Contribute  :  Web Resources  :  Site Statistics  :  About  :  RSS  
    Bart's Soapbox talk about technology and politics    

A few notes on software and patents

   

PoliticsAfter having a long and interesting discussion with someone knowledgable on the patent system in the USA, I have been thinking a lot about why the whole concept of software patents feels so wrong to me as a person who deals a lot with software development.

The first issue seems to be that to the law, a general purpose computer running some specific program makes for 'an aperatus' and as such may be subject to patents.

For a software developer, a general purpose computer is a mere tool that is needed to make the program execute. The running of a specific program does not change the general purpose nature of that computer, at least not in the eyes of a software developer.

One could argue that in many cases, the view of the software developer is factually correct, only a small percentage of the general purpose computers nowadays is used to execute one specific program only, and a large majority runs a multitude of programs at the same time.

The consequence of this from a developers point of view is that any patent that seems to implement an in itself obvious* idea in software on a general purpose computer, looks like claiming non obviousness just because of using a tool in the exact way it is intended to be used (a computer to run software that implements a pre-existing idea)

An other issue is that for as far as software goes, the first person to succesfully implement something is seldom the person who first thought of it. It is usually either the first one actually faced with the problem and needing a solution, or the first one having access to a computer capable of performing it.

The consequence for the patent system should be that 'obvious to someone skilled in the art' should be assumed for software (at least on the level of software algorithms).

One problem is that the legal definition of obvious is not quite what you'd expect as becomes clear in the slashdot discussion mentioned above. Hence, while most things would be obvious to someone skilled in the art for as far as the English language goes, not so in legal language.

Still, the mere application of a tool in the way that tool is intended to be used should not make enough of a difference to claim a new, non obvious invention.

Many companies that are currently pushing for allowing software patents in Europe argue that most software runs on embedded devices that do make for an aperatus, and I can quite see the validity of that argument.

The question becomes if now that many a piece of electronics can be replaced by a general purpose device (being it a personal computer, a fpga or any other programmable device) running a 'program', if patentebility of such things shouldn't be reviewed instead of allowing programs running on a device to be patented including specific software algorithms.

I can see how a temporary exclusive right can be an encouragement to strive for new things, but this is quite some right to give to anyone in what is supposed to be a free market and I think the idea should be to allow fewer instead of more of them.




What's Related

Story Options

A few notes on software and patents | 12 comments | Create New Account
The following comments are owned by whomever posted them. This site is not responsible for what they say.
A few notes on software and patents
Authored by: back_pages on Thursday, April 28 2005 @ 03:43 AM CEST
Thanks for the compliments. (I am the guy from Slashdot)

I think you're quite right and describe things pretty accurately from the software developer point of view. I had a similar conversation the other day that hit on similar points but was closer to what you said on Slashdot. My analogy was that software engineering is like manufacturing a car, except all the costs of going from design to final product are basically eliminated and it happens instantly. It's astonishingly easy to go from "idea" to "implementation" to "product" that it is hard to see the distinction sometimes. Nobody would seriously argue that a car is just a temporary arrangement of molecules that exist in nature, yet people DO argue that software is just a big number or an abstract algorithm. Of course this is technically true, but it's an absurd argument. I'm not posting this with my abstract algorithm machine - software is, in many ways, a manufactured product like a car.

The second difficulty about this issue is that there is a LOT of misunderstanding of how the patent system works and adjusts. The law that broadly determines what can or cannot be patented is 35 USC 101. It is one short sentence. The words "useful process" are the core of all the software-related patents. Those two words have produced volumes of court decisions (where judges interpret the law), board decisions (where patent judges decide if the USPTO has followed the court decisions), and no shortage of internet debate. But for all of that, software-related patents are only about 10 years old, dating back to Diamond v. Diehr. The USPTO is still examining cases with identical sets of claims for "computer-implemented methods", "computer program code on a computer readable medium", and "executable instructions in a data carrier wave" in the same application. Why? Because those applications were filed 3-4 years ago and nobody - attorneys, judges, USPTO, and inventors - had ANY idea what was going to be the "right" patentable phrase in 2005. The whole venue is so new and unstable that you had to play a guessing game with your patent application. Two sets of claims would be thrown out but hopefully one of them would be patentable when the case was up for examination.

Why does that matter? It shows how the issue of patentability is still a rapidly moving target for software-related inventions. Some patents have been issued that will be invalidated with more recent court decisions. The patents issued in 2006 will have different requirements to pass than those issued in 2004. As time goes on, the case law will get better, the board and the USPTO will know how to enforce that case law consistently, and the wackiness will die down.

The second reason that matters is how the moving target idea applies to my first point - the fact still remains that software is not patentable by itself. A tangible embodiment of that software could be patentable. That software has to actively do something in the claims. It has to have a practical application. Working backward from that point, a naked algorithm is unpatentable. An encryption algorithm is unpatentable. An encryption algorithm that is claimed as taking input and producing output is unpatentable. When that algorithm is part of a tangible computer system, taking input and producing output, and that output is still useful (via decryption, most likely) then you have a candidate for a patent. But that might change next year, and it certainly was NOT the case 5 years ago. That also reflects the difference between the idea of a car, the blueprint for a car, a factory that could produce a car, and an actual car itself. You can't patent the idea, the design, or the code, but you could patent the code as it executes on a computer system.

Shew, pant pant, and if that's not enough inane legal analysis, the USPTO is up to its neck in this crap as well. They do all they can to stay up to speed with the legal requirements, but it is a moving target. My friends on the inside inform me that about 4 times a year they get new and updated flowcharts and guidelines to help them decide whether or not a software-related invention is patentable under 35 USC 101. They get about 20 hours per application to get -everything perfect-. This particular topic is never black and white, so how much of a fight can they give the attorneys? (Answer: not much - they can only afford to spend time fighting the fights they KNOW they can win. 35 USC 101 is almost never a sure win.)

It was great talking to you. Best of luck in everything.

A few notes on software and patents
Authored by: bart on Saturday, April 30 2005 @ 09:59 PM CEST
Thanks for your explanations (both on slashdot and on here).

There is one aspect that I think is missing from this discussion as well as the legal arguments made about patents.

The US constitution allows congress to make laws that grant a short term 'monopoly' like patents and copyrights with a very specific purpose: "To promote usefull inventions and works of art".

I would argue that when the USPTO as well as the courts have to decide on the patentability of some new type of 'invention', they should also look at if allowing patents is in fact promoting usefull inventions or (as I believe is the case with software) actually hinder usefull inventions.

The argument as to why patents hinder usefull inventions in software have been made by many professionals in the field, including people like Linus Torvalts, Alan Cox and many others involved in colaborate software development projects.

I have yet to see a reasoned argument as to why patents do in fact promote usefull inventions in the area of software development. The fact that they do promote usefull inventions in other areas is not a reasoned argument, those areas are as you point out yourself quite different, specifically with regards to the efford (and cost) involved in going from idea to an actually working device (product or otherwise).
A few notes on software and patents
Authored by: back_pages on Sunday, May 01 2005 @ 07:05 AM CEST
That's a very good point. However, I think all reasonable people can agree that it's extremely difficult to create a legal definition of a relative concept such as whether or not a particular invention promotes science.

Regarding the software-related arts, this is an argument that the courts have not yet decided definitively. I've read and understand the arguments by all the geek celebrities (if you'll allow such a label). When reading those open letters and arguments, I'm first struck by their lack of expertise in the law, which isn't surprising but is nevertheless their weakness. Secondly, I observe that they tend to hold an extreme point of view that isn't satisfied by comprimise. I can't help but notice that historically this is a tactic that is rarely satisfied.

To address your implied question directly, I feel that it isn't so proper to ask, "Why does this software-related invention promote innovation?" because we could easily agree that some software-related inventions do indeed promote innovation. We could bicker about where to draw the line, but from SETI-at-home to protein folding simulations to streamlined internet commerce to gmail, somewhere we can probably all agree that there has been some legitimate invention that could deserve patent protection. The -critical- problem is that the USPTO -must- draw that line somehow in a way that doesn't unreasonably burden society or the courts. The more appropriate question is, "What is the black and white distinction between inventions that promote progress and those that do not?" Man, beats me. I work in this industry and I honestly could not break that down into a black and white distinction.

I fully understand that many people believe that the current practice burdens society, but the current practice also burdens the board of appeals with an 18-24 month appeal process, and that's if nothing goes wrong. And keeping in mind that a patent is not an infallible document, I don't think that the current line of patentability at the USPTO is radically far from where it should be. Should it be adjusted? Of course - and it always will be, in pliers, drugs, light bulbs, and software.

Creating legislation is an exact science for an inexact world - if you provide a definition of a "good and patentable invention" that I can violate, yet reasonable people would agree it was a "good invention", your definition would fail in the courts. This is the -critical- problem; the veritable million dollar question. How can you distinguish between a "good software invention" and a "bad software invention", without failure or error, in one paragraph, and in a way that won't be offensively ludicrous to the existing standards in industry and law?

A few notes on software and patents
Authored by: bart on Monday, May 02 2005 @ 01:51 PM CEST
I can agree that at least part of the leading voices in OSS development take an extreme point of view in this, but I see no difference with thosw who promote software patents, quite a few there seem to be at least as extreme in their statements. Both only look at one side of the issue and their arguments should be viewed that way.

They are of course not legal experts, and in many cases have limited to very limited understanding of the legal aspects of this. The same applies to me, tho I am trying to understand (thanks a lot for your help in that)

The flipside of that is that few legal experts seem to understand much about the actual process of developing software and inventing new methods used in software. This is a pot/kettle argument imho, what needs to happen is that both sides need to help eachother with at least understanding the fundamentals of eachothers field of expertise.

Also, we should not forget that the legal system is not there to serve itself, it is there to serve and protect society. When the workings of the legal system conflict with that, the legal system, not society needs to change.

With regards to promotion of science and usefull inventions, I disagree with your argument. The issue is not if an invention in itself promotes other usefull inventions and/or science, but if the fact that it is patented promotes usefull inventions and science, and if the now patented invention would not have been made without such patents.

Patents burden society because they effectively establish a temporary monopoly (or to be more precise, grant exclusive rights for a limited time). A monopoly burdens society for the simple reason that it limits choice. A well known result of this is that prices for a product based on such an invention are determined by what the market is willing to pay, and not by competition and actual cost of production. This means that society pays a higher price then needed.

Besides this financial burden there is also the burden of basicly creating a minefield for anyone who wants to write software either for hobby or commercial purposes.

In some fields it is quite possible to get around that by creating an equivalent product with an alternative for the patented invention but in the field of software this is often not possible. This is in part a result of the nature of software development, and partially because of the way the claims in many software patents made. Many such claims can be fought in court, but few people have the financial means to do this (and when they do, we are back to it being an economic/financial burden on society)

So, while it may be possible to find individual cases where a pateted invention promotes science, I believe we should first look at the bigger picture.

For example, in case of molecular modelling, used for example to research possible new drugs, there is more then enough motivation to do this because of the resulting ability to create new drugs. I do not believe there needs to be an extra motivator in the form of exclusive rights, and seeing how there has been a lot of software development and inventions before the courts started allowing for software patents, I strongly believe that history of software development supports this opinion.

Also, we should not forget that software developed for such a thing (or any other use for that matter)is already protected by copyright, so a competitor can not just take it and use it, they will have to do a similar investment in developing equivalent software.

You are of course right that promotion of usefull inventions is not something that can be measured easily, and that is actually rather relative. Reading the relevant part of the constitution and the letters from those who wrote it, I get the strong impression that such exclusive rights should be the exception, not the rule. This imho means that such exclusive rights should not be granted unless there is good reason to believe they do in fact promote usefull inventions. Civil courts can work with concepts as reasonable doubt, so I don't think that a somewhat relative measurement should be that big a problem.
A few notes on software and patents
Authored by: back_pages on Tuesday, May 03 2005 @ 04:15 AM CEST
With regards to promotion of science and usefull inventions, I disagree with your argument. The issue is not if an invention in itself promotes other usefull inventions and/or science, but if the fact that it is patented promotes usefull inventions and science, and if the now patented invention would not have been made without such patents.

That's actually a very elegant and relevant distinction - does the act of patenting a particular invention promote technology in a way that the invention itself would not? I don't think there is an easy way to assess that in a way that would actually move the controversy forward. This is probably related to the general idea of "obviousness" in many ways. Of course, this is how the legal concept of obviousness began, but turning it into law made it the "combination of prior art" concept that it is today. You know what they say about laws and sausage - yet law is indeed what runs defines this topic.

Patents burden society because they effectively establish a temporary monopoly (or to be more precise, grant exclusive rights for a limited time). A monopoly burdens society for the simple reason that it limits choice. A well known result of this is that prices for a product based on such an invention are determined by what the market is willing to pay, and not by competition and actual cost of production. This means that society pays a higher price then needed.

I would quibble with this argument - you are blurring the notions of "patent monopoly" and "commercial monopoly". They are related but quite different. The percentage of patents that are commercially successful is tiny - something near 5%. The other 95% cost more money to patent than they're ever worth. A patent monopoly -requires- public disclosure of the technology -specifically for the purpose- of enabling the public to improve upon it and, in effect, create a better alternative. It is true that a patent monopoly is a burden on the public, but not in the way you have described.

You have described a commercial monopoly, and this only coincides with a patent monopoly in those very few cases where a particular technology is patented, is commercially successful, and for one reason or another, improvements are not commercially viable. The compact disc is probably the truest example of this, but I hardly hear people complaining that the compact disc is an anchor that drags down technological progress.

For example, in case of molecular modelling, used for example to research possible new drugs, there is more then enough motivation to do this because of the resulting ability to create new drugs. I do not believe there needs to be an extra motivator in the form of exclusive rights, and seeing how there has been a lot of software development and inventions before the courts started allowing for software patents, I strongly believe that history of software development supports this opinion.

I'm not sure I follow this argument. Certainly it is conceivable that a drug company could produce a revolutionary new line of drugs by heavily investing time, money, research, and intellectual property into molecular modeling techniques. However, the incentive to do so for a corporation is the financial reward. If that financial reward is inextricably tied to protecting the results of their research - the modeling technique - then it would be far riskier to make that investment in a world that refuses them patent protection. (That's based on the assumption that a competing company, armed with free use of the new modeling technique, could come to market with the same drugs but without paying for the research and development of the modeling technique.)

Copyright says you cannot copy their code, but it says nothing about implementing their methods on an analogous system. Patents say you cannot copy their METHOD, but they have to disclose what it is - and patents also guarantee that someone else can improve that method, disclose that improvement, and similarly protect that research investment as well. Who knows - develop and patent a significant improvement to their method and you'll probably find yourselves enjoying a lucrative cross-licensing deal, including the right to use their method.

I'm not so naive or ignorant of current events to say that philosophy works fantastically all the time, but the opposition wants to throw the baby out with the bath water. I think they would be far more effective in their arguments if they focused on a specific, focused topic rather than broadly arguing against "all software patents". The former would probably find a lot of support at the USPTO while the latter is an argument that could be dismantled by someone with a few hours of training in the patent system.

Also, we should not forget that software developed for such a thing (or any other use for that matter)is already protected by copyright, so a competitor can not just take it and use it, they will have to do a similar investment in developing equivalent software.

But that's the problem. Remember the software as a manufactured product dilemma? Making that equivalent software is FAR easier than coming up with the new method. I'm not an expert on copyright law, but apparently the "derivative work" clause is not strong enough to sufficiently protect software. Bolstering that copyright protection might improve the patent controversy, but could easily become a 1st Amendment disaster here in the US. Patents protect you against equivalent software but, in principle, motivates people to improve upon your methods. Whether or not it's working is one argument, but I think it's very clear that this is a completely different protection than copyright provides. Copyright basically protects your authorship, patents protect the meaningful implementation of your concept.

Bleh this got long. Good conversation though. By the way, are you familiar with 37 CFR 1.99? That's the rule that allows 3rd parties to submit prior art. That's a good one to know..

A few notes on software and patents
Authored by: bart on Wednesday, May 04 2005 @ 06:29 PM CEST
I'm not sure I follow this argument. Certainly it is conceivable that a drug company could produce a revolutionary new line of drugs by heavily investing time, money, research, and intellectual property into molecular modeling techniques. However, the incentive to do so for a corporation is the financial reward. If that financial reward is inextricably tied to protecting the results of their research - the modeling technique - then it would be far riskier to make that investment in a world that refuses them patent protection. (That's based on the assumption that a competing company, armed with free use of the new modeling technique, could come to market with the same drugs but without paying for the research and development of the modeling technique.)

I believe that while this will definitely apply to a lot of technology out there, in case of software this reasoning has some flaws.

  1. The modelling techniques are a tool, they enable new ways of researching molecular structures in order to make a drug, in themseles they do nothing. It is an enabling technology in order to make a product (the drug). A competitor will still have to do the actual research to make the product.
  2. In software the challange is seldom to come up with a new concept, it is making it work and usable. Even when someone takes your concept, they cannot easily get around this difficulty unless they effectively copy your code
  3. In software, a new method or concept is often conceived before there exists hardware that allows implementing it. This means that the first one to implement it is often not the one who thought it up, but the one who is first to have access to such hardware (which in itself can be patented) or the first one who actually needs the specific method or concept, or at times the first one who manages to make it work.
  4. It is often cheaper and more practical to buy a (copyrighted) program instead of making your own implementation, unless you want to try to sell your own implementation.
  5. Software is different from almost any other product in that there are no 'secret methods' to be disclosed. All the methods are in the actual product. If you obtain the product, you have the methods (alltho often not in a directly usable form for making your own re-implementation or improvements).
In my opinion, this means that both the disclosure and the 'protection of investment' arguments are at least a lot weaker for software then for most other areas of technology. Copyright says you cannot copy their code, but it says nothing about implementing their methods on an analogous system. Patents say you cannot copy their METHOD, but they have to disclose what it is - and patents also guarantee that someone else can improve that method, disclose that improvement, and similarly protect that research investment as well. Who knows - develop and patent a significant improvement to their method and you'll probably find yourselves enjoying a lucrative cross-licensing deal, including the right to use their method.

Or find yourself locked out of the market because the patent holder does not want to give up their leadign market position in their specific market (Apple, IBM and others tried this approach, at times succesfull also)

As argued above, I do not believe the disclosure argument is a very strong oen in case of software, often this disclosure already takes place because the methods used are an integral part of the software product.

One could easily achieve the same effect with regards to disclosure by 'banning' the 'no reverse engineering' clauses so common in the typical ULA. Removal of rules makes the system simpler, causing a lower burden on the courts and society.

I'm not so naive or ignorant of current events to say that philosophy works fantastically all the time, but the opposition wants to throw the baby out with the bath water. I think they would be far more effective in their arguments if they focused on a specific, focused topic rather than broadly arguing against "all software patents". The former would probably find a lot of support at the USPTO while the latter is an argument that could be dismantled by someone with a few hours of training in the patent system.

I understand that in order to prove infringement of a patent, one has to show that all elements of a specific claim are met by the infronging product (invention).

When you want to prove the damage done by the patent system, it is enough to show that enough (but by far not all) elements of your claim are indeed true.

This means that you cannot judge a claim of damage done by the same standards as you'd use for judging patent infringement.

The relevant question I believe for the USPTO and the courts is if there is a substantial reason to believe the patent system does more damage then good when applied to software inventions. When enough elements of damage can be shown beyond reasonable doubt then I believe patents should not be allowed for a specific area (software in this case).

One should understand that quite a few who are using the GPL for their software have chosen to use an entirely different approach to address virtually the same thing as patents do, ensure that those who are able can create and improve software, can build upon eachothers work (as to prevent reinventing the wheel). In fact, they have a system that enforces disclosure of improvements (unlike the patent system where one can simply opt to not patent and disclose an improvement).

Quite a few of the more vocal people there seem to regard this and patents to be mutually exclusive, and it will be very hard to convince them that both can coexist for software.

So it may seem they want to throw away the baby with the bath water but that is not the case, rather, they believe they simply have something that achieves the same things for society while being much less of a burden, and I believe there is a good reason to believe this can work for a commercial enterprise as well, seeing IBM's success with Linux on their system 390 hardware for example.

You can most likely dismantle a substantial part of the arguments being made against software patents, but I strongly doubt you can dismantle enough of them to show that they do more good then bad.

I believe that in order to get a cleaner and better discussion, we should start by defining what the purpose of patents is. I did a shot at it when refering to the US constitution in one of my earlier posts, and I'd be interested if the USPTO has defined for itself somewhere what patents are here for, and how they describe this. Could you enlighten me a bit about this?

A few notes on software and patents
Authored by: Anonymous on Saturday, May 07 2005 @ 03:10 AM CEST
Ahem, I'm just an anonymous user who stumbled upon this discussion..

The relevant question I believe for the USPTO and the courts is if there is a substantial reason to believe the patent system does more damage then good when applied to software inventions.

Actually, the USPTO isn't concerned with that question. The right to consider this question is granted to Congress and interpreted by the judicial branch. ::shrug:: The USPTO just does what it's told by those branches.

One should understand that quite a few who are using the GPL for their software have chosen to use an entirely different approach to address virtually the same thing as patents do, ensure that those who are able can create and improve software, can build upon eachothers work (as to prevent reinventing the wheel). In fact, they have a system that enforces disclosure of improvements (unlike the patent system where one can simply opt to not patent and disclose an improvement).

This is true and cannot be understated. Where US patent examiners have trouble is finding dated prior art. Walk the halls of the USPTO on any given day and you'll hear screams of, "Aaargh, why don't they date their work?!?!" I wish I had a nickel for every website that offered the "latest version" and you took $20 for every website that provides a dated backlog of their work - I'd be a millionaire. The GPL is great, but it only works for public disclosure when the material is dated and catalogued backwards through the last 5 years.

So it may seem they want to throw away the baby with the bath water but that is not the case, rather, they believe they simply have something that achieves the same things for society while being much less of a burden, and I believe there is a good reason to believe this can work for a commercial enterprise as well, seeing IBM's success with Linux on their system 390 hardware for example.

I'm not terribly familiar with IBM on system 390, but in principle you are quite right. Something that is publicly disclosed under the GPL (or any other public disclosure) that is adequately dated and documented will preclude patenting on the same material, including derivative material that can be shown as legally obvious.

But, again, how much publicly documented GPL software is concerned with 1-click shopping? There's a lot of great GPL software, but attorneys will specifically avoid claiming something that's clearly and publicly disclosed. Whaddya gonna do?

You can most likely dismantle a substantial part of the arguments being made against software patents, but I strongly doubt you can dismantle enough of them to show that they do more good then bad.

Hehehe, if you're talking to an examiner, you're talking to someone who is paid a lot of money to argue. If you were talking to an examiner ;), he'd say that you can doubt all you like. In fact, have a license to doubt. All the doubt in the world doesn't prove the converse - that you can show that they do more bad than good. Of course, on any given week of the year, the USPTO issues somewhere between 150-400 patents on computer-related inventions (450-1600 per month), but you only hear about (let's be generous here) 50 per year as absurd. Hm, QED?

I believe that in order to get a cleaner and better discussion, we should start by defining what the purpose of patents is. I did a shot at it when refering to the US constitution in one of my earlier posts, and I'd be interested if the USPTO has defined for itself somewhere what patents are here for, and how they describe this. Could you enlighten me a bit about this?

I certainly can't speak for the other poster, but the purpose of patents are clearly stated in the US Constitution. What that means on a daily basis in the trenches is certainly a matter of judgement. However, the USPTO is FAR more bound by following the law than some vigilante force that gets to toss out bullshit inventions that they don't like - quite the opposite. They're literally bound by law to issue legally valid patents on bullshit inventions from time to time. It's up to the court system to recognize that it was a bullshit invention, change the law, and allow the USPTO to toss it out the next time one shows up. Like it or not, that's how it goes. Precious few people really understand the previous three sentences.

By the way, nice site. Keep up the good work. An application for an invention titled "Inventions" that includes a sink liner, a video game system, and a package for brown sugar was recently published. Find that puppy and you might provoke some more discussion.

A few notes on software and patents
Authored by: bart on Saturday, May 07 2005 @ 06:37 AM CEST
Ahem, I'm just an anonymous user who stumbled upon this discussion..

Always nice to get some opinion and feedback :)

Actually, the USPTO isn't concerned with that question. The right to consider this question is granted to Congress and interpreted by the judicial branch. ::shrug:: The USPTO just does what it's told by those branches.

I understad that, but untill a judge decides otherwise, it is the USPTO who interprets what the law means. You are of course right that congress should ask itself this question first and try toi answer it.

It is clear to me that quite a few people at the USPTO may not be expert on software development, but they are on the subject of patents. Just for that reason their opinion matters.

This is true and cannot be understated. Where US patent examiners have trouble is finding dated prior art. Walk the halls of the USPTO on any given day and you'll hear screams of, "Aaargh, why don't they date their work?!?!" I wish I had a nickel for every website that offered the "latest version" and you took $20 for every website that provides a dated backlog of their work - I'd be a millionaire. The GPL is great, but it only works for public disclosure when the material is dated and catalogued backwards through the last 5 years.

Often such a backlog is available in the sense of a cvs commit log and repository.

It seems that what is available is mostly aimed at those involved with writing the software, but its quite difficult to manage a software development project of some size without the help of some version control system, so most at least have some sort of documented history with regards to what was put in when.

I'm not terribly familiar with IBM on system 390, but in principle you are quite right. Something that is publicly disclosed under the GPL (or any other public disclosure) that is adequately dated and documented will preclude patenting on the same material, including derivative material that can be shown as legally obvious.

Well, the point was that the GPL simply does not need patents to get the benefit of public disclosure. While it is possible to hold a GPLed project within a circle if all members of that circle decide to do so, its not possible to enforce such a situation. If you let anyone other then yourself get the resulting 'product' or 'invention', the GPL requires you also let them get access to the souce code needed to 'build' it. When comparing this to a physical (hardware) invention, it is like getting a sample of the product and the instructions on how to make more of them with the cloner device that you have at home already.

The simple result for people using the GPL is that patents on things they use in software they develop are merely a distraction and not needed at all, not to mention they are incompatible with the GPL and similar licenses.

But, again, how much publicly documented GPL software is concerned with 1-click shopping?

I think these guys would be quite interested in using the 'technology' for example.

There's a lot of great GPL software, but attorneys will specifically avoid claiming something that's clearly and publicly disclosed. Whaddya gonna do?

Me? I am not too interested in proving prior art or obviousness on everything, I rather believe that patents on software algorithms should be treated identical to mathematical algorithms, and that I have yet to hear a good case being made in favor of the current situation.

Hehehe, if you're talking to an examiner, you're talking to someone who is paid a lot of money to argue. If you were talking to an examiner ;), he'd say that you can doubt all you like. In fact, have a license to doubt. All the doubt in the world doesn't prove the converse - that you can show that they do more bad than good.

Which in itself is no problem. I believe that when granting exclusive rights as with a patent or copyright, it is upto those who seek such exclusive rights to show beyond reasonable doubt that they do more good then bad.

Of course, on any given week of the year, the USPTO issues somewhere between 150-400 patents on computer-related inventions (450-1600 per month), but you only hear about (let's be generous here) 50 per year as absurd. Hm, QED?

Well, the real problem is not the few really absurd patents out there, the problem is that software development is something that can be done by individual people at least as well as by large companies in many cases. Patents strongly favor the companies simply because of the financial aspects of writing it properly and being able to enforce it through the courts if needed. Software development is becomming more like a minefield then an area of technological development in many cases. To me this seems to conflict with the basic purpose of that provision in the constitution for patents.

This all would put a large burden on society with a smaller payoff then that 'society' manages with this thing called GPL., or at least that is what quite a few people seem to argue.

I certainly can't speak for the other poster, but the purpose of patents are clearly stated in the US Constitution. What that means on a daily basis in the trenches is certainly a matter of judgement.

The constitution is very clear about why congress can allow for such exclusive rights for a limited amount of time, yes. This is why I believe it is upto those who want to get such exclusive rights in a new area of technological development must show why it is benificial to society to allow patents for it.

However, the USPTO is FAR more bound by following the law than some vigilante force that gets to toss out bullshit inventions that they don't like - quite the opposite. They're literally bound by law to issue legally valid patents on bullshit inventions from time to time. It's up to the court system to recognize that it was a bullshit invention, change the law, and allow the USPTO to toss it out the next time one shows up. Like it or not, that's how it goes. Precious few people really understand the previous three sentences.

I do understand them and I quite agree with you there.

By the way, nice site. Keep up the good work. An application for an invention titled "Inventions" that includes a sink liner, a video game system, and a package for brown sugar was recently published. Find that puppy and you might provoke some more discussion.

Thanks, will see if I can find it when I have some more time.

A few notes on software and patents
Authored by: Anonymous on Sunday, May 08 2005 @ 06:06 AM CEST
Often such a backlog is available in the sense of a cvs commit log and repository.

Unfortunately, that is not sufficient to use for prior art. Here's why: 1) The examiner needs to figure what the reference is doing in about 10 minutes. Sometimes he's concerned with the details of implementation, but sometimes he's concerned with the user interface. A CVS is an extremely inefficient way of getting documented evidence of the big picture. 2) He needs to present the reference with a persuasive argument to an attorney. Source code accompanied by the allegation that it does what the examiner says it does is not persuasive. The attorney is under no obligation to compile the code and see what it does - he is merely under the obligation to read the code and see what it says. If the comments, for example, don't clearly explain what's going on, then that source code is completely useless as a reference. (Also, the board of appeals and other courtrooms will disregard the source code. It would take a defendant in a litigation proceeding to actually compile the code and demonstrate, for the judge, what is going on.) 3) Source code is rarely conducive to effectively teaching claim limitations. "Encrypted network connection to communicate user demographics data" is part of one sentence in a normal reference but could cover 1,000 lines of source code. Using that source code is an extremely dismal proposition.

Dated, publicly available manuals is the key. Examiners love pdfs of what features were supported in what versions and when those versions were publicly available.

Well, the point was that the GPL simply does not need patents to get the benefit of public disclosure.

True, but the GPL software needs clear, concise, dated disclosure to enjoy benefit of doubt from the patent system. This is a situation where patent examiners would love to use GPL software, but doing so without shooting themselves in the foot requires something ELSE from GPL projects, something that is not often provided or readily available. By no means are examiners against using GPL software, it's just that GPL projects are very rarely documented adequately enough to be useful to the patent system. Hm... I smell a lucrative career as a consultant...

By the way, here's an excerpt I found from a Bruce Perens open letter (reprinted with explicit permission):

We [open source software and community] are especially threatened by royalty-bearing software patents that are embedded in industry standards. In many cases, it is impossible to achieve compliance with a standard without infringing upon the patented algorithms that are specified by that standard. Standard compliance is critical for interoperability, and thus software patents in standards can make an un-communicating island of a Linux system. For example, the IEEE 1394 FireWire standard is encumbered by patents that apply to the software interfacing to it, and a patent royalty pool is operated in connection with that standard. Linux implementations of FireWire are potentially infringing, and prosecution could result in our software becoming legally unable to access FireWire devices.

With all due respect, what a bunch of crap. So FireWire is a patented interface - big freaking deal. Here's an idea, don't support a patented interface at all. The minute you try to endorse a patented interface with some work around, you contribute to its proliferation and acceptance.

If someone patents a bolt and wrench design, the arguments that you can't make your own wrench to interface with that bolt without infringing the patent would be absurd. Seriously, "I can't interface with patented Bolt without infringing the desing of patented Wrench," is a testament to the fact that it's patented, not a problem with the patent system.

Only because this issue has to do with computers, and everybody likes free stuff on computers, and many people are used to free stuff on computers is this an issue. They have this great agenda to push that linux should be the next great thing that supports everything and everybody's digital camera, and they need to interface with FireWire to get there. It looks to me like FireWire is that patented bolt, and supporting that patented bolt in your open source software is just like endorsing the patented bolt itself. Hey, maybe if that bolt was really unpopular, it wouldn't be an issue? Nah, that's crazy - go ahead and make sure that patented bolt is super popular and support it anyway you can.

Just make sure you can pin the blame on somebody else when you're sued for infringing on that patented wrench.

Honestly, the heroes of OS software are the most incoherent voices in this argument, and it's NOT because I don't understand software. Quite the contrary, I'm educated in and worked in software. The problem isn't a lack of education or ignorance on the part of the PTO, but rather a failure of the OS community to educated itself regarding patents and a refusal to recognize software as the work of manufacture. (Hell, some people are still so ignorant that they believe the USPTO patents algorithms, which is as true as 1=0.) You can't reach people that are deliberately ignorant. They're determined to be really pissed off and the world will be content letting them be really pissed off long after they're forgotten.

Becoming educated and adapting their philosophy will return them to their position of marketable flexibility and creative dominance. If the Open Source community required clearly dated, readily accessible, coherent, extensive documentation in order to "participate" (at FreshMeat or to use the GPL or whatever), and people used 37 CFR 1.99 (3rd party submissions of prior art for pending applications), and somebody had some non-retarded input regarding case law and legislation, we would be seeing a lot of problems getting solved. Instead, we see a lot of deliberately ignorant people getting pissed off while the clock is ticking.

A few notes on software and patents
Authored by: bart on Friday, December 16 2005 @ 10:08 PM CET

A bit late, but this is asking for a reply still:

Unfortunately, that is not sufficient to use for prior art. Here's why: 1) The examiner needs to figure what the reference is doing in about 10 minutes. Sometimes he's concerned with the details of implementation, but sometimes he's concerned with the user interface. A CVS is an extremely inefficient way of getting documented evidence of the big picture. 2) He needs to present the reference with a persuasive argument to an attorney. Source code accompanied by the allegation that it does what the examiner says it does is not persuasive. The attorney is under no obligation to compile the code and see what it does - he is merely under the obligation to read the code and see what it says. If the comments, for example, don't clearly explain what's going on, then that source code is completely useless as a reference. (Also, the board of appeals and other courtrooms will disregard the source code. It would take a defendant in a litigation proceeding to actually compile the code and demonstrate, for the judge, what is going on.) 3) Source code is rarely conducive to effectively teaching claim limitations. "Encrypted network connection to communicate user demographics data" is part of one sentence in a normal reference but could cover 1,000 lines of source code. Using that source code is an extremely dismal proposition.

There is one very basic issue with your reasoning:
The legal as well as the patent systems serve society and not the other way around. In the specific case of the patent system, it serves to 'promote usefull inventions' to be exact. In order to do so, it must use the language, methods and standards that apply to the inventions it is trying to promote.

If patents on software are allowed at all, the type of documentation that gives indisputable information about the claimed invention is the source code and NOTHING else then the source code. Almost any software patent that does not contain source code implementing said invention fails to forfill one of the basic requirements, that a person skilled in the field of the invention can reproduce the patented invention based on the claims in the patent. The target audience as stated by the law, and the only one that makes any sense logically, is a person skilled in the field, and while an attorny can definitely also be skilled in the field of the specific invention, attornies are NOT the target audience of patents, and if you are trying to argue that in the current system they are the target audience, then that is just more proof of the current system being broken by failing to forfill its purpose.

Then:
With all due respect, what a bunch of crap. So FireWire is a patented interface - big freaking deal. Here's an idea, don't support a patented interface at all. The minute you try to endorse a patented interface with some work around, you contribute to its proliferation and acceptance.

A small reality check might be in order here... If some 90% of the computers come with a specific piece of hardware which people also want to use, then not supporting that piece of hardware makes your software undesirable. Arguing that this is the other way around, esp. in this specific case, shows a serious lack of insight into how the market for commodity software and hardware works.

If someone patents a bolt and wrench design, the arguments that you can't make your own wrench to interface with that bolt without infringing the patent would be absurd. Seriously, "I can't interface with patented Bolt without infringing the desing of patented Wrench," is a testament to the fact that it's patented, not a problem with the patent system.

If that means you cannot make a specific wrench that 'interfaces' with a specific bolt and that combination has an outstanding advantage over any other combination then that is an acceptable situation for a limited time. If it means you cannot make a wrench in general then it is absolutely unacceptable.

A patent that rewards its inventor while creating oppertunity for others (and the inventor itself) to improve on the invention fits the goals of the patent system as allowed by the constitution. A patent that creates a monopoly in a specific market (as opposed to for a specific technique) however conflicts with the purposes of the patent system and is undesirable from a free market point of view.

Only because this issue has to do with computers, and everybody likes free stuff on computers, and many people are used to free stuff on computers is this an issue.

It has nothing to do with free stuff. If you care to look around, you will see that many small comapnies that produce software, and in theory should benefit from a WORKING patent system are opposed specifically to patents on software. You will notice that many individual developers are opposed to patents on software. You will also notice that among the proponents of patents on software you will find a few huge companies and companies that do not produce much and live from buying and then licensing someone elses inventions.

The arguments against are really simple:

  1. There is no proof whatsoever that allowing patents on software promotes usefull inventions in the field of software while there is a substantial number of examples of them hindering usefull inventions in this specific field. Because of that patents on software fail to forfill the primary purpose of patents.
  2. Software is not like manufacture at all. Software is design, and is already properly covered by copyright. THe manufacture part of software is a mere matter of copying a number of bytes to some kind of medium. Software can also be a tool for manufacturing other things, but talkign about manufacture of software is simply nonsense.
  3. The first person to implement an algorithm in software is seldom if ever the first person to think it up, rather, it is the first person to have an actual need for it or often the first person havig access to a system that makes said algorithm feasable. Also, many such algorithms are discovered or are the direct consequence of laws of physics and mathematics. Seldom are such algorithms invented.
  4. Designing and creating software may take a lot of efford and time, it seldom takes a huge investments in other resources, which makes this an activity that can easily be done by individuals and very small companies. Those will automatically get protection for their specific implementation through copyright. Adding patents to the mix excludes those and completely shifts the balance towards the few big players in the market, which again undermines the free market and discourages instead of encourages usefull inventions.

Because of those things, software patents do more damage then good, regardless of if the PTO has a clue.

They have this great agenda to push that linux should be the next great thing that supports everything and everybody's digital camera, and they need to interface with FireWire to get there. It looks to me like FireWire is that patented bolt, and supporting that patented bolt in your open source software is just like endorsing the patented bolt itself. Hey, maybe if that bolt was really unpopular, it wouldn't be an issue? Nah, that's crazy - go ahead and make sure that patented bolt is super popular and support it anyway you can.

Just make sure you can pin the blame on somebody else when you're sued for infringing on that patented wrench.

No, the patent system should encourage peopel to do usefull inventions, not waste their time with legal bullshit.

Honestly, the heroes of OS software are the most incoherent voices in this argument, and it's NOT because I don't understand software. Quite the contrary, I'm educated in and worked in software.

Well, for one you still don't understand that this is a matter of belief in two mutually exclusive ideas. Not only does the GPL have no need for patents, the very nature of patents conflicts with the nature of the GPL.

You should nnot be surprised that those who strongly advocate the GPL also have an issue with a system that conflicts with theirs.

Then, with all respect, your opinions about software and how it is created and should be treated sound about as uneducated to me as Bruce's argument seems to sound to you, and to me it reads like two different people arguing something based on belief and with disregard of most arguments.

The problem isn't a lack of education or ignorance on the part of the PTO, but rather a failure of the OS community to educated itself regarding patents and a refusal to recognize software as the work of manufacture.

As long as you argue that it is a work of manufacture that to me and to many others is a clear sign of being ignorant with regards to the development of software. Creating software is a design challange, not a work of manufacture.

For what I know the USPTO also argues that it is a work of manufacture, which to me shows they really don't know enough about it either.

(Hell, some people are still so ignorant that they believe the USPTO patents algorithms, which is as true as 1=0.) You can't reach people that are deliberately ignorant. They're determined to be really pissed off and the world will be content letting them be really pissed off long after they're forgotten.

No, in the case of software, the USPTO grants patents on expressions and implementations of such algorithms, not on the algorithms themselves. Because those are expressions and are covered properly by copyright already, there is simply no need for patents there. If you want to patent what is underlying those expressions then you ARE patenting algorithms.

Becoming educated and adapting their philosophy will return them to their position of marketable flexibility and creative dominance.

Uh.. no idea what marketing book you grabbed that nice expression from but really....

'return to their position of marketable flexibility' ????

wtf? there is nothing to return to here because that was never lost. There is no problem with the marketability or flexibility of open source software. There is the problem of an increasingly dense minefield called patents choking the possibility to create open source softwsare, despite its marketability, or maybe because of it, seeing who the proponents of software patents are.

If the Open Source community required clearly dated, readily accessible, coherent, extensive documentation in order to "participate" (at FreshMeat or to use the GPL or whatever), and people used 37 CFR 1.99 (3rd party submissions of prior art for pending applications), and somebody had some non-retarded input regarding case law and legislation, we would be seeing a lot of problems getting solved. Instead, we see a lot of deliberately ignorant people getting pissed off while the clock is ticking.

Why should they? When the government or on its behalf the USPTO, grants a temporary monopoly to a company or individual, they better make sure there is a good reason to grant this because in general doing such a thing damages society. Proper care in this is the task of the USPTO, and given the reasoning behind the patent system, it is silly to demand from inventors to spend more and more time on legal documentation and research because that discourages usefull inventions (and it gives a huge advantage to large corporations at the expense of individuals)

Also, following your own bolt and wrench reasoning, why should OS developers support the patent system and thereby legitimize it further while they in fact believe it is absolutely ndesirable because it conflicts in concept with the syetem they believe works best?

Bottomline, patents are not a goal in themselves, they are a tool for promoting usefull inventions to the benefit of society as a whole. When the act of patenting something conflicts with those goals, a patent should never be granted. Because of the nature of software, which is different from physical things in that the efford is in design and not in manufacture, combined with the fact that the 'efford' part is already protected by copyright, software cannot be treated like a typical work of manufacture, and there is ample reason to assume patents on software are damaging, while there is no evidence of the opposite.

A few notes on software and patents
Authored by: bart on Monday, July 03 2006 @ 06:50 PM CEST
I was reading this discussion again and one thing occured to me..

Nobody would seriously argue that a car is just a temporary arrangement of molecules that exist in nature, yet people DO argue that software is just a big number or an abstract algorithm. Of course this is technically true, but it's an absurd argument. I'm not posting this with my abstract algorithm machine - software is, in many ways, a manufactured product like a car.

I always felt that this argument is flawed, but so far I found it hard to express why exactly. In the end it seems simple however.

To me it seems that a manufactured product suggests that there is a blueprint that gets used to create a product with some manufacturing process.

The equivalent step in software engineering is compiling your source code.

Now, the problem with looking at software as a manufactured product is that you normally don't repeat this 'manufacturing' step over and over, rather, you can just directly copy the end result of the manufacuring process. For all practical purposes, software is like a cloned product (something which has no equivalent yet in the physical world). The protection it gets is covered by copyright which matches well how the product is created.

Unlike you suggest, the problem with software engineering is in fact implementation, ie, transcribing an idea into practical source code that not only works in theory, but actually compiles, works also in odd corner cases, deals well with the unexpected, and at the same time is efficient.

The result of the implementation step, the source code, compares very well to having created a blueprint for some manufactured product, and can be protected in the same way (copyright, trade secret). The complexity of this step is usually quite substantial for software, often far more so then when designing a physical device.

So, the difficult step in creating software that should be considered for protection is going from idea to implementation. If an idea is patentable in general, then the fact that it is implemented on a computer should make absolutely no difference, and should not be seperately patentable as such.

I'm not posting this with my abstract algorithm machine -

But you are, or more precisely, you are posting from your general purpose computer, which is indeed a device that in most typical uses applies abstract algorithms to some data and stores or somehow communicates the result.

See, software may be a product in itself (it can be boxed and sold in a shop etc), but it is in fact only a component of another product, lets call it 'electronic office' or 'entertainment center' or whatever other purpose you have in mind for your general purpose computer and some piece of software. You who start a program on your general purpose computer are the one who 'manufactures' the end product right there and then.

Why does that matter? It shows how the issue of patentability is still a rapidly moving target for software-related inventions. Some patents have been issued that will be invalidated with more recent court decisions. The patents issued in 2006 will have different requirements to pass than those issued in 2004. As time goes on, the case law will get better, the board and the USPTO will know how to enforce that case law consistently, and the wackiness will die down.

And meanwhile those who are actually trying to invent things get bogged down by all this unclarity, and the risk of ending up in a patent minefield.

This situation is typical for new fields of technological development. What is not typical is that software was initially not patentable at all. The effect of this is clear when comparing rate of early development of software to the rate of development of electronics. While the later field develops quite rapidly at the moment, it has had very big and important inventions being hindered and forestalled for decades esp in the earlier years of development. One of the better documented cases involves the electronic video camera, RCA and a guy called Fernsworth.

Another quite classic example of a patent hindering a new field of technology a lot more then help it, dates back a lot further in time. We have to go back to the Netherlands in the second half of the 1500s (then still a bunch of provinces in the Spanish empire, but at the brink of an independence war).

A guy who called himself Leeghwater (empty water) had invented a way to drive large pumps with a windmill. He then went on and became rich and famous thanks to his ability to claim land from the sea (and more often, internal lakes) with help of his invention. He also obtained a patent for his invention.

Oh, and it gave the Netherlands one of its symbols of course.

Another guy (whoms name I forgot) came along and invented a way to use Leeghwater's invention to drive interesting devices like multiple synchronized saws and a practical wind-drivensawmill was born.

Due to Leeghwater and his patent (and his unwillingness to license it to what he considered to be an unworthy farmboy, the guy who invented that sawmill), the inventor of this sawmill could not make practical use of it in the territories where Leeghwater's patent was valid and enforcable, so after lots of attempts to negotiate a deal, he left for what is now called the 'Zaanstreek', an area in the north west of the Netherlands, where the patent was not recognized and set up shop there. What resulted is history. With help of wind driven sawmills, wood could be processed at a speed and accuracy never dreamt of before, and the area became fundamental to the Netherlands becoming a maritime superpower a bit later. (one can make an interesting comparison to what the steam engine did a few enturies later on a much larger scale)

Now, this 'escape' allowed this development to begin decades earlier, and that in turn played a substantial role both in financing and providing war equipment to fight a 80 years war of independence.

In the meantime, the southern parts where the patent was enforcable, quickly lost its initial position of being the rich and more powerfull part of the territories (and later country)

No doubt his patent served Leeghwater well, but is rather doubtfull that he would have been off much worse without it because his abilities were a lot more into how one would use this invention to get the water out in a practical way)

Of course it eliminated any competition in that specific area, which definitely made things easier for Leeghwater. The price for that however was that any territory that recognized his patent was outcompeted in all other markets where his invention might be usefull untill some 2 decades later, and if it wasn't for some territories not recognizing this patent, not only would another inventor have been left witout any incentive for the extremely usefull invention he did, the country would not have been in any position to try to fight the Spanish empire for its independence either.

A few notes on software and patents
Authored by: bart on Saturday, March 11 2006 @ 04:53 PM CET
Another interesting and imho good discussion on the subject of patents can be found here.
 Copyright © 2014 Bart's Soapbox
 All trademarks and copyrights on this page are owned by their respective owners.
Powered By Geeklog 
Created this page in 0.33 seconds 
Powered by Apache Powered by FreeBSD