attorneys don't like process claims much, though, because they can only prove infringement if they catch the infringer actually carrying out that process.
Now for a surprise: software is patentable. At least if you define it as an algorithm or a series of instructions governing a process. It must be a technical process, of course (more to it later). What is not patentable, according to most European legislation, are "computer programs". This means you can get a patent with a claim concerning a process to be executed by a computer (so, an algorithm), but not for the computer program as a device. Which means the infringement takes place only when the program is actually executed, but that a support carrying the program does not infringe the patent.
This is a big problem for the industry, and I don't mean by any means only the software industry. The "computer program" exclusion was written into law in the early '70s, when computers were still big unwieldy devices and nobody could dream that we would have programmable computers at the core of almost every device. When engines were controlled with carburettors, you could get a patent covering a carburettor and the control process it executed. When we came to injection engines controlled by purpose-built electronic circuits, it was still possible to protect the process and the circuit executing it. But now that we have engines controlled by programmable, generic CPUs, the device actually behind the process, the computer program, can no longer be patented.
The obvious solution, attempted by the directive, is to allow claims for computer programs, as long as they have a technical character. Thing is, it is damn difficult to define technical character. A program for controlling a car engine, a machine tool or say a camera focussing mechanism has technical character. A program for comparing investments clearly not. But would a program for, say, enhancing the image in that same camera, or for controlling wireless networks have a technical character or not. This is what the serious debate has been all about. To further muddy the waters, a few years ago the US Supreme Court in a rather surprising decision almost entirely did away with the technical requirement by declaring patentable "everything under the sun", thus opening the way to Amazon's much-mentioned "One-Click" patent and more.
Finally, to the issue of patent offices (notice the plural: there is not "one" patent office, but one per country, and a European Patent Office too) and trivial patents. As said, the subject of a patent should be inventive. While it is quite easy to say whether something is new or not, to determine that something is "not inventive" or "trivial", particularly in the face of a determined attorney, is less than straightforward. And as patent examiners are human, and therefore fallible, they may also blunder: hundreds of thousands of patents are applied for every year, it is unsurprising if a few howlers get through. At least in Europe the process of opposing a patent grant is far easier than in the US. (And regarding those "perpetuum mobile" patents: do they harm anybody? No. A patent only serves to prevent others from copying your idea. If your idea is stupid, who will want to!? Few patent examiners will waste much ink trying to convince somebody that his invention does not work...)
posted by Skeptic at 11:54 AM on July 6, 2005
I have been told by lawyers that software copyright only covers the source code, so if someone copies your program by writing it themselves (even if they translate the code literally into another programming language) then the copyright is no longer applicable.
So if I invent a novel software concept, then a huge corporation with lots of resources can copy the general concept (but not necessarily the implementation) and out-market me. I hope someone can refute this.
posted by krunk at 12:04 PM on July 6, 2005
So if I invent a novel software concept, then a huge corporation with lots of resources can copy the general concept (but not necessarily the implementation) and out-market me. I hope someone can refute this.
posted by krunk at 12:04 PM on July 6, 2005
I don't know if this analogy would be helpful, or more confounding.
M. Night Shaymalan, applies for and receives patents for:
1. Any movie plot in which ghosts and non-ghosts communicate, and
2. Any movie ending in which a final development recasts what happened previously.
In both cases, what is at stake is a method or strategy. Both might be widely used, but the sheer number of patents applied for makes it practically impossible for a full "prior art" investigation to be done.
The patent office has long had a policy of making mathematical formulae unpatentable. Most software patents are little algorithmic tweaks or approaches to creating software--they are more properly considered part of software craft. But because technology is arcane, and not understandable by all, the patent office is granting monopoly rights to simple building blocks of programs.
If this approach were applied to more accessible crafts--like woodworking--it would be easier to see how ridiculuous and counterproductive software patents are. Imagine:
6191928: Sanding a smooth surface using a circular motion.
5158682: Using four legs, one one each corner, to hold the surface of a table above ground level.
What software patents do promise is to make it virtually impossible to write an non-infringing computer program. Widely-employed programming approaches and techniques have been patented, decades after they became common practice. Checking a program for use of infringing techniques would take 100 more man-years than writing the program would have taken in the first place.
Software patents are just a bad idea. Beyond that, it is virtually impossible to make a reasonable determination of prior art, or of derivative intellectual property. As enforcement of software patents in the U.S. increases, it will lead to the downfall of the software industry.
posted by curtm at 12:13 PM on July 6, 2005
M. Night Shaymalan, applies for and receives patents for:
1. Any movie plot in which ghosts and non-ghosts communicate, and
2. Any movie ending in which a final development recasts what happened previously.
In both cases, what is at stake is a method or strategy. Both might be widely used, but the sheer number of patents applied for makes it practically impossible for a full "prior art" investigation to be done.
The patent office has long had a policy of making mathematical formulae unpatentable. Most software patents are little algorithmic tweaks or approaches to creating software--they are more properly considered part of software craft. But because technology is arcane, and not understandable by all, the patent office is granting monopoly rights to simple building blocks of programs.
If this approach were applied to more accessible crafts--like woodworking--it would be easier to see how ridiculuous and counterproductive software patents are. Imagine:
6191928: Sanding a smooth surface using a circular motion.
5158682: Using four legs, one one each corner, to hold the surface of a table above ground level.
What software patents do promise is to make it virtually impossible to write an non-infringing computer program. Widely-employed programming approaches and techniques have been patented, decades after they became common practice. Checking a program for use of infringing techniques would take 100 more man-years than writing the program would have taken in the first place.
Software patents are just a bad idea. Beyond that, it is virtually impossible to make a reasonable determination of prior art, or of derivative intellectual property. As enforcement of software patents in the U.S. increases, it will lead to the downfall of the software industry.
posted by curtm at 12:13 PM on July 6, 2005
curtm Re. Movieplots: Yes, a method or strategy, but without any technical character whatsoever. Moreover, the idea that the sheer number of patents applied for makes it practically impossible for a full "prior art" investigation to be done
is rather wrong. In fact, patent offices are very well equipped to search prior art (well, some more than others). Especially previous patents. Many people reckon that the main problem the US Patent and Trademark Office has had after the US Supreme Court opened the floodgates is that it lacked both the archives and the expertise to search software prior art. But I hear it's getting better at it, as it should be expected.
As enforcement of software patents in the U.S. increases, it will lead to the downfall of the software industry.
Thing is, despite some real excesses, the US software industry is still doing rather better than the European software industry, funnily enough...
posted by Skeptic at 12:43 PM on July 6, 2005
is rather wrong. In fact, patent offices are very well equipped to search prior art (well, some more than others). Especially previous patents. Many people reckon that the main problem the US Patent and Trademark Office has had after the US Supreme Court opened the floodgates is that it lacked both the archives and the expertise to search software prior art. But I hear it's getting better at it, as it should be expected.
As enforcement of software patents in the U.S. increases, it will lead to the downfall of the software industry.
Thing is, despite some real excesses, the US software industry is still doing rather better than the European software industry, funnily enough...
posted by Skeptic at 12:43 PM on July 6, 2005
Skeptic
As for "technical character," ask a screenwriter. Screenwriting is an odd amalgam of creativity, craft, and rigorous structure, just as programming is. My point is that patents are awarded for mere tradecraft (e.g. patents awarded for the mouse "doubleclick", or a linear search through an array (for i = 1 to limit...). They make absolutely no sense within the context of software development: what's being patented are not inventions, but building blocks.
Prior art on an abstract concept (which is what most software patents cover) is a nightmare to establish, and it's even harder to disprove. Imagine, for a moment, that there are one million existing granted software patents: a prior art search would require that each be reviewed, interpreted, and compared, to see if the underlying concept is the same: there is no universal representational language or symbology that can make these searchable, or comparable. What's more, prior art must consider the universe of general practice, and a propertly-granted patent must consider the 40+ years of software practices, structures, methods, and techniques. What kind of resources does the patent office devote to such analysis? And even if the resources were adequate to the task (and they can never be), is it good or bad public policy to allow patents on what are essentially approaches to addressing a task?
If prior art searches were even vaguely competent, the patent office would not grant patents on algorithms that have been in wide use for 40+ years. I know two people who were granted patents (on behalf of their employers) for techniques they copied out of an undergraduate programming text (and both patents were granted within the last five years).
The U.S. software industry is the world's strongest not because it has software patents, but because of characterics of American culture: entrepeneurialsm, pragmatism, a low regard for social heirarchy, an embrace of individual achievement, etc. Add to that a substantial head start in these technologies, largely thanks to research funding and education in the 60s and 70s. I don't think it is wise to be complacent about our lead, though. Software patents are shifting the market advantage from the innovative and resourceful to the powerful and litigious.
posted by curtm at 1:22 PM on July 6, 2005
As for "technical character," ask a screenwriter. Screenwriting is an odd amalgam of creativity, craft, and rigorous structure, just as programming is. My point is that patents are awarded for mere tradecraft (e.g. patents awarded for the mouse "doubleclick", or a linear search through an array (for i = 1 to limit...). They make absolutely no sense within the context of software development: what's being patented are not inventions, but building blocks.
Prior art on an abstract concept (which is what most software patents cover) is a nightmare to establish, and it's even harder to disprove. Imagine, for a moment, that there are one million existing granted software patents: a prior art search would require that each be reviewed, interpreted, and compared, to see if the underlying concept is the same: there is no universal representational language or symbology that can make these searchable, or comparable. What's more, prior art must consider the universe of general practice, and a propertly-granted patent must consider the 40+ years of software practices, structures, methods, and techniques. What kind of resources does the patent office devote to such analysis? And even if the resources were adequate to the task (and they can never be), is it good or bad public policy to allow patents on what are essentially approaches to addressing a task?
If prior art searches were even vaguely competent, the patent office would not grant patents on algorithms that have been in wide use for 40+ years. I know two people who were granted patents (on behalf of their employers) for techniques they copied out of an undergraduate programming text (and both patents were granted within the last five years).
The U.S. software industry is the world's strongest not because it has software patents, but because of characterics of American culture: entrepeneurialsm, pragmatism, a low regard for social heirarchy, an embrace of individual achievement, etc. Add to that a substantial head start in these technologies, largely thanks to research funding and education in the 60s and 70s. I don't think it is wise to be complacent about our lead, though. Software patents are shifting the market advantage from the innovative and resourceful to the powerful and litigious.
posted by curtm at 1:22 PM on July 6, 2005
Skeptic
As for "technical character," ask a screenwriter. Screenwriting is an odd amalgam of creativity, craft, and rigorous structure, just as programming is. My point is that patents are awarded for mere tradecraft (e.g. patents awarded for the mouse "doubleclick", or a linear search through an array (for i = 1 to limit...). They make absolutely no sense within the context of software development: what's being patented are not inventions, but building blocks.
Prior art on an abstract concept (which is what most software patents cover) is a nightmare to establish, and it's even harder to disprove. Imagine, for a moment, that there are one million existing granted software patents: a prior art search would require that each be reviewed, interpreted, and compared, to see if the underlying concept is the same: there is no universal representational language or symbology that can make these searchable, or comparable. What's more, prior art must consider the universe of general practice, and a propertly-granted patent must consider the 40+ years of software practices, structures, methods, and techniques. What kind of resources does the patent office devote to such analysis? And even if the resources were adequate to the task (and they can never be), is it good or bad public policy to allow patents on what are essentially approaches to addressing a task?
If prior art searches were even vaguely competent, the patent office would not grant patents on algorithms that have been in wide use for 40+ years. I know two people who were granted patents (on behalf of their employers) for techniques they copied out of an undergraduate programming text (and both patents were granted within the last five years).
The U.S. software industry is the world's strongest not because it has software patents, but because of characterics of American culture: entrepeneurialsm, pragmatism, a low regard for social heirarchy, an embrace of individual achievement, etc. Add to that a substantial head start in these technologies, largely thanks to research funding and education in the 60s and 70s. I don't think it is wise to be complacent about our lead, though. Software patents are shifting the market advantage from the innovative and resourceful to the powerful and litigious.
posted by curtm at 1:26 PM on July 6, 2005
As for "technical character," ask a screenwriter. Screenwriting is an odd amalgam of creativity, craft, and rigorous structure, just as programming is. My point is that patents are awarded for mere tradecraft (e.g. patents awarded for the mouse "doubleclick", or a linear search through an array (for i = 1 to limit...). They make absolutely no sense within the context of software development: what's being patented are not inventions, but building blocks.
Prior art on an abstract concept (which is what most software patents cover) is a nightmare to establish, and it's even harder to disprove. Imagine, for a moment, that there are one million existing granted software patents: a prior art search would require that each be reviewed, interpreted, and compared, to see if the underlying concept is the same: there is no universal representational language or symbology that can make these searchable, or comparable. What's more, prior art must consider the universe of general practice, and a propertly-granted patent must consider the 40+ years of software practices, structures, methods, and techniques. What kind of resources does the patent office devote to such analysis? And even if the resources were adequate to the task (and they can never be), is it good or bad public policy to allow patents on what are essentially approaches to addressing a task?
If prior art searches were even vaguely competent, the patent office would not grant patents on algorithms that have been in wide use for 40+ years. I know two people who were granted patents (on behalf of their employers) for techniques they copied out of an undergraduate programming text (and both patents were granted within the last five years).
The U.S. software industry is the world's strongest not because it has software patents, but because of characterics of American culture: entrepeneurialsm, pragmatism, a low regard for social heirarchy, an embrace of individual achievement, etc. Add to that a substantial head start in these technologies, largely thanks to research funding and education in the 60s and 70s. I don't think it is wise to be complacent about our lead, though. Software patents are shifting the market advantage from the innovative and resourceful to the powerful and litigious.
posted by curtm at 1:26 PM on July 6, 2005
krunk writes "So if I invent a novel software concept, then a huge corporation with lots of resources can copy the general concept (but not necessarily the implementation) and out-market me. I hope someone can refute this."
Well here's the thing:
1) You probably don't have the capital to patent your idea. And if you do you probably don't have the money to protect it. SCO had to get RBC on board to sustain their suit against IBM.
2) IBM has something like 250,000+ patents. Unless your idea is a software implementation of cold fusion (the energy producing kind) or something your probably violating at least one of IBM's.
3) IBM has way more laywers than you. Which bizarrely hasn't stopped SCO from sueing them.
4) Patents (and this is one of the things MS's team is after) basically kill many GPL'd projects, not because they infringe but because of #1.
5) Abuse. The Amazon 1-Cick thing is an absolute classic abuse. The idea wasn't inovatative. It was an obvious (to computer programmers) solution to an obvious marketing request. If the patent had been proposed for a brick and mortar place it would have been laughed at. But because it was on the shiny new internet a patent was granted. The GIF patent is another classic case of abuse.
To sum up the potential benifits are mostly unobtainable and wildly out weighed by the harm.
posted by Mitheral at 2:21 PM on July 6, 2005
Well here's the thing:
1) You probably don't have the capital to patent your idea. And if you do you probably don't have the money to protect it. SCO had to get RBC on board to sustain their suit against IBM.
2) IBM has something like 250,000+ patents. Unless your idea is a software implementation of cold fusion (the energy producing kind) or something your probably violating at least one of IBM's.
3) IBM has way more laywers than you. Which bizarrely hasn't stopped SCO from sueing them.
4) Patents (and this is one of the things MS's team is after) basically kill many GPL'd projects, not because they infringe but because of #1.
5) Abuse. The Amazon 1-Cick thing is an absolute classic abuse. The idea wasn't inovatative. It was an obvious (to computer programmers) solution to an obvious marketing request. If the patent had been proposed for a brick and mortar place it would have been laughed at. But because it was on the shiny new internet a patent was granted. The GIF patent is another classic case of abuse.
To sum up the potential benifits are mostly unobtainable and wildly out weighed by the harm.
posted by Mitheral at 2:21 PM on July 6, 2005
Mitheral:
I work for smallish software company that can afford to patent our software, and in fact we do just that. Our main competitor is one of the largest companies in the world; if we don't patent our work, what is to stop them from copying our ideas and getting contracts that we both compete for?
I'm not for or against software patents, I'd just like to know.
posted by krunk at 2:27 PM on July 6, 2005
I work for smallish software company that can afford to patent our software, and in fact we do just that. Our main competitor is one of the largest companies in the world; if we don't patent our work, what is to stop them from copying our ideas and getting contracts that we both compete for?
I'm not for or against software patents, I'd just like to know.
posted by krunk at 2:27 PM on July 6, 2005
If this approach were applied to more accessible crafts--like woodworking--it would be easier to see how ridiculuous and counterproductive software patents are. Imagine:
6191928: Sanding a smooth surface using a circular motion.
5158682: Using four legs, one one each corner, to hold the surface of a table above ground level.
Um, if no one had ever done those things, they would definetly be patentable, especialy the 'circular motion' one.
posted by delmoi at 2:42 PM on July 6, 2005
6191928: Sanding a smooth surface using a circular motion.
5158682: Using four legs, one one each corner, to hold the surface of a table above ground level.
Um, if no one had ever done those things, they would definetly be patentable, especialy the 'circular motion' one.
posted by delmoi at 2:42 PM on July 6, 2005
Nothing really. But that's no different then the situtation with patents. IF MS or IBM or heck anyone with deeper pockets comes after you they can kill your company by outspending it on litigation. If they take notice of you, you are undoubtably infringing one a patent or two of their's. They threaten a suit, you work out a cross patent library agreement, the lawyers earn their retainer. You haven't protected your contracts just forwarded a bunch of profits to your laywers.
I'm not saying patents are bad. Just that software patents tend to be by their nature (because they are mostly ideas) very vague and open to broad interpretation. This leads to abuses and rarely helps small companies while at the same time giving big companies legal sticks to beat their smaller competitors with.
posted by Mitheral at 2:44 PM on July 6, 2005
I'm not saying patents are bad. Just that software patents tend to be by their nature (because they are mostly ideas) very vague and open to broad interpretation. This leads to abuses and rarely helps small companies while at the same time giving big companies legal sticks to beat their smaller competitors with.
posted by Mitheral at 2:44 PM on July 6, 2005
curtm: I'm familiar both with the art of storytelling and patent law. I'm certain that scriptwriting involves a lot of technique, just not technical character in the sense of (European) patent law.
They make absolutely no sense within the context of software development: what's being patented are not inventions, but building blocks.
Which is in fact the essence of good patent drafting. If you claim the combination of two such building blocks, you'll only get protection for that combination. If you claim only one of the building blocks, you get protection for anything involving that building block. If something is a solution to a given problem, it is considered an invention. If that building block is there for a reason, addresses a particular task in a way that has not been done before, it is an invention.
Prior art on an abstract concept (which is what most software patents cover) is a nightmare to establish, and it's even harder to disprove. Actually, that is one of the things I do for a living, and I know all too well how incredibly hard it can be.
Finally, I know that the US software industry is not so strong because of the patent system. Indeed, I wrote it was so despite its excesses. But I don't think that the source of the US patent system's troubles lie in allowing patents for computer programs, but rather in some of its more particular aspects (bars to reviewing a granted patent, to present third-party evidence during granting procedures, highly complicated rules governing prior art and, definitely not least the first-to-invent rule), as well as to general problems of the American civil justice system (very high costs and litigation culture).
posted by Skeptic at 2:49 PM on July 6, 2005
They make absolutely no sense within the context of software development: what's being patented are not inventions, but building blocks.
Which is in fact the essence of good patent drafting. If you claim the combination of two such building blocks, you'll only get protection for that combination. If you claim only one of the building blocks, you get protection for anything involving that building block. If something is a solution to a given problem, it is considered an invention. If that building block is there for a reason, addresses a particular task in a way that has not been done before, it is an invention.
Prior art on an abstract concept (which is what most software patents cover) is a nightmare to establish, and it's even harder to disprove. Actually, that is one of the things I do for a living, and I know all too well how incredibly hard it can be.
Finally, I know that the US software industry is not so strong because of the patent system. Indeed, I wrote it was so despite its excesses. But I don't think that the source of the US patent system's troubles lie in allowing patents for computer programs, but rather in some of its more particular aspects (bars to reviewing a granted patent, to present third-party evidence during granting procedures, highly complicated rules governing prior art and, definitely not least the first-to-invent rule), as well as to general problems of the American civil justice system (very high costs and litigation culture).
posted by Skeptic at 2:49 PM on July 6, 2005
krunk:
A company like the one you work for is often building its patent portfolio for (at least mostly) defensive purposes. When somebody else comes along claiming violations, they start looking at the accuser's software, and start counter-suits for all the infringements they find. Then both sides settle with crosslicensing agreements.
Also, does the current patent system manage to stop them from doing that? As if yet, I've never seen even one case where it actually works that way.
posted by atbash at 11:24 PM on July 6, 2005
A company like the one you work for is often building its patent portfolio for (at least mostly) defensive purposes. When somebody else comes along claiming violations, they start looking at the accuser's software, and start counter-suits for all the infringements they find. Then both sides settle with crosslicensing agreements.
Also, does the current patent system manage to stop them from doing that? As if yet, I've never seen even one case where it actually works that way.
posted by atbash at 11:24 PM on July 6, 2005
« Older Obituaries of the Future | MMORPG Sweatshops Newer »
This thread has been archived and is closed to new comments
Pretty damning defeat, though - especially for something which (at least initially) most of the ministers seemed to not understand...
posted by Chunder at 7:08 AM on July 6, 2005