Talk:Software patent debate/Archive 1

From Wikipedia, the free encyclopedia

Archived 29 March 2007

This is an archive of past discussions. Do not edit the contents of this page. If you wish to start a new discussion or revive an old one, please do so on the current talk page.

Violent Clean Up

I am performing a violent clean up of this rambling page. My goal has been

To focus the page on the debate itself. Information on patents in general, their history etc. goes elswhere.
To provide a structure for the discussion.
To remove rambling repetition, of which there was much.

I think that the work is now fairly clean, please take care.

Aberglas 08:17, 4 Feb 2005 (UTC) aberglas

I think its time for another clean up. I propose that "arguments for" and "arguments against" be simple statements of arguments with citations of sources. No citation, to argument. I also propose that we stick with affirmative statements and avoid point / counterpoint. I'll try two as examples and see how other's feel.--Nowa 00:56, 30 September 2006 (UTC)

I totally agree. The article has just become a mess with too many people doing nothing more than voicing opinions. I've been watching this article recently with an eye to doing something with it, but I just can't see how to begin. Best of luck Nowa! GDallimore 01:28, 30 September 2006 (UTC)

Nowa, I've made a start in trying to tidy up the "arguments for" section - I can't face the even more fractious and lengthy "arguments against" section. I kept the point/counter-point style since I thought it nicely illustrates the lack of consensus. However, if you do decide to try to change this style, hopefully my amendments will make it a bit easier for you; I think I've removed most of the repetition and the statements that are more blatantly based on personal opinion. GDallimore 11:33, 7 October 2006 (UTC)

What if we set up a table with point - counterpoint? Point on the left, counterpoint on the right?--Nowa 21:28, 18 October 2006 (UTC)

Purge of Stubs

I have now purged my stubs out of the discussion entirely.

Aberglas 08:17, 4 Feb 2005 (UTC) aberglas

What "inventions"?

"The topic of granting patent protection to software inventions"

This is biased wording. According to EPC 52 Software there is nothing such as a "software invention", it is not regarded as an invention (as such)

There is a two steps question:

1. Are software (ideas, concepts, generalisations, funaction) an invention 2. Shall patents be granted if we agree that there is an invention ... etc.

When you write "software invention" you already agree that there is such a thing as a "software invention".

What I also miss is the Economist's test for applying patent law.

Disputes

There are quite a lot of disputes

legalistic

a) Is software an invention?

  Is Software technical?
  Is data processing a field of tehcnology
  Shall program claims be admitted?

b) are there international obligations or restrictions

normative

a) Do software authors need additional patent protection b) Cui bono? c) Who is in charge: The patent institutions or the lawmaker d) What abour Venture capitalists

economically

a) Shall patents be granted for services as well b) Two scenario model

  b1) patent law in the field of software
  b2) no patents for software (Free market)
  b1 better b2? For whom? Possible Winners and losers, compensation ecc.

c) Are the effects attributed to patent law working in the field of software

empirical

a) US patenting as a success model?

Stakeholders

Patent attorneys
Patent institutions and institutionalised interests
Software authors
SME
US corporations
lobbying groups
consumers
economists
politicians

It would be important to split these interest groups and analyse their stake in the debate.

I would prefer to split the questions and rather create a pro-con article

Answers to objections

These studies are generally written by economists that do not understand patents or copyrights.

I find this sentence particularly stupid. Sould it stay in the article ? Peco 07:27, 28 Mar 2005 (UTC)

Patent research cost

I think the article forgets perhaps the most important argument against software patentability: The arguably prohibitively high (at least for SMEs and independent developers) cost of the study to ensure that the produced, or about-to-be-produced, SW does not infringe any existing patents.

However I think this also needs an counterargument (I assume one exists) before it can be added to the page. Anyone with knowledge?

NPOV doesn't mean that there has to be a counterargument for every argument. However, what we really should have is a source that can roughly quantify the cost of a patent search. Rl 07:07, 1 Apr 2005 (UTC)

Why Wikipedia Community doesn't make any actions against software patents?

Moved from article to talk page. Not a fact, just a signed question.--Edcolins 11:53, Apr 7, 2005 (UTC)

Well I think at least some proper banner have to be on Main Page.--Popski 11:37, 7 Apr 2005 (UTC)

http://www.ljudmila.org/patent_overload_general.gif

Because the Foundation exists to provide resources for and promote Wikis and open content. It isn't a legal advocacy organization. You're looking for the Electronic Frontier Foundation. --Slowking Man 06:24, Apr 10, 2005 (UTC)

Overcomplicating products for marketing purposes

If there is a solution for something that is so simple that it cannot be exploited economically by a company, it might market a more complex and more expensive method. I suppose this is common in all industries. Now, if a company patents such an idea, does this mean that a simpler method can no longer be offered because the patent is about the function and not the implementation? DirkvdM 09:41, 2005 May 6 (UTC)

I don't really get your point. Why would something simple not be economically viable?

An extreme example is when a solution to a problem is simpler than a specific product, eg certain stains can be removed with (a combination of) simple household items, but that's an idea, which can't be sold, so a producer would rather bottle that mix separately, so they can sell it (in stead of tell it). Another extreme is that a couple of Dutch inventors reinvented the wheel in a much simpler form. The tv show that presented this invention (a few decades ago) said "if you never hear form this again, that'll be because it's been bought by Michelin or some other company, not to use it but to prevent it from being used, because it's a much cheaper wheel, which will decrease their turnover (and, consequently, profits). And then there's standardisation. That simplifies matters for the consumer, but also means that if you buy a product from producer A (say, a printer or camera), you'll be able to combine it with products form producer B (ink or memory cards), so A prefers to make the 'secondary product' (what should I call that?) proprietary, forcing buyers to stick to their products. You'd think that the companies (if any) that follow a standard will win, but Sony's memory sticks are still around.

In software something similar exists with those little tools that abound in the Open Source community, which are too small to be marketable. So a commercial company will rather make a complete set of such tools, which they can then sell that as a program, which includes all sorts of things you'll never need. DirkvdM 09:30, 2005 May 7 (UTC)

Well, it seems this issue is more about marketing than patents... --Edcolins 17:17, May 7, 2005 (UTC)

Let me specify; suppose someone discovers a cheaper way to perform the function of a memory stick (they're very expensive). To make this marketable they'd have to use the same interface. If the memory stick is patented, then would that still be possible? This would be a specific example of the general principle that a company can come up with an expensive solution, patent that and thus prevent the emergence of a cheaper solution.

Notice that, here, I've moved away from the original question to something related: Does a patent for something specific automatically include a patent on an aspect of it that should be a standard and therefore not in the hands of some private company?

Now something closer to the original question: If someone comes up with a little program that does the same as one of the tools of the Norton Utilities suite. Would that then infringe on a (possible) patent by Norton? The reason I ask this is that in the software patent article it says that a patent "covers the programming method itself, independently of any implementation in code. Thus usually reimplementing a program will avoid copyright infringement, but not patent infringement." DirkvdM 07:03, 2005 May 8 (UTC)

Regarding your second question, it depends on what has been invented and on what is claimed in the patent. Is this the simple solution, the complex one or a solution which is to be found in both the simple and the complex solutions? Also, is the simple solution known in the art? --Edcolins 12:22, May 6, 2005 (UTC)

Well, my question really is what is patentable. I understood that one patents the idea of a solution, not any particular solution. Which would mean that, even if there is no solution yet, any solution falls under the patent. So my question really is whether the assumption is correct (and therefore the conclusion). I believe prior art has nothing to do with that because that's about copyright and patents go above that, right? DirkvdM 09:30, 2005 May 7 (UTC)

Ideas and ideas of solutions are not patentable. Inventions are (but not all inventions). Under European patent law, an invention can be viewed as a technical solution to a technical problem. If there is no solution yet, there is no invention, and then there cannot be any patent (or if there was one, it would be invalid). In order to obtain a valid patent, it is not sufficient to have an idea of a solution to a problem, you must make a workable solution and you must describe it in a manner sufficiently clear for the person skilled in the art to practice it. Not all inventions are patentable, only those which are new, involve an inventive step and are susceptible of industrial application are. --Edcolins 17:12, May 7, 2005 (UTC)

Please Beware NPOV

Some time ago this page had degenerated into a long rample. I had then restarted this page with a relatively NPOV analysis, trying to keep the number of words for and against roughly equal.

I must admit that some of my arguments were a little cheeky, but they largely repeated what others had said. If you read my last edit carefully you will know which side of the debate I am on, but it should not be obvious. [1]

I am concerned that the article is again degenerating into an obviously anti S/W patent diatribe. Keeping it succinct, sharp, and as NPOV as possible will make a much better and more useful, quotable addition to the debate.

Thus I believe that tighening it up is the key. It must not get any longer. I also think that the arguments should be kept to just two levels. Less is more.

As to the new table at the front, in its current form it is distracting because it duplicates to some extent the body of the article. If you really want to add a table then you need to do a good job of integrating it with the whole article, not just plonking it in front. That sort of thing is what happened last time. (My feeling is that a simple format will do for this article.)

So I think that a deletion/clean up phase is really required. Aberglas 05:12, 16 Jun 2005 (UTC)aberglas

The claims of a patent clearly define what is being protected

For varying degrees of "clearly". I should patent a software implementation of the patent process. Then I could sue the government for infringement. But then again, why would I waste my time making a program intended to waste my time, without it even being entertaining. -- Bob the Cannibal 1 July 2005 22:05 (UTC)

Economic Overview

I've reverted the table, because I think it's helpful for the article which can arouse such partisan feelings to start out by setting the stage with a NPOV quote from one of the leading academics who has done research in this area.

I also personally found the separation of economic consequences into the different headings - innovation, competition, transaction costs - useful in clarifying my thinking.

In response to the allegation of bias by the person who proposed to cut it, I'm slightly at a loss - do you think it is biased pro-patent or anti-patent. (FYI, the article the table originally appeared in was a fairly trenchant criticism of business method patents).

The point I would make about the table is it is a list of some of the factors to look for in /any/ industry as a result of patentability. There are some industries, for example the classical one patent = one product high development cost industry, where arguably the positive factors on the left are particularly strong, the negative factors on the right particularly weak. This tends to be the archetype that most legislators start out, which is why if you believe software patents are a bad thing, you need to explain what it is about the software industry which makes the "classic" positive factors on the left weak, and the negative factors on the right strong. (The FTC report is a particularly good for this).

Aberglas wrote above:

I don't agree. I think the table is helpful short intro teaser to give readers a structured and useful compass on the kind of factors to look out for, which is useful before they are thrown into the full mass of contending opinions in the next section.

I also restate that I think it's useful to start with an abstract of the factors to look out for in /any/ industry as a consequence of patentability, before getting into the detailed specifics of the software industry, and the discussion as to how relevant or irrelevant each of those factors actually are for software. -- Jheald 23:02, 9 July 2005 (UTC)

Quotes

Shouldn't the second Gates quote be considered a quote against patents? It sounds like he's saying that the current system allows large companies to prevent anyone from entering the software market without a sizeable patent portfolio, and that can't possibly be good.--Jdcope 11:39, 19 July 2005 (UTC)

Is there such a thing as a software invention?

I am a US patent agent specializing in software and business method inventions.

One of the biggest challenges for both myself and my clients is to determine where exactly the "patentable invention" is when they come up with a new idea for software. Very often they think it's the basic idea they came up with. This is rarely the case. A little bit of prior art searching usually shows that others have thought of the basic idea in the past. The conversation then moves on to what inventor had to do to make the idea work in a practical sense. That's usually where the invention lies. --Nowa 21:51, 23 October 2005 (UTC)

Cleanup needed again

I have added the "cleanup-verify" tag since many arguments in the sections "Arguments for patentability" and "Arguments against patentability" lack source citations. This article starts to look like a talk page. --Edcolins 08:29, 10 October 2005 (UTC)

Agreed. Added some references and did major pruning of unsupported assertions in "Arguments for patentability".--Nowa 01:27, 19 October 2005 (UTC)

Reverted. There's a lot of sense in a lot of these arguments, on both sides, and they give a reasonable flavour of the "debate". If you think that individually particular arguments can be cleaned up, or better grounded with references, then do so. But just deleting the lot, en masse indiscriminately, is mere vandalism. Jheald 15:17, 22 October 2005 (UTC)

Good point. I admit I was getting a little carried away. It's probably best to take them one at a time.

I presume that we only want to show arguments that have some support as being valid arguments. If not please let me know. --Nowa 23:01, 22 October 2005 (UTC)

Patents not about R&D ?

I deleted the point that "Patents are designed to protect manufacturing processes, not R&D" because to the best of my knowlege, I don't know anyone in the patent business who would agree with it. (I am a registered US patent agent specializing in software and business method patents. Prior to that I was an R&D engineer at a major chemical company where my inventions earned 17 patents). On the contrary, patents are designed specifically to protect the inventions that result from R&D. For example, from Pfizer's recent annual report, their general counsel, Jeffrey Kindler stated "patent laws exist to support and encourage medical innovators". I could put this in as a counter point to the assertion that patents are not intended to support R&D, but since I am unaware of any support for this position, I think the article would be best that it be left out or at least reworded to be more accurate. If I'm mistaken, please let me know.--Nowa 23:01, 22 October 2005 (UTC)

A point which is maybe worth considering is that protection may be needed to reassure an inventor he will be able to recoup all of the costs of bringing an invention to market -- eg retooling a manufacturing line, establishing product awareness, etc -- not just the R&D. Otherwise the invention will have no value to him. This I think is a point worth bearing in mind, and maybe worth making somewhere; though not necessarily just here. Jheald 11:31, 23 October 2005 (UTC)

Excellent point. Any ideas where it should go? How would we support it?--Nowa 12:03, 23 October 2005 (UTC)

Possibly the deleted statement/s could be re-edited, so it was more clear that this was the point being discussed. But I don't have any particular wording to mind ATM. Jheald 12:43, 23 October 2005 (UTC)

Another point that maybe should be considered is that in software context, "all costs of bringing an invention to market" may be just R&D and distribution. If efficient method is used for distribution, the distribution cost is very close to zero. However, currently, there are all kinds of barriers to companies using this method, patents are just one of those. It seems this is one source of inefficiency. esap 14:36, 23 October 2005 (UTC)

esap, What do you mean by "this method"?--Nowa 21:50, 23 October 2005 (UTC)

I mean the method of developing software where R&D and distribution are the only things needed to take the software to market. esap 17:53, 24 October 2005 (UTC)

All software is obvious

In the European patent office, all pure software inventions are rejected as being "obvious" (i.e. no inventive step). I think this is worth mentioning as an argument against patentability. Any seconds? Any counter arguments?--Nowa 21:55, 23 October 2005 (UTC)

It's not being rejected as "obvious" (art 56 EPC alone), it's being rejected as "no inventive step" (art 56 + art 52 together) -- ie the advance on the prior art may be non-obvious, but it's not patentable subject matter. There is a difference.

All obviousness objections are raised under Arts 52 and 56 together, whether they concern CIIs or not- Art 52 says that to be patentable an invention must involve an inventive step, and art 56 tells you what an "inventive step" is. Specifically, art 56 says that having an inventive step means "not obvious", so as far as I can see, "Inventive step" and (non) obviousness are synonymous.

I'm not sure I understand Nowa's comments. The Guidelines for examination at the epo suggest rejecting unpatentable CIIs as lacking an inventivestep since this is generally more procedurally efficient- there is generally less scope for argument. This practice certainly dooes not suggest that the EPO think all pure software patents are obvious- the objection will only be raised on the basis of a properly argued obviousness objection based on the "problem-solution" approach, with reference to the closest prior art. If no such argument is plausible, no such objection will be raised. ~~M Bell

(See patentable subject matter for more detail. A succinct summary of the U.S. position after State Street and Ex Parte Lundgren for that article would be very much appreciated: discoveries not being inventions, "everything under the sun made by man", etc). -- Jheald 22:29, 23 October 2005 (UTC)

seebusiness method (patent) for succinct treatment of ex parte lundgren--Nowa 22:55, 30 October 2005 (UTC)

Microsoft patent application

I wasn’t sure what the prior editor’s point was regarding Microsoft’s pending US patent application on multiple desktops [2]. Microsoft does make mention of KDE and GNOME prior art but has since successfully convinced the patent office that their particular invention would not be obvious to a person of ordinary skill in the art even in view of the KDE and GNOME inventions. You can see the exchange between Microsoft and the patent examiner here: [3] Download the “notice of allowance” to see the reasons why the examiner allowed the claims.

It seems to me that the Microsoft example refutes the argument that software patents only cover trivial extensions of existing art. I don’t think, however, that that was the point the author was trying to make.--Nowa 23:25, 17 January 2006 (UTC)

Nowa can you defend that unsupported statement? I looked at the patent. It is all very ordinary virtual destop technology.

The fact that this Microsoft patent application uses ordinary virtual desktop technology is irrelevant (I don't actually know whether your unsupported statement that it is ordinary is true, by the way, but it does appear to be that way). Rather, it seems that the "invention" Microsoft are claiming is the USE of known techniques in a NEW way. The question the USPTO are seeking to answer is whether this new use of the underlying technology is obvious.

Few software patents claim some new algorithm or programming technique. Most of them are about applying old ideas in a new way. This is why putting source code in software patents would usually be of no benefit to anyone. Actually getting the computer to achieve the desired result is probably easy once you know what the desired result is. What the Patent Office are there to decide is whether or not it would have been obvious in the first place to want to program a computer to achieve this new result.

It's also important to remember that the link provided by Nowa only goes to the original application - NOT the granted patent, which is here. Claim 1 is quite a bit longer in the granted patent than it was when the application was originally filed. Be careful not to form opinions on software patents when you've only read the original software patent application

Having said all that, it doesn't look like much of an invention to me. BUT it is vital when having a debate to understand WHY the patent was actually granted. It was not granted because the examiner did not realise that virtual desktops were around before this patent application (this is what your comment seems to imply). It was granted because the examiner thinks that this particular use of virtual desktop technology is novel and inventive and useful. GDallimore 11:16, 29 September 2006 (UTC)

GIF

A prior editor made reference to the fact that the Unisys and IBM patents on GIF precluded the use of these inventions in free software. See the GIF patent issue. The reference points out, however, that the last of the gif patents will expire in Feb 2006. At that time, the subject matter disclosed in these patents will be freely available to the public.

It's nonetheless a good example of how patents that are disclosed too late can cause problems, because the existence of the patent was understood by the development community many years after the technology had been in universal use. It's unlikely that it would ever have gained that level of use, if information about the patent had been available earlier. esap 00:06, 23 January 2006 (UTC)

- Esap, I think you "hit the nail on the head" as far as the big problem with patents for software inventions. The patent process is orders of magnitude too slow to be of value to the software industry. Disclosure and examination should take minutes, not years. The open source approach solves the speed problem and uses a sort of natural selection approach to examination. It's difficult approach to take, however, for those with a proprietary business model. Their best approach is to keep the really important inventions secret. Do you know of anyone who had given thought to an approach that would work for those with a proprietary business model? --Nowa 11:48, 23 January 2006 (UTC)

- - No sorry, I don't have good references for that. esap 17:27, 24 January 2006 (UTC)

As a software inventor, I have given thought to the problem of trying to sell proprietary technology, as it is one I've run into myself. The most difficult part of software invention that I have found is determining that something hard can be done. If one way is found to do it, often another approach that is a transformation of the first or even totally different will do the same job, and not be covered by the patent. Often the critical innovative step seems obvious in hindsight (look at quicksort), and thus is unpatentable. Thus, as soon as a real patent (one that tells people in sufficient detail how it works) is published, it is easily reproduced and worked around. This may take both legal (with some thought) and illegal (simple copying that isn't detected) forms. This, combined with long lag times makes patents risky for real innovators, and convinced me to use trade secret law instead. A better approach: Have a patent "abstract" visible to the public that covers what the patent does only at the most basic level. Only publish this once the patent is approved. Have the inventor send in his claims, along with a full working implementation including unobfuscated source code, which get kept secret by the patent office. Have this special form of "secret" patent last for 2 years (or some comparable term) and be approved in a more rapid process, spending what is necessary to get this done quickly (higher fees?). When the patent expires after 2 years, it should be fully publicized, including the full working invention. Comparing detailed claims for infringement purposes should be handled by a trusted intermediary (the patent office?) Doing this in an efficient, inexpensive manner is the hard part of this approach. When there is infringement, it should be rapidly adjudicated in a special system, possibly a form of arbitration. A detailed set of "reasonable royalties" should be built as guidelines.

This solves: The value of software is greater the freer and more available it is; making it open source in 2 years greatly reduces the cycle time and enhances innovation. Letting the inventor make some money off his innovation, which should still be significant with a 2-year lead in software, without having to give the whole idea up on a public site. This also encourages him to commercialize it quickly.Cuberoot31 00:38, 26 September 2006 (UTC)

Bill Gates 1991

Please note that what Gates said in 1991 is misquoted here as in many other places. Find the original memo; it's named "Challenges and Strategy" and it's around. --Antonios Christofides 11:23, 8 March 2006 (UTC)

Neutrality

This article seems to have taken sides. There are more points relating to the software patent opposition, and also more quotes relating to that viewpoint. Rather than add more to the pro-patent side, someone with some expertise should think about cleaning up the article and making it more balanced (not longer). —Preceding unsigned comment added by 63.230.148.173 (talk • contribs)

I am not sure that "equal time" is really a road to article neutrality. For instance, I don't see that in an argument for and against evolution, it is somehow "neutral" to give equal space and credibility to anti-evolutionists or groups who claim the earth is only 5000 years old.

Supposedly, one of the sides in a dispute is right. A couple aspects of that rightness that we should expect to see manifested are 1) a plurality of sound (and different- economic, social, technical) arguments and 2) a large number of credible, sober parties "coming down" on one side or the other.

It is an encyclopedia's job to present the truth, that is reality, as well as it can. In some cases, reality is in dispute, true, but that doesn't mean that every holocaust denying or evolution denying band of zealots is somehow entitled to "equal time." If the arguments are more numerous for the anti-patent side, so be it.

This is not journalism attempting to please a readership who is split down the middle on political issues and can't afford to offend any special interest groups no matter how small- this is the presentation of reality in a disinterested but accurate way.

It seems the pro-patent lobbyists have taken to removing valid anti-patent points from this wikipage on a pretty regular basis. It is not acceptable to attempt to bias this page by removing appropriate material and / or substituting weak arguments.

I did remove some of the arguments in that section, including some of yours. However, I did so a carefully as possible. I deleted repeitious, non-sequitous or irrelevant points from both sides of the debate, and actually added a number of new arguments against software patents, so accusations of bias are hardly friendly or appropriate.

My main reason for removing some large chunks were where the arguments for and against were completely unrelated to the main point - eg the whole section about the clarity of the claims, to which you then added a large, somewhat rambling IMHO, chunk about micro-niches filled by software developers. With a bit of refinement, that section could have found a place in the "arguments against" section, but I couldn't find a place anywhere in the "arguments for". I actually tried to find a place for it by adding the section about the EPO only granting patents that provided a technical effect - to which, you will note, I also added two quite strong counter arguments - but it just didn't follow on in a sensible way.

I also deleted or refined some sections that led to confusion about what a "software patent" is. Under a broad definition, cds, mobile phones etc are indeed covered by software patents - they are certainly covered by CII patents and many people feel that the two terms are synonymous. You appear to have a somewhat narrower definition judging from your newly added counter argument to that section of the debate. A big part of the debate, is that nobody is quite sure what all these different terms mean. I've tried to clarify this by adding things like "pure" software patents as compared to software patents used by the manufacturing industry. I've referenced a paper in the "arguments against" section that discusses the number of software patents granted to different companies, although I probably should have referenced it in this section, too.

Your newly added comment about large companies using lots of software patents as a weapon is a good one, although phrases such as "virtually all code could be said to violate some software patent" are hardly NPOV (or correct, IMO!). However, the point had not been made in the article before I edited it so your new addition is hardly repairing any damage done by me. The same can be said of your new argument about the US having market dominance.

I do hope you don't simply undo the edits I made due to bitterness about my deleting some of your comments. As I said, I think they could find a prominent place in the arguments against section - perhaps some section about the problem with software patents being seen as obvious by programmers. GDallimore 22:27, 7 October 2006 (UTC)

This quote

R. Buckminster Fuller 1939

"The patent files are glutted with relative nonsense." (Fuller, R. Buckminster, Nine Chains to the Moon, Doubleday Anchor, 1971 p 277)

Can the above quote really be considered against software patents? He could very well be opposed to software patents if he were still alive but I don't think we can discuss whether he would have been opposed to software patents based on this quote. It could be against patents in general but that's not particularly clear. It might just be the case that Fuller felt there were major problems with the way the patent system was working at the time allowing lots of stupid patents but that he was not opposed to patenting in general.

Also, I don't think there issue of whether the patent system on a whole is working well or has major problems should be touched in quotes. It might be fair to state the opinion (if it can be sourced) that there are major problems with the patent system and that these problems are going to have an ever bigger/worse effect when it comes to software in the arguments section however. Nil Einne 10:33, 24 June 2006 (UTC)

I see no harm in removing the quote. I don't think he was criticizing the patent office. He was a big proponent of patents and licensed his own. I think, rather, he was criticizing inventors for filing patent applications on relatively unimportant inventions.--Nowa 15:20, 24 June 2006 (UTC)

I agree. Since nobody's raised any objections, I've gone ahead and deleted the quote. I also agree with Nil Einne that quotes aren't the best method of presenting the arguments, but I'll leave that cleanup to someone else... Crispy 08:17, 29 June 2006 (UTC)

Citing Sources

A good deal of this page seems to be one sided and a good number of the arguments for both sides are lacking in sources. Also the arguments for software patents section is basically made up of arguments against software patents. -- 209.204.181.196 23:02, 24 June 2006 (UTC)

The same is true of the arguments against software patents section. It just seems to be the way someone decided to present the arguments. Crazywolf 07:27, 2 July 2006 (UTC)

The rebuttals in both sections are very enlightening; they give readers a chance to see the arguments answering each other ala the Supreme Court, rather than leaving the two sides to talk past each other. The rebuttals are clearly distinguishable. Without this format, the reader is left to try to find a corresponding argument in the opposing side's section. I don't see a lot of low-quality argumentation on either side.

Independent economic studies responses

The original economic study located at http://www.researchoninnovation.org/swpat.pdf has a response to the argument in the introduction that counters the rebuttal in this article. I am planning on revising it in one of two ways, and would like input on which you prefer. The first follows the precedent already set for presenting counter-arguments, but the latter is more concise and does not give the initial impression that the researchers conceded that their research proved nothing.

Independent economic studies have observed that a rise in software patents has not been accompanied by a rise in software R&D (see software patents).
- These studies are emprirical in nature and by their own admission do not provide a basis for causal links between patents and their effects on R&D.
  - They do refute the theory that making software patents more readily available will increase R&D spending, however.

Independent economic studies have observed that a rise in software patents has not been accompanied by a rise in software R&D (see software patents). This does not show the causal links between patents and their effects on R&D, but does refute the theory that making software patents more readily available will increase R&D spending.

Crazywolf 07:57, 2 July 2006 (UTC)

Actually there are few economic studies on this point. There is a tendency on both sides of the argument to get absolutist about this type of thing. The beauty of economic research is that it rarely leads to a straight answer, and certainly not an uncontestable law. Bessen & Hunt's research is certainly detailed, but is not without its critics. For instance, Hahn and Wallsten, in the way that only academics can, have rather torn into the research citing

"The data, analytical methods, and empirical results, however, do not support these conclusions. In this note we explain in detail the problems in the paper and how those problems prevent one from drawing any firm conclusions. Our critique is based on a concern that this unpublished manuscript will unduly influence policy decisions and is not intended to diminish the importance of the questions the authors address. We therefore conclude the note with recommendations for future research."

On this basis, we have to be careful not to make assertions based on little evidence. If this were a medical debate, would we conclude that we had the cure for cancer after reading a single article? I don't think so. There needs to be a body of research before people can scientifically establish that hypotheses are (probably) true.--Baggie 14:24, 11 August 2006 (UTC)

900 patents a day

This is probably an accurate figure since 4000 a week = avg of 800 a day. But the paragraph should still be out as it is grandstanding. --Trödel 12:52, 29 September 2006 (UTC)

Most of the statement was completely false. Patents are only granted by the USPTO on Thursdays - hence my 4000 a week figure. So saying that 900 software patents are granted "each and every day" cannot possibly be correct. Added to that the suggestion that over 100% of granted patents are software patents, and you begin to see the statement falling apart. If I recall correctly, the majority of granted patents still lie in the electrical/communications and chemical fields, so there are at most 10-20% patents granted that involve software.

The rest of the statement was, I agree, grandstanding or soapboxing, and I was sorely tempted to revert all of the additions made by the author in question. I'm looking forward to seeing Nowa's edits. GDallimore 01:52, 30 September 2006 (UTC)

IBM alone had 3,000 patents issued last year.

Look, the statement that 900 software patents are granted every day is just plain wrong, and is not supported by the website you cite. That website says that, on 3 October 2006, 803 software patents were granted. The fact that these were granted on a particular day is essential, not academic, because if you look at any other day in that particular week, you will see that no patents at all were granted. Hence, there were 800 software patents granted that week, total (not per day). Note that 800 patents times 52 weeks is about 41,000, agreeing with the projected yearly grant of software patens.

The TOTAL number of patents granted on that day (and therefore during that week) was 3792, which means that about 21% of patents granted were "software" patents. That is just above the high end of the range I gave in my previous comment, but none of my figures are "way off". In any event, the method used by the site to try to identify software patents is contested since it would include lots of patents that do not fall within many definitions of what counts as a "software patent" (see documents in Software_patent#economic_studies), so I stand by my figures as being supportable. GDallimore 09:16, 22 October 2006 (UTC)

I was not attempting to rebut the conclusion. Conducting infringement searches is difficult and time-consuming. However, a potentially valid point backed up by completely innacurate data is not suitable content for any already shaky article. My personal opinion (which is a personal opinion that I have therefore not expressed in the article) is that it is in no way necessary to review every granted "software" patent to discover whether you can do something in software or not, so the total number of "software" patents granted "per day" is not the most relevant factor to the problem of conducting infringement searches. I have seen nothing which supports the assertion that a large number of "software" patents makes specific infringement searching difficult. Again, I'm not saying that infringement searching is easy, but there are different reasons why it is difficult. One factor is, indeed, the number of patents granted, but it is more to do with the number granted in specific fields. The range of subject matter covered by this definition of "software" patent is so broad that it cannot serve as a useful indicator of how many patents are granted in more specific fields. That's like saying that because we have 210 (or so) bones in our bodies, it can be difficult to work out that we've broken our leg rather than our arm.

Note that I am still referring to "software" patents in quotes since, if you actually took the time to read the economic study articles I cite above, you will see how divided people are about the accuracy of the method used to obtain the 800 "software" patents a week figure and what this figure actually represents in terms of the breadth of subject matter being monopolised. GDallimore 15:49, 8 November 2006 (UTC)

Here is the search used by www.pubpat.org to conclude that there were 40,964 SOFTWARE patents issued in 2006.

http://patft.uspto.gov/netacgi/nph-Parser?Sect1=PTO2&Sect2=HITOFF&u=%2Fnetahtml%2FPTO%2Fsearch-adv.htm&r=0&f=S&l=50&d=PTXT&RS=%28%28%28ISD%2F20060101-%3E20061003+AND+SPEC%2F%28software+OR+%28computer+AND+program%29%29%29+ANDNOT+TTL%2F%28%28%28%28chip+OR+semiconductor%29+OR+bus%29+OR+circuit%29+OR+circuitry%29%29+ANDNOT+SPEC%2F%28%28antigen+OR+antigenic%29+OR+chromatography%29%29&Refine=Refine+Search&Refine=Refine+Search&Query=ISD%2F%281%2F1%2F2006-%3E12%2F31%2F2006%29+AND+SPEC%2F%28software+OR+%28computer+AND+program%29%29+ANDNOT+TTL%2F%28chip+OR+semiconductor+OR+bus+OR+circuit+OR+circuitry%29+ANDNOT+SPEC%2F%28antigen+or+antigenic+OR+chromatography%29

If anything, it misses some software patents. 41 thousand a year is about 800 a week- real software patents. I don't see any point in minimizing the the vastness of this problem; if a solution is going to be found to the software patent problem, then the discussion has to start with what is or is not the case. 800 a day each and every day. That is reality.

Can't believe this old chestnut has come round again. 800 software patents a day is not reality. It is a lie, or an error, based upon a complete lack of understanding of how patents are granted by the USPTO and that data provided by pubpat.

The Bessen/Hunt technique for identifying software patents, used by pubpat, includes false positives and false negatives as they themselves admit. Therefore, while it is true to say that it "misses some software patents", it also includes some that aren't software patents, so there is no way to know whether the actual figure of "real" software patents is higher or lower than this.

In any event, what is meant by a "real" software patent? Take a look at list of software patents and you'll see that there is a huge range of subject matter covered by this definition (such as search engines, video compression and playstation controllers). No person wishing to perform a specific task with a computer would have to look through 800 patents a week to see if what they are doing is an infringement of any of them, hence my point that there can be no knowledge of how much the total number of software patents under this very broad definition of the term impacts on the difficulty to perform an infringement search. Instead, it is the number of patents granted in each of the different fields of computer technology that matters.

That's a fascinating search. If anything it points to the fact that programmable machines are an integral part of a substantial fraction of all new inventions. Why, however, do you leave out "chips", "antigen" and other related terms? It seems to me that creating new systems to design, manufacture, and distribute semiconductors and pharmaceuticals [4] is a substantial fraction what software companies do. --Nowa 12:33, 17 January 2007 (UTC)

It's all in here See p.9 and the appendix.GDallimore 13:02, 17 January 2007 (UTC)

GD, Great reference. Bob Hunt and I are actually old buddies.--Nowa 18:02, 17 January 2007 (UTC)

Well this study says that software patents do NOT incentivize innovation and in fact serve as a substitute for innovation.

Categories: Talk archives