Jump to content

Talk:Software patent/Archive 1

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia
Archive 1

Major cleanup

As you can see, I have split this article which was getting really long. The resulting new articles are

I kept this one to a minimum, as factual as possible, with a listing of relevant definitions (or lack of definition), a short discussion of these definitions, links to the distinct patent regimes regarding the patentability of software and that's all for now. The article on abortion, a highly controversial (and very different) issue, was somehow my model (that one does not even contain a "debate" section). Any comments? --Edcolins 22:40, Nov 5, 2004 (UTC)

Unfortunatly the c;eanup needs a cleanup. A "means plus function" example was used to show how a patent could sneak in software. However, means plus functionclaims have a usually a speacial interpreataion, limiting them to the means disclosed in the specification. Thus unless the specification describes the use of software as the "means" it is unlikely that the claim could be read as encompasing software. In other words, "means for . . . " is a really bad example. Colmmack 00:22, 6 Nov 2004 (UTC)
Thanks for this comment about functional features and means-plus-function claims. Although US patent law does not seem to allow means-plus-function claims to sneak out in software implementations when there is no explicit support in the description, certain jurisdictions such as the European Patent Convention allow this to happen provided that it would have been clear for the man skilled in the art that a software implementation could be used, even though the description does not explicitly mention that a characteristic can be "softwarized"... I have added the limitation "under certain jurisdictions". More on this may be explain in the "Software patents under..." articles. --Edcolins 11:47, Nov 6, 2004 (UTC)
What you are here discussing is an equivalence point -- i.e., it is well known that an equivalent of a 'hard' claim element may exist in software, and thus substituting software will be considered to be infringing means. Technically, such an arrangement might also give rise to §112(6) equivalence under US law as well. Still, I would be very cautious about using a means-plus-function element as an example in this context. The key issue is equivalents and indeed, where an invention in toto is novel, holding the replacment of a hardware element with a well known software equivalent to be infringing would not seem to me to be particularly controversial. Colmmack 19:04, 6 Nov 2004 (UTC)
I understand that, under US patent law, some decisions have restricted the scope of a means-plus-function claim to the equivalents of the elements described in the description. To the best of my knowledge, this does not apply in European patent law (as far as the European Patent Convention is concerned at least). The doctrine of equivalency according to the European patent case law seems to serve only to extend the protection beyond the wording of the claim, when a structural feature is claimed. When a functional feature is claimed, the doctrine of equivalency does not apply to decide whether the functional feature is disclosed "in a manner sufficiently clear and complete" according to Art. 83 EPC [1]. Thus, it may be possible to draft a claim which encompasses software implementations without describing explicitly this type of implementation in the description (provided that it would be obvious for a person skilled in the art to trade the disclosed hardware implementation for an undisclosed software implementation). By my opinion. --Edcolins 08:19, Nov 8, 2004 (UTC)
It's somewhat more complex - there are two forms of equivalence in US patent law, the doctrine of equivalents which is seperate from §112 paragraph 6 equivalence in means-plus-function claims. The first is an essentially legal doctrine of claim interpretation, the latter is a statutory provision based on the language in 35 USC § 112 that allows means-plus-function claims ---- so its not decisions limiting means-plus-function, it is the underlying statute.
§112 equivance is essentially the same as the doctrine in European patent law -- a structural element that is replaced by an obvious equivalent to what is disclosed in the specification -- but §112 is narrowly drawn. EU patent agents fall easily into using means-plus-function language in US claims without realizing that it may limit claims rather than make them broad -- i.e., all the intended means and their equivalents need support in the specification. I belive that the same would tend to apply to a means-plus-function element in a European patent, i.e., the scope of the means would be limited to that obvious and which corresponds to the means in the spec Colmmack 19:30, 8 Nov 2004 (UTC)
Actually, from a European perspective, there is no statutory or precedent based doctrine of equivalents. Article 69 EPC (and the Protocol on Interpretation) set out the correct (and seemingly only) way of interpreting claims. Essentially, and as confirmed by Lord Hoffman in Kirin-Amgen v. TKT et al., there is no Doctrine of Equivalents in the UK, and probably also in Germany. In fact, Hoffman was quite rude about US law on this point. His point is that there is only one possible interpretation of the claims of a patent, and no amount of legal wrangling can move the boundaries of the scope of protection. This may sound flippant, but it seems to get us to a position of more certainty for third parties. The Kirin judgement is also important as it moves the burden of interpreting patents from patent attorneys to the nominal 'skilled man'. As a patent attorney, you'd think I would be against this, but I actually think this is a fairer position as it is the 'skilled person' who is best able to judge the plain meaning of terms of the claims, read in the context of the patent as a whole. Baggie 15:26, 30 April 2006 (UTC)

Hi Rl, I modified your version [2]. Neutrality is also about fairly representing all sides of a dispute, which should mean not stating only what is considered by the open source community, but also by economists, the industry, the patent profession and so forth. However, it would be cumbersome to state each and every position in this general article, while the explanations of positions are clearly belonging, I think, to the article about the directive we are talking about (which is partially the case as you can see in the "reactions" section). Stating that the proposed directive is controversial might be enough in this more general article by my opinion, and might be enough to lead the reader towards a more detailed explanation of the European directive, if he wishes to know more about the subject. --Edcolins 20:55, Nov 19, 2004 (UTC)

Well, in fact you removed a side of the dispute and imposed your view of the dispute. Phrases like Intellectual Property are Us-centric. NPOV means document all sides and identify the stakeholders.

Whoever you are, this is the most preposterous and ill informed statement I've seen for a while. IP is a competely univerally accepted term. I'm not from the US, and I've used it for years. What would you have us call it instead? Baggie 15:29, 30 April 2006 (UTC)
IP is quite modern concept which stresses intellectual achievements. It tries to mix copyright and patents. That is a sort of propaganda from the US. In Europe we speak of Immaterial goods. Podmok 09:54, 8 June 2007 (UTC)
I'm from Europe and have never heard that term. It's also completely misleading. The whole point of intellectual property law is that it defines a way in which an immaterial thing (such as an invention or a design or a trademark or the expression of an idea) can be turned into a "material" possession that can be bought and sold like other "material" goods. Just because Richard Stallman fails to understand that doesn't mean it's a US-centric propaganda term. GDallimore (Talk) 10:38, 8 June 2007 (UTC)

Please cite your sources, preferably authoritative ones

I removed this paragragh:

On a broad definition, software patent is sometimes taken to categorise any patented invention which could be fully specified by the code of a computer program operating particular known conventional hardware.

Please cite your sources, preferably authoritative ones. Thanks. Please also justify why the term "computer-implemented inventions" would be broad (... close to the broad definition above). I don't think this is correct. --Edcolins 11:35, Nov 25, 2004 (UTC)

FOLDOC definition added

I've added the definition from Foldoc, not because I think it's particularly definitive, but because I think a very simple, even over-simplistic, definition along the lines of "a software patent is a patent which can be infringed by software" is a good place to start the article.

I think it's better for the reader to give them at least somewhere to stand, and then explain what's wrong with it, rather than throw them straight into a swirl of confusion.

The fact that Foldoc is by no means necessarily a definitive authoritative legal resource I think if anything underlines the point made in the subsequent paragraphs.

Furthermore, it's a definition which is out there on the net, so I think it's no bad thing to present it and critique it.

My text I think could use some copy-editing for better readability, but I hope it's a useful verion 0.1 enhancement.

The FOLDOC definition is fine enough. However the one you added is slightly different, a rather misleading: "a software patent is a patent which can be infringed by software". A patent can only infringed by an action (using, selling, importing, producing), not by a product, nor a process. I tried to reword you version, but I am not satisfied yet about the criticism of the FOLDOC definition. --Edcolins 12:10, Nov 27, 2004 (UTC)
I removed the part criticising the FOLDOC definition unless there is some facts behind this comment, which looks like "original research" to me. --Edcolins 21:24, Dec 8, 2004 (UTC)
I don't know that I'd necessarily agree with that comment; but I think the article does read better with your cut, now leaving the discussion of different places where people have suggested drawing the line re acceptable and unacceptable subject matter entirely to a single useful catalogue at the end of the section. (To which I have added the additional point of view that no software should ever be infringing whatever any patent says, because this is another legislative suggestion which is often made, whatever one might think of its pros and cons).
I have also moved the terms "software-enabled invention", "software-related invention", and "software-operated invention" back in with "computer-implemented invention", because as far as I can see all of these terms are entirely synonymous; and in clear distinction to the discussion in the remainder of the section as to different notions of how one might identify unacceptable "pure" software patents. --Jheald 20:41, Dec 9, 2004

"his forbids the direct copying of any part of the program code. Applying for, and being granted a patent gives much stronger protection. The invention achieved by the code may then be protected, and others who use the same invention may be sued to stop, or forced to licence the patent (even if they have come up with and developed the idea independently).

This gives the patent-owner much stronger protection for his invention. But it also means he has been granted a state-backed monopoly, and the chance to close out all competition, if the patented invention cannot be avoided. The question of whether in the area of software this is on balance a good thing or a bad thing for society has attracted intense and heated legal, academic and political disagreement. This is reviewed in the associated article Software patent debate."

This paragrpah is wrong, because software is no "invention" according to Art 5.2 EPC. In fact there is nothing such as a "software invention", professionals in the field do not speak of inventions but ideas, concepts, abstractions ecc.

"gives much stronger protection" is inaccurate as patents covers something different. Copyright law differs in member states of the EU. Copyright extends to more the just 1:1 copying. Stronger is no neutral term.

The real question is rather whether patent law was an appropriate tool to provide protection for software ideas/concepts. There are many indications that patent law fails. Hartmut Pilch for instance proposes a copyright style system.

Additionally the so called "protection gap" is pure fiction put forward by patent attorneys. Software developers don't ask for a software idea protection and if the burden of proof lies with the proponents of patents to show that it was the appropriate tool given the fact that patent pratice is developed under different market structures.

---Andre, 5.12.2004


Restructure Page

aberglas starts

I've tried to clean up the article and keep it tight. Still far from perfect, for example, a better history would be good. The last two sections are also rough. I hope people don't object.

What about a few typical examples? And a test suite with characteristics.

Note that this is NOT the place to have the Software Patent debate.

Rational:-

  • Software patents and Copyright. This is the main definition section, clarifying the different types of intellectual property.
Why compare Patents and Copyright. Why not Software trademarks and .. copyright?
  • Brief History. Given the substantial changes, I thought this was useful. Maybe add a new section after it "Current Status"?
  • Outline of Controversy. I felt that the controversy could not be completely ignored in this top level page. Especially as the Software patent debate page is such a mess. I'm hoping that others will tighten this, but not make it much bigger. Just the key points of the debate.
  • Scope of Software Patents. This is a rework of a previous section that was not well written IMHO. It is sort of definitional, but also a bit of a technicality. I would not object to someone deleting it entirely.
  • Law. I have not touched this. I think it probably should be removed, and merged into a Current Status topic. Software patents are all about law, so I don't see the point of the title.

I've also gone through this discussion and deleted the rambling discussion which predates the split into a separate [Software Patent Debate] page by Edcolins.

Aberglas 04:12, 29 Dec 2004 (UTC) aberglas

edcolins replies, definitions...

I have to say I have some reservations about your amendments...
Software patents and business process patents do have close ties. And both relate to services.
the other question is whether software patents will subsequently introduce bm-patents. Just think of ERP Systems.

A link to a new article about business method patent is in order but mixing them is not. A business method can be purely implemented without any kind of software, while a software can obviously relate to something else than a business method.

  • I am convinced a "Definitions" section should start the article. Otherwise it is impossible to understand the issues at stake. The "Software Patents and Copyright" section should come later since it is important to understand first what considered to be a software patent. A comparison is interesting but at a later stage. I do not understand why you belittle the use of a "Definitions" section.
  • The different type of intellectual property can be found and should be found in the Intellectual Property article.
  • Please cite your sources for you "definitions varying from broad to narrow". There is no point inventing new definitions instead of the ones already put forward.
I reverted some parts, but let other. Software patent is all about law and this should then come first to my humble opinion. --Edcolins 21:40, Dec 30, 2004 (UTC)
All about law? This is what von hayek criticized in the 40th. It is dangerous to leave this public policy debate to the lawyers. The horse rides the rider so to speak.
Patents are, and were about "public interest" and "incentive". They are all about granting a "temporary monopoly" for "incentive" to invent/create things which should eventually become "public domain". A definition expanding on what this actually means in terms of software patents should be stated along with what makes a software patent different from other types of patents, different from copyright and then brief references to debates for and against how the patent system serves this aim. We must be careful using terms which colour its perspective. EdColins must be careful not to let his own corporate interest/anti-OpenSource view override the neutrality as much as those with an OpenSource/public interest view must also keep this article neutral. Disputed terms such as "intellectual property" should be left in quotes and couched as being debated, as with all the debated parts, so instead of going one way or another, the debates are merely alluded to in an encyclopaedic manner. Keeping them in quotes, and referring to the fact that they are disputed is more helpful than avoiding them, or successive editors with different views scrubbing them out and putting them back in. A lot of propaganda and rhetoric don't belong in this article, and there are plenty of sites/forums/blogs on the internet to cover the debate, which could be then be linked to from the content under a heading "The Software Patent Debate". Obviously it is worth making sure that this set of links is representative and not one sided. --DannyStaple

aberglas response

Thanks to edcolins for reviewing my work.

Business methods

Business methods are only patentable when implemented in software, which is implemented on computing machinery. They are thus very much the same thing. Eg. is the the Amazon one click patent a "Software Patent" or a "Business Method Patent" -- the distinction is not useful. But I'll leave this alone.

However, the current first paragraph is a bit woffly IMHO, so I have tightened it up again, but leaving out Business Method Patents.

Definitions

I think the general definition of a software patent is a patent is a patent that concerns software. This is not worth stating. I don't think that there is much controversy with this general definition -- it is just a was to focus an article.

Where there is controversy is that the EU bans "programs for computers" and some advocates of patentability have tried to interpret this along the lines of "pure software that does not do anything useful". This is very much a technical legal point, so I have moved it under the Law section. I don't think that anyone goes out of their way to define "Software Patents" as such, but they do go out of their way to define the meaning of "programs for computers" etc.

I wonder about the utility of other quotes in the article. For example it's great that someone looked up the the FOLDOC definition "A patent intended to prevent others from using some programming technique". However it is essentially meaningless in IMNSHO, unless you think that there some unstated but meaningful distinction between a "programming technique" and a "software technique". I have left this in, but not because I think that it is helpful.

Where there is potential source for confusion for someone new to the area is with copyright vs patents. I take your point that these could be discussed at a more general level of intellectual property. However, in software this distinction is particularly important and easy to misunderstand. That is why I think that the distinction should be clarrified for software specifically. It is also why I made it promenant in this article -- in a practical sense it is definitional.

So I have left a brief Definitions section because edcolins things it is useful, but I'd be happy to see it removed. Others might want to expand it a bit.

I also moved the Controversy summary to the Software Patent Controversy page, I'm sure edcolins will agree with this one!

Aberglas 02:51, 1 Jan 2005 (UTC) aberglas

Aberglas,
  • I respectfully don't agree with this one: "Business methods are only patentable when implemented in software, which is implemented on computing machinery." In the States, it goes further than that. The State Street Bank decision and the statement "anything made under the sun by man is patentable" makes it reasonably clear (I think) that you can patent business methods, without the need for them to be computer-implemented...
  • Alright with spliting the definition section, it make sense.
  • Quotes are essential, if you should meet the wikipedia "cite your sources" criterion...
  • I changed this "However the European Patent Convention, Article 52 specifically excludes "programs for computers", so these patents are probably not be enforcable." to "However the European Patent Convention, Article 52 specifically excludes "programs for computers" as such." The assumption you make is doubtful to my humble opinion. The Case law of the Boards of Appeal of the EPO is consistent, and reasonably followed by national jurisdictions I think.
No. The EPO is an independend institution which did the reinterpretation on its own, anticipating a change of law. However, many courts in memberstate stick to other interpretations. This was the reason for the EU directive by the Commission: to harmonise law of member states in order to provide legal certainty for those patents issued. The EU directive will only cover this enforcability in the member states. The EPO case law only applies to the EPO pratice, it has no legal validity in the member states although member states are influenced by EPo pratice. Same the EU directive has no legally binding purpose with regard to the EPO pratice, however the EPO will have to politically correct its interpretation if patents in those EU memberstates are not enforcable because of the directive. Some states in Europe have a strong statutory law tradition. Member states are bound by the EPC, not by the interpretation of the EPC by the EPO.
Most EPC Contracting states have amended their national laws to correspond to Art 52 EPC. UK (PA 1977 Section 1(2)), France (CPI Art. L 611-10 (3)) and Germany (Patentgesetz S1 (3)) at least use substantially the same wording, and in particular the wording of Art 52(3) (.."as such"). You should also make clear the distinction between EU member states, and EPC Contracting states- the two intersect, but neither is a subset of the other. ~~M Bell
  • Moving the controversy section was indeed a good move..
Thanks for your work. --Edcolins 15:15, Jan 1, 2005 (UTC)

I think that EdColins comments make sense. I think that we are done for the time being. The Software Patent Debate article is a much messier project though, especially maintaining NPOV. All help gratefully accepted. Aberglas 01:02, 2 Jan 2005 (UTC) aberglas

NPOV for the worse

AB responses by Aberglas 01:31, 7 Jan 2005 (UTC) aberglas Your comments would have more weight if you registered a username!

I think now the article is not NPOV anymore

AB Interesting. I was concerned that it was getting too anti-patent. I supose that if everyone thinks that I am biased the other way I am doing well.

a) the term intellectual property is used which reflects a certain view and is terminology used in International lobbying

AB Well, they've been successful, it is the term that is used.

only in the US. It was introduced by WIPO.
Sadly that's not necessarily true - the Charter of Fundamental Rights which forms part of the proposed EU Constitution contains a clause "Intellectual property rights will also be protected". --Ryano 16:14, 19 Apr 2005 (UTC)

b) "Software patents are often confused with software copyright." In fact nobody involved in the debate confuses Patents with copyright. This is however often said by the patent attorney community. Clarifying position which nobody has. In fact there is no proof or evidence given for this remark.

AB Sure, but the page is not just or even mainly for "people involved in the debate". The media certainly gets these confused.

What kind of confusion is it? It is not enough to say: this is a this is b and media gets it often mixed up rather than starting with this "people are dumb" kind of accusation?

c) "Computers powerful enough to run complex software have existed since the 1950s." In fact software is not based on computers = sw execution machines but existed prior to hardware, e.g. on a Turing machine.

AB And this is the type of useless software that everyone agrees does not have any technical effect.

What meaning does it have that it has "no technical effect", would you say that software run on a computer has a tehcnical effect in the sense of the patent law?? this is a legally dubious remark. IMHO Software is always to be run on a "virtual machine". The VM can be a computer, or a interpreter that is itself run on a computer, e.g interpreted code, p-code, Java binary, prolog. I do not understand why the sentence matters here. Regarding virtual machines you shall get informed. Wikipedia is a wonderful source to bridge the knowledge gap.

d) The term "computer-implemented invention" was put forward by the European Commission It was put introduced by the EPO

AB ??

Some persons say it was cast by IBM's Fritz Teufel.

e) "Applying for, and being granted a patent gives much stronger protection." I disagree with that, it is another form of protection. "stronger" is not NPOV. The question is whether this kind of protection suits the market

AB Disagree. It is obviously stronger. That is what upsets lots of people.

? patent = 20 yrs. copyright >= 50 yrs, what upsets people is that classes of possible implementations are protected.
not "stronger", it protects a different subject matter. You cannot compare copyright and patents. And: Patents are a weaker form of protection e.g. regarding the duration. Cpr and patents are two entirely different systems. Would you say that trademarks are a stronger protection than copyright? Probably a "broader" protection. However it is the term "strong" that implies that it was better. And this is not NPOV.

f) while modern copyright has an indefinite life Us-centric

AB yes, but we need to keep things succinct. Australia is following. Don't know about Europe.

Wikipedia has to stick to the international standards. It is the Berne Convention or the CPR protection. USA and its Sbono act is a special case, nobody will follow.


e) Patenting software has become popular. Popular derived from latin means "common in the public". The cited examples are not individuals

AB Nonsense, I think that the meaning is clear.

wording counts here, how do you measure popularity. polpular among which community?

f) The EPO -- explain the shortcut

I believe this page shall not be for the debate but for the terminology only.

AB there is another page Software Patent Debate I have addressed your other concern about "software invention", although it is a bit trivial IMHO.

EPO is a shortcut that shall be explained. and terminology is often very fuzzy in the debate

"Practical effects of software patents" section

I am concerned about some parts of this particular section. Some bits should clearly be moved to Software patent debate. Other are pure speculations. In order to keep the article clean, factual and neutral, information should not be added unless supported by external sources. Look at this section George W. Bush#Public perception and assessments, where many sources are cited. --Edcolins 09:51, Jan 7, 2005 (UTC)

Please be more specific.

One the one hand I have tried not to put the debate in the patent section. But on the other hand I try to keep the debate page free of basic facts.

I think that the test for what goes in the main softare patent page should be facts that

Someone would want to read that wants to understand software patents but has no interest in public policy. For example a software engineer or CTO that needs to develop policies for dealing with them.

So to such a person the number of patents is important, as are the cross licencing deals. Are they actually being enforced? Can they be enforced (maybe this should move to the law section)? The new line of business section emphasizes the fact that software patents are very real and need to be planned for (maybe the word "useful" is gratuatous"). Etc. with the open source perspective, inventive step (maybe should move to law and be elaborated), and searching. The last paragraph may fail this test and could be moved to the debate page.

I don't feel the need to cite references to things that are obviously true, eg. "Many open source developers fear software patents...". Possibly the other sentences in that paragraph could be tightened up.

Anyway, tell me your gravest couple of concerns and tighten up as needed.

Aberglas 01:38, 8 Jan 2005 (UTC) aberglas

For instance, this is unsupported by facts:
"In practice, software engineers rarely search patent databases and applications looking for new inventions that could benefit their projects. This may be because 1. the lack of inventive step in many software patents, 2. the obscure language with which software patents are described, and 3. the risk of being assessed for triple damages for knowingly infringing one. Many infringements are for independent inventions."
Generalizing what all software engineers would do, without your citing sources, is odd. This paragraph should be removed to my opinion. "This may be because (...)" also is pure speculation. --Edcolins 22:36, Jan 11, 2005 (UTC)
In fact all the leading books about software engineering do not mention patenting and I do not know a single programmer who browses patent databases except myself. I know a lot of them. I affirm the second remark. The third does only apply for the US legal situtation as far as I know.
Point 1. is just daft. The 'lack of inventive step' is a reason to engage with the patent system, not just ignore it. If more software engineers took notice of the damage ignoring IP issues can do, and not just stick their heads in the sand, the system would be much more workable. Fair enough, the USPTO's performance from around 1998 to around 2002 was hardly glorious, but my personal experience recently is that the chances of getting a duff US software oriented patent are rapidly diminishing. There are ample opportunities to apply for re-examination, or to oppose patents, rather than just jumping up and down about how unfair the system is.
Point 2. is based on reality, but far too sweeping. I'm not sure how much benefit software engineers will get from the average patent disclosure. To be reasonable about this, I would also say the same of technical papers - most of these are completely incomprehensible. However, I think that the better patents, which are in turn also probably the strongest, are pretty good and my clients (mostly software engineers themselves) tell me that they learn plenty from patent disclosures.
Point 3. is a US point and is not relevant elsewhere. However, ignoring the problem of potential patent infringement seems risky, whatever the penalty. There is a naivete to the way that certain software engineers seem to think that the business risk attached to competitor IP doesn't apply to them. Baggie 15:42, 30 April 2006 (UTC)
The points about triple damages, use of patentspeak rather than source code, and the difficulties of interpreting claims and their boundaries were all sharply flagged in the 2003 Federal Trade Commission report (see especially [3] and [4]), and can also be found widely attested elsewhere. Note especially the FTC's summary: "[Representatives] discounted the value of patent disclosures, because they do not require the disclosure of a software product's underlying source code". The questionable quality of many software patents also appears again and again as a running theme. --Jheald 23:59, Jan 11, 2005 (UTC)
Thanks Jheald for the link. But I should be clear as to why I put the paragraph in in the first place. It was not to argue that software patents were bad. It because someone not familiar with software patents but uninterested in public policy might want to know whether they should spend time searching to see if they might infringe some. In many peoples view this is not productive. (EdColins does a good job of keeping us on our toes citing references. The article is better because of this exchange.) Aberglas 01:21, 12 Jan 2005 (UTC) aberglas

suggested Definition swpat

http://www.ffii.se/erik/EPIP/img8.html

Instead of "suggested", I thought I would make it definite.--Nowa 21:36, 9 January 2007 (UTC)

Exploitation of cross-licencing agreements by a third party

Suppose two companies have a cross-licensing agreement and one of them infringes on a patent of the other, to which that other company then does not react. Now suppose a third party notices this and then comes up with its own solution, also infringing on that patent. If the 'damaged' company then sues them, can they then claim that that company should first sue the first 'infringer' and if they don't, then obviously they don't value their property as much as they claim (or something similar)? Would this hold up in court (in various contries), would it simplify matters and, most importantly, would that reduce the cost of the court case sufficiently to make it affordable for others than large companies? The best outcome would be that a judge rules that the 'damaged' company will first have to sue the other company. Which they won't, so you're in the clear because it will never appear before court. I don't know much about law (and of course this also depends on which court we're talking about (eg US or EU)), but I can imagine something like this might work. Also a central question there is if cross-licensing agreements have a legal basis. DirkvdM 11:14, 2005 May 6 (UTC)

I'm not sure that patents imply any obligation to defend the rights granted therein, unlike trademarks, for example. If it's shown that you're not taking action to defend your trademark, you can lose it, but I don't believe there's any such obligation on patent holders. So patent holders can basically pick and choose who they sue and who they don't sue. But IANAL --Ryano 12:49, 6 May 2005 (UTC)
If company A owns a patent PA, if company B owns a patent PB, and if company A grants a license to company B for exploiting patent PA while at the same time company B grants a license to company A for exploiting patent PB, company A and company B are said to have concluded a cross-licensing agreement. This means company B can exploit the subject-matter claimed in patent PA and that won't constitute an infringement. So if "two companies have a cross-licensing agreement" and the situation in which "one [A or B] of them infringes on a patent [PB or PA] of the other [B or A]" cannot arise.
If a third party, let's call it company C, infringes patent PA, company A can sue company C for infringement. In most legislations, in this case, the licensee, company B, can even become party to the court proceedings and claimed damages from company C. I would say company C has no chance at all arguing that company doesn't value its intellectual property in this case.
Your last question is very interesting: "Are cross-licensing agreements legal?". To a large extent, yes (otherwise this could completely block the exploitation of a technology of which two or more inventions are patented), but this can easily become a complex issue, involving (as far as the EU is concerned) Art. 81 and 82 of the EC Treaty (abuse of dominant position, etc) as well as licensing directives, cartels, etc. TINLA. You might wish to post this question on Talk:Patent instead. --Edcolins 13:06, May 6, 2005 (UTC)

"Claims" vs "relates to"

I've changed 'claims' to 'relates to' which is how the EU patent directive is worded:

"(3) The provisions of paragraph 2 shall exclude patentability of the subject-matter or activities referred to in that provision only to the extent to which a European patent application or European patent relates to such subject-matter or activities as such."

Carmack's paragraphs

Under "Practical effects of software patents" there is the story of the Carmack/Creative patent issue, containing the sentence:

It has since been discovered that Carmack himself had prior art in the source code of Quake3 and id Software would have likely won any litigation and invalidated the patent at the expense of Doom3 being delayed, which they could not afford.

This doesnt make any sense to me having re-read it numerous times, what exactly is "prior art in the source code" supposed to mean? Can someone please rephrase it to make it clearer.


  • It means that Carmack wrote _and_disclosed_ parts of the Quake3 source code before the date of Creative's (supposed) invention, and that said parts encompass techniques/patentable-subject-matter/etc. constituting prior art with respect to the claims of Creative's patent. (Quake3 <> Doom3)
I'm confused. If Carmack has prior art, surely he could get a settlement. My suspicion is that his assertion of prior art is not as strong as first made out. There seems to be an assumption here that Carmack (respected, non-patenting software pioneer) must be somehow better than Creative (big, bad patenting corporate). This seems to me to be an unrealistic simplfication. I also disagree that the patent concerned has in some way hampered independent innovation. This sentence makes so many assumptions, I'm not sure where to start. Maybe Carmack didn't actually have the invalidating disclosure to overturn the patent. Maybe he coudn't prove his assertion as to the timeline of his creation. I obviously don't know the answers here, but I suspect that neither does the writer of this passage. Baggie 15:52, 30 April 2006 (UTC)

Information on havens from patent-based persecution?

There is information on major jurisdictions of the world and how they treat software patents. Does anyone have any information on jurisdictions, however small, that effectively ignore or reject patents, or has WIPO, etc. managed to destroy all such possible havens. —Preceding unsigned comment added by 65.39.91.183 (talkcontribs)

In a certain sense, all nations are havens from persecution from the patents of other nations since patents can only be enforced within a given nation. For example, if an inventor is granted a patent in the US but is not granted a patent in Taiwan, then then a third party is free to practice that invention in Taiwan, but not the US. The problem is that at least for the US, a patent that claims a computer system (hardware + software) is deemed to be infringed if any part of the system resides in the US. This is what got Research in Motion in trouble as far as the NTP, Inc. patents goes. Their servers were located in Canada, but the mobile BlackBerry(tm) devices were in the US.--Nowa 02:14, 24 April 2006 (UTC)

copyleft is offtopic

"Copyleft (that include several open source and free software licences) license agreements are examples of how copyrights are used to encourage the public disclosure of improved versions of a particular piece of software. The license agreements prevent third parties from copying a given piece of source code unless said third parties agree to make their improvements to the source code available to the public under similar open source terms when they distribute the program."

This is off-topic. Stick to the point, it is about patents, not special arrangements under copyright jurisdiction —Preceding unsigned comment added by 84.136.200.121 (talkcontribs)

Perhaps such a thing belongs in an "Alternatives" section as merely a link to CopyLeft. 213.46.174.149 13:28, 10 November 2006 (UTC)

Trivia

"Patents are not granted automatically. Inventors must file patent applications in each and every country (such as Japan, China, US or India) or region (such at the European Patent Office) that they want protection in. These patent applications must disclose how to make and use the invention in sufficient detail so that another person of ordinary skill in the art can reproduce the invention without undue experimentation. If it is not self evident, then a patent application must also disclose what practical utility an invention has. Patent applications also have one or more claims which set out the boundaries of the invention that the inventor asserts is his/her original and non-obvious invention.

Most, but not all, countries or regions then assign one of their patent examiners to the case to determine if the inventor is entitled to the patent claims they are requesting. If so, then the claims are allowed and, upon payment of a fee, a patent issues.

Different countries have different standards for allowing patents. This is particularly true of computer implemented inventions. Thus a particular computer based invention may be patentable in one country, such as the US, but not patentable in another country, such as the member states of the European Union.

Once a patent issues in a given country, the patent owner can prevent others from making, using or selling the claimed invention in that country. Patents are generally enforceable for up to 20 years."

Is this about software patents? No. It is an explaition of patenting. E.g. it has to asked whether software solutions are "inventions" at all. The paragraph talks about patenting of inventions. —Preceding unsigned comment added by 84.136.200.121 (talkcontribs)

none at all?

Red Hat is another special case: it has no original intellectual property of its own

Is this literally true? none at all?--203.6.205.131 06:01, 5 September 2006 (UTC)

No, it is wrong. I removed the sentence. Please see Red Hat#Business model where a couple of red Hat's patents and patent applications are listed. Thanks for pointing out this wrong statement. --Edcolins 07:06, 5 September 2006 (UTC)
The article should never have contained a sentence so silly. No patents of its own would be a concrete, if wrong, statement. No "intellectual property" is rubbish talk. Gronky 11:14, 5 September 2006 (UTC)

defensive patents

Did a fairly major cleanup of this section and added references. Comments? Suggestions?--Nowa 00:36, 24 September 2006 (UTC)

Software patents specifics

The sole purpose of the section Software patents specifics appears to be to link to this document

I've had a look at this document, and it contains numerous factual errors - referring to refused patent applications as granted patents, claiming that the 1-click patent has been grand in Europe etc.

Add to that the highly opinionated nature of the cited document as a whole and I wondered if anyone would anyone object if this entire section were deleted. Is there anything of merit in the cited document? I didn't want to remove the section without raising the issue first as it's been there for a while without objection. GDallimore 02:28, 30 September 2006 (UTC)

This is off-topic in an article about software patenting:

" Ironically, open source solutions are creating copyright infringement problems for proprietary software companies. The developers of proprietary software often incorporate open source code into their systems. This may violate the copyright licenses of said open source code if the proprietary companies do not make their code similarly available. Systems, such as blackduck[1], are available to scan software to determine if it contains copyrighted open source code. "

An article on software patents properly includes a discussion of their relationship to software copyrights. See the section Software patents vs copyrights--Nowa 17:07, 6 December 2006 (UTC)
please finally tell the world what is the relationship of GPL-violations to patents ?? or as a simple test: please make a sentence containing the words "patents" and "copyright-violations" which is more useful than "both somehow got something to do with intellectual property". --81.210.202.103 00:32, 7 December 2006 (UTC)
Good suggestion. I think GDallimore's latest edits do the job nicely.--Nowa 22:24, 7 December 2006 (UTC)

First software patent: references...?

I have removed the following bit from the article:

The first software patent ever granted is probably a patent for a "computer having slow and quick access storage, when programmed to solve a linear programming problem by an iterative algorithm, the iterative algorithm being such that (...)" applied for in 1962 by British Petroleum Company<ref>http://www.cippm.org.uk/pdfs/JILT%20kretschmer%2011_03.pdf, see end of page 3</ref>. The patent relates to solving simultaneous linear equations.

The reference given does not mention BP and this alleged patent, but it mentions a "1966 UK decision allowing a claim to a computer 'when programmed to solve a linear programming problem by an iterative algorithm' since it may be regarded as 'a machine that has been temporarily modified'." (sic) Linear programming are not linear equations, btw.

It would be very interesting, though, to know if the information is true (this is a most blatent case of purely algorithmic patent), who granted the patent (US or UK ??) and whether the corresponding law has changed since then.

More info on the (probably unrelated, or much distorted) 1966 UK decision would also be valuable. --Josce 11:31, 4 December 2006 (UTC)

Actually, by the same author as the reference there's a text on the web, which has this additional info about 1966 patent: Slee & Harris's Application, 1966, R.P.C. 194. According to Beresford (2000, p.4), this UK Hearing Officer decision is the first reported software patent decision worldwide The book quoted is Beresford, K (2000) Patenting Software under the European Patent Convention, London: Sweet & Maxwell.
But there's no more detail. --Josce 11:39, 4 December 2006 (UTC)
I have added more information on the patent GB1039141. The complete specification can be found on espacenet. --Edcolins 20:23, 4 December 2006 (UTC)
Thanks! I have corrected "solving linear equations" to "linear programming", this is quite different and (along with the discrepancy of dates) was the reason of my doubts.--Josce 10:45, 5 December 2006 (UTC)

Why not do use "protection"

The article mentions "protection" in a few places, but I think this is inappropriate. Patents only offer "protection" from one point of view - that of someone with a business model based on patents, be it a patent office (such as the European Patent Office, which is funded by the patents it accepts), or be it a private company. Wikipedia should describe patent holding in a neutral way. So instead of saying "get patent protection in India", we should say "get a patent that is enforceable in India". That's a fact, from any point of view, the patent is enforceable. Whether it protects or not is not a fact from all points of view. Gronky 22:30, 7 February 2007 (UTC)

To clarify: I'm not saying that "protection" is never valid in this article. For example, it's not wrong, IMO, to say that patents can be used for the purpose of protecting a revenue stream. Gronky 22:33, 7 February 2007 (UTC)
I think I see your point. Patents offer "protection" from the viewpoint of a patent holder. From the standpoint of someone who practices a patented invention without a license, however, patents provide a threat. What they seek is protection from a patent. Have I captured your view on this correctly?--Nowa 22:52, 7 February 2007 (UTC)
Enforceable is also not appropriate. It makes patents sound like they are only weapons that people use to bash others over the head with in court. Compare the number of patents with the number of court cases, and you'll see that's hardly the case. Suggest removing adjectives entirely! GDallimore 08:24, 8 February 2007 (UTC)

GD, I like your wording.--Nowa 16:28, 8 February 2007 (UTC)

Disputed

Running the Bessen/Hunt search technique (see List of software patents) on the USPTO database as far back as it can go (my search ran from 1900 to 1976) found patents that are classed as software patents going back to 1971. Presumably, earlier patents are not full text searchable, hence the 5 January 1971 cut-off date.

The statement that the USPTO traditionally did not grant software patents therefore needs verification or removal. As it stands, it sounds very NPOV as it sounds like an anti-software patent assertion. (see talks by Stallman, for example, although he at least recognises that software patents rarely describe or claim specific software algorithm implementation but instead claim processes that are to be carried out under computer program control).GDallimore 10:53, 8 February 2007 (UTC)

GD, Have you tried Bessen/Hunt on Google patents? That does full text for the pre 1971 patents.--Nowa 16:31, 8 February 2007 (UTC)

For example U.S. patent 3,403,386 issued in 1968 claims and "apparatus" which comprises a "comparator means". See claim 1. The specification discloses that the comparator means may either be hardware means or software means (col 6 lines 30 - 35). So I would argue that this is a software patent in the sense that it covers a computer running a particular piece of software. In that sense, a patent on any machine that describes the set of steps that the machine carries out is a "software patent". --Nowa 16:50, 8 February 2007 (UTC)

Good thinking, although the interface isn't powerful enough to do the proper search. I did a quick search just for patents with the exact phrase "computer program". One that jumped out was US 3316539  granted in 1967 (so I beat yours!).
It's interesting looking through these older patents and seeing where the truths and the falsehoods lies in accusations that patent offices have recently started extending the scope of patentable subject matter. When it comes to processes carried out by a computer, that seems to be misguided or just plain wrong. There are plenty of processes carried out by computer that have been patented since computers were invented. What HAS changed are the computers themselves. Now, a chap sitting at home can easily code a computer himself to achieve a desired result or perform a desired process, whereas in the 60s and even the 70s programming a computer was a task only performed by a very elite group highly skilled experts modifying the very apparatus itself. What seems to have happened is that patent infringement has been brought into the home due to the extension of technology, not due to the extension of patent laws. As a comparison in the field of copyright - copyright hasn't changed significantly in a long time, but home users can now make copies of their music collection with ease. Selling said copies is also much easier, thus bringing copyright infringement into the home. Is it any wonder, therefore, that patents and copyright have such a bad rap these days? Once upon a time, only business or highly devoted individuals could even dream of doing something that infringed on a patent or copyright in a way that would draw the attention of the right holder. Now, a teenager in his bedroom can create thousands of infringing articles due to the power of his PC.
Where I am seeing a recent extension in the scope of patents, however, is in the field of business methods. Run a search for "business methods" on the Google site and you find practically nothing until the early 90s. OK, not a very scientific test, but revealing. I am afraid I cannot give my personal view on this in such a public forum. :)
Sadly, I'm not sure anyone's noticed these trends, so putting in the article would be OR. Maybe I need to write an article, get it published in a reputable source and then cite it here :) Sadly, from past experience, newspapers and the like only care about patent disputes and scandals. Although I've had some dry articles published in patent periodicals, my attempts to garner more general interest have failed dismally. GDallimore 17:35, 8 February 2007 (UTC)

If you read the The opinion in Diamond v. Deihr, you will see that the reason it went to trial was because the patent examiner believed that software did not fall under statutory subject matter. The examiner cited the supreme court case of GOTTSCHALK v. BENSON, 409 U.S. 63 (1972) where the court rejected the patentability of a program that converted a Binary Coded decimal numbers into pure binary numbers. In that case however, the Court also said that "It is said that the decision precludes a patent for any program servicing a computer. We do not so hold." The fact that such a case was brought to trial, and that the Court had to clarify the matter in such a way, and that the issue was again raised a few years later in the Diamond case, indicates that the patentability of computer programs was at least uncertain at the time.

I propose the change include the following elements: 1) The differing views of software as both mathematical (old view) and practical process(present view). 2) Statutory subject matter restriction on scientific truths and mathematical formulas 3) That the patentability of software before Diamond v. Deihr was unclear. Cite the software patents issued before this date as well the PTOs position in Deihr.

Nelziq 22:59, 19 April 2007 (UTC)

I think you (1) is potentially misleading and walks into propagandist territory that "software is nothing more than mathematical algorithms". As far as I can tell, software is not viewed differently now than it was. Instead, software is being used for different things now than it was in the past. In the past, software was used to do nothing more than perform mathematical operations such as converting "Binary Coded decimal numbers into pure binary numbers" - this was the only effect of running the program. These days, software is more commonly used to achieve some effect which is more than a mathematical process. The uses of software have changed, not the way software is viewed in patent law. As far as I can tell it is, and always has been, the effect of running the computer program that matters, not the computer program itself.
Something that has changed in patent law, as I point out above, is the assessment of which of these uses of software are patentable. If the effect was an improved business method it never used to be patentable in the US and is not in Europe, but these practices are being questioned and have changed in the US of late.
I came across a nice concise summary of the practice in Europe, which might help to understand. In Europe, a patent may only be granted for an invention that solves a technical problem. Getting a computer or network to run faster or moving a robot arm more efficiently may be technical problems, but the act of programming the computer itself is not. GDallimore (Talk) 09:59, 20 April 2007 (UTC)

Licensing of software patents

This section is a mess, tangenting off into Patent troll as well as general patent licensing issues. Maybe the section would better serve the problems that patent licenses are not often compatible with GPL? There's the question of whose fault that is - the licensees for requiring unsatisfactory terms or the GPL drafters for refusing to find a comprimise and thus concluding that free software and proprietary software are wholly incompatible? The rest of the stuff can be deleted or moved to the appropriate articles, although I doubt that anything would be lost from Wikipedia as a whole if it were just deleted.GDallimore 10:57, 8 February 2007 (UTC)

In fact developers didn't request patents. Patents invaded their profession. I find the section quite informative. Despite that Propaganda as " Other patent holders are in the business of inventing new computer implemented inventions and then commercializing the inventions by licensing the patents to other companies that manufacture the inventions." needs to be deleted. Podmok 09:51, 8 June 2007 (UTC)
It may be informative, but the point is that it's information that is already elsewhere in Wikipedia and is not needed or particularly relevant to this article. The alleged "propaganda" you cite is essentially a definition of a patent troll and should be deleted because it's more about patent trolls than it is about software patents, not because there's anything wrong with it. Don't come here making bald statements that "patents invaded their profession" and expect a good reaction when you start claiming that other things are propaganda.
One day I might get around to properly tidying up this article, like that licencing section, with proper sourcing, but I'm finding it's very difficult to work on articles relating to topics that you know far too much about from experience... GDallimore (Talk) 10:44, 8 June 2007 (UTC)

How many years ?

Hello. Could someone add to this article the duration of patents for the United States, Japan, the United Kindom and Germany ? Thanks in advance.--Youssef

Please see Term of patent. --Edcolins 08:46, 24 February 2007 (UTC)

The companion article to this one, listing important software patents, has just been nominated for featured list status. Please take a look, see how that article might be improved and leave comments. Let's prove that there can be at least one uncontroversial article on this topic! Thanks. GDallimore (Talk) 15:09, 21 April 2007 (UTC)

Definition

I thought it best to start the article with an affirmative statement as to what a software patent is as opposed to what it is not. Clarification can be made in the body of the article.--Nowa 10:53, 29 May 2007 (UTC)

Neutrality

The new definition may not be neutral and may not come from a neutral, reliable source. When was it published? By whom exactly? Is the author recognized amongst all parties involved in the debate? Is the definition widely accepted by lobbyists, academics, lawyers, MEPs, and so on?

To meet WP:N, one widely accepted neutral definition or several alternative definitions representing the various opinions should be provided on top of the article. Providing a unique definition supported by the FFII, very much involved in the software patent debate, is not neutral. If there is no widely accepted definition, the reader should not be misled in thinking that there is one. And if the use of the term "software patent" is not widely accepted, the article should say so...

Finally, the expression "patent on computer-implemented inventions" is not merely a "related term", but it refers to the same concrete reality! The expression "related term" is misleading IMHO. The article should be about the concrete reality behind the expressions "software patents", "patents on computer-implemented inventions" or whatever you want to call them. Although a section could be dedicated to the definition and the expressions, the double-speak and the loaded terms, the introduction should be about the reality.

Thanks. --Edcolins 19:58, 29 May 2007 (UTC)

Ed, I actually think that the FFII definition is fairly NPOV, even if they are strong adherents to a particular point of view. Absent a more neutral source, why not leave it as is without casting aspersions?--Nowa 00:35, 30 May 2007 (UTC)
Since I was the one who added the definition, here's my thinking on the subject.
From my purview of the literature, "Software Patent" is typically perceived as a negative phrase, but without any real understanding of what it means. Often, any patent that has anything to do with computers and looks a bit dodgy is labelled a software patent and then swiftly followed by a "this evil must be stopped" kind of statement. The Bessen Hunt technique doesn't help understand what a software patent is, just lords itself as a way of identifying lots of them very quickly, which is why I removed a reference to it from the introduction.
To try to counter this negative perception, I searched everywhere for a sensible definition of "software patent". Definitions of CII's from the directive wouldn't work since it is strongly disputed as to whether the two are the same (although I would say that they are "related", this being the most NPOV word I could think of in the face of anti-sofpats saying they're synonyms and the EPO saying they don't grant sofpats). Eventually, someone pointed out Erik's seminar with it's definition of a software patent that, although based on the CII directive, was not the same as anything proposed there.
Personally, I think it's a pretty poor definition (no worse than any definition propsed during the CII directive debate, I hasten to add!) which ignores problems about whether hardware implementation would avoid the definition, but it has one redeeming feature which I think makes it a suitable basic definition for this article: "performance of a computer". In my view, this cuts down the scope of what is meant by a software patent adequately in the face of the FFII position that anything that falls within that scope should not be granted a patent. Therefore, while the FFII is not a neutral party to this, their attempt at a definition seems acceptable since it does not try to suggest that anything that looks a bit dodgy and happens to involve a computer or computer programs should be unpatentable, only those computer programs that control the performance of the computer itself.
I would therefore like to go back to the sourced statement here (thaks Nowa for finding a new link after the previous one died) saying that there is no accepted definition of what constitutes a software patent, but here is one definition that has been suggested. A section on "definition" could go into more detail if necessary, but for a pithy introduction that combination of useful information but highlighting the confusion sounds about as perfect a balance as would be possible. I'll make the edits and people can see what they think.GDallimore (Talk) 08:53, 30 May 2007 (UTC)

inventive step

The inventive step and non-obviousness requirements that should be met for a patent involving software.

inventive-step (EPC) and non-obviousness (USA) are the same. These are trivial conditions which apply to all patents, its off-topic Podmok 09:45, 8 June 2007 (UTC)
Firstly, you're wrong to say that the two are the same. If you want a source for that, read the January 2007 edition (I think it was that month) of Patent World discussing the (at that time) forthcoming Supreme Court decision on the topic. It discussed at length how they were similar in some respects, but different in others. Besides, inventive step and non-obviousness is the title of the article on the subject.
Secondly, I agree that inventive step applies to all patents, but the debate about software patents in particular seems to have a great deal of focus on "obvious" patents and the allegation that too many "obvious" patents being granted. It is therefore not off-topic. GDallimore (Talk) 13:03, 1 June 2007 (UTC)
either you talk about inventive-step or you talk about obviousness but no jurisdiction discusses both terms. Obviousness is just the corresponding US legal concept. I don't say that US and European law is essentially the same. Of course it is not. As a French brevet is not a UK patent. But it is the very same concept. Podmok 09:45, 8 June 2007 (UTC)

Aspects of the debate

Erik has suggested that "Machines/Process involving software" is political and has instead change the introduction to whether "software is patentable subject matter". This is not appropriate since the debate is far broader than this and highlights a lack of understanding on Erik's part.

Take a computer, for example, performing a certain task. That is not software, that is a computer. Apparently, there is no debate about whether this should be patentable according to Erik. However, a computer performing a certain task does "involve" software. Consequently, it is part of the debate, as it should be. Therefore my wording is correct.

I'm not going to engage in an edit war, but strongly recommend that Erik's last edit be reverted since it is plainly wrong. GDallimore (Talk) 14:05, 1 June 2007 (UTC)

addendum. Erik may be confusing "implemented in" with "involved". I have never heard any suggestion that "involving" could be a politically motivated obfuscation maneouvre.GDallimore (Talk) 14:53, 1 June 2007 (UTC)

GDallimore, two things: One, I think what the law says is patentable is one of the core issues of the debate. And two, the word "involving" is unclear and blurs the issue. As I see it, the question is whether the achievement of a programmer, formalised in a patent applicaltion, passes the "subject matter test". This is quite clear from the EPO guidelines of 1978: "If the contribution to the known art resides solely in a computer program then the subject matter is not patentable in whatever manner it may be presented in the claims.". You can call it "achievement", "contribution to the known art" or "invention", or whatever, but it *is* something. What "the achievement" involves is not a core issue in the software patent debate. Best regards. //Erik

You're just not understanding. Ask the question "is software patentable subject matter" and people will avoid the question by saying "I'm not trying to patent software". I've spent years working on things that have some connection with software but hardly any of those have relied on a particular programming technique. Take a machine for mixing chocolate. One of the claims might say that the machine could be controlled by a computer. Is that patenting software? No, it's patenting a machine that is under computer control, but software is involved.
Perhaps you're confusing what is patented with what what would infringe that patent. In the above example, someone might write and sell a computer program that would control the machine. Part of the debate MUST be whether do this would/should count as contributory infringment. And what about the person who uses that computer program to control a machine? is that an infringement? Yet, if we just limit to the debate to patenting software, then these important questions do not arise. GDallimore (Talk) 09:09, 4 June 2007 (UTC)
Without getting into the debate, I did take the commented suggestion and tried to improve the wording while keeping it neutral. I moved the comment here:
The term "process/machine" has been used because "invention" would be deemed POV by
one side of the debate while the term "idea" would be deemed POV by the other side
of the debate. NPOV improvements to the language here would be welcomed.
CRGreathouse (t | c) 13:03, 5 June 2007 (UTC)

That just made no sense whatsoever to me. I've deleted it all as unreferenced to avoid the argument. Find a reliable source which explains what the debate is about and maybe something can go back in. GDallimore (Talk) 15:05, 5 June 2007 (UTC)

If someone has a better wording they can change it. Since I think the material should stay I added in some references—trying to keep them on both sides of the argument as far as possible. What phrasing do you prefer? CRGreathouse (t | c) 15:59, 5 June 2007 (UTC)
You just added [not in citation given] for The Anatomy of a Trivial Patent (and another link) for non-obviousness. I'm a little confused by that; the whole article is about the question of obviousness vs. non-obviousness in software patents. Am I missing something? CRGreathouse (t | c) 20:02, 5 June 2007 (UTC)
Nevermind, I see you added a link of your own.
CRGreathouse (t | c) 20:05, 5 June 2007 (UTC)
What was missing was the term "software process" which I do not think actually means anything. GDallimore (Talk) 20:44, 5 June 2007 (UTC)
I don't like the phrase myself and don't mourn its passing. I chose it only to attempt to stay the middle ground between the wordings in the comment. CRGreathouse (t | c) 13:48, 6 June 2007 (UTC)
Cool. Sorry if my irritation at lobbysists claiming that every phrase except the ones they want to use is biased spilled over at you.GDallimore (Talk) 13:52, 6 June 2007 (UTC)
No problem. Good luck keeping them at bay. CRGreathouse (t | c) 17:49, 6 June 2007 (UTC)

Legalistic

"There is intense debate over the extent to which software patents should be granted, if at all. Criticisms of software patents include:

  • Whether software is patentable;[3] and
  • Whether the inventive step and non-obviousness requirement is too easily satisified for software.[4]"

--> both are just legalistic aspects. The economical aspect needs to be added.Podmok 09:41, 8 June 2007 (UTC)

This article is not the place to have the debate or even list ALL of the issues in the debate. Besides, there's hardly anything in the article about the economics of software patents, so it would hardly be appropriate to mention such things in the introduction. If you can find reliable sourced information about the economic effects of software patents (there is hardly anything out there, by the way, so good luck) and add a discussion into the article, THEN the introduction can be expanded. GDallimore (Talk) 09:48, 8 June 2007 (UTC)

POV anon edit

The opening was changed to

Criticisms of software patents include:

  • Whether software is patentable;[3] and
  • Whether the inventive step and non-obviousness requirement is too easily satisfied for software.[4]

Support for software patents includes:

  • Encouraging innovation [4]
  • Encouraging disclosure of inventions to the public to expand the public knowledge rather than keeping them secret [5]
  • Encouraging commercialization of ideas [6]

The first (preexisting) list are questions to be addressed in the debate, and the second list are arguments for software patents. The NPOV remedies would be to add counterbalancing points with actual criticisms, or to drop both the supports of software patents and return the "criticisms" to the neutral debate points that they were in the original version.

"Whether software is patentable" is not an argument for or against patents; it is a feature of the debate. Those who support software patents argue that it is patentable and those who oppose (often) argue that it is not.

The strength of the non-obviousness requirement is likewise a feature of the argument -- although this should be better written. Those who oppose software patents say that this step is too easy; supporters (often) argue that it is not.

Of the new points added, I think that "encouraging innovation" if properly reworded would be a good debate point to include. I've read pro-software patent arguments that suggest that software patents encourage innovation, and anti-software patent arguments that suggest they stifle innovation.

As I don't want to get into a war over this (and frankly prefer the WP:0RR anyway) I'm not making changes to the article, at least not now. Any thoughts on the ideal wording? I'm thinking

The software patent debate covers:

  • Whether software is patentable;
  • To what degree software patents encourage innovation; and
  • Whether the inventive step and non-obviousness requirement is too easily satisfied for software.

Suggestions?

CRGreathouse (t | c) 14:55, 23 September 2007 (UTC)

Template:Computer programs, software and patent law

I suggest the name to be changed to "Software patent" or "Software patents" according to the article name. There is a discussion on the template's talk page.--Kozuch (talk) 10:44, 18 March 2008 (UTC)

Opening section, and proposed rename

Hi 02barryc (also 81.157.171.8 as you said), thanks for your message on my talk page, and the stimulating comments. Indeed, I didn't revert your edit to the opening section. However, I concur with the reasons provided for reverting it: "what's software "design" supposed to mean? Revert unreferenced definition" (by User:GDallimore). The definition you introduced:

"A software patent is a legal instrument, intended to protect software designs in the same way as a regular patent would"

raises issues IMHO (and is different from writing "a software patent is a patent for software" - if it is the intended meaning). What does "software design" mean? This seems slightly unclear to me. But more importantly, your definition tends to suggest that a software patent would be like a "regular patent", but would not be a regular patent. I must disagree. A software patent is nothing more than a name, with no legal value whatsoever (at least to my knowledge), given to some patents, using various criteria (ranging from rigourous keyword-based identification - see for instance Bessen/Hunt_technique - to rather arbitrary, unexplained criteria). That's a problem with the topic (unclear boundaries). I concur that the actual opening section is not excellent, and I would rather rename the article to Patentability of software, but no consensus so far. Any proposals would be of course helpful, to improve the opening section or the title...

Besides, my purpose was not to be off-putting in any way. I might have overused a template for letting you know of the revert. If this can fuel a discussion, good enough maybe... Cheers, --Edcolins (talk) 21:02, 5 August 2008 (UTC)

Fair enough, just as well I assumed good faith! I now fully understand the reasoning why my edit was reverted. As you probably understand, the opening paragraph is, in my mind, completely correct and encyclopedicly poor, for the reason that I if were someone who didn't know the slightest bit about software patents, and I couldn't be bothered to go into too much depth, the opening paragraph (and I think the large part of the article) fails to give any definition, which is probably what I was looking for.
However, having looked at patent offices (UK PO, EPO, USPO to be precise), they seem to make no distinction between a patent and a software patent. This seems to indicate to me, along with other stuff, that really this article should, as you recommend, be moved to an article on 'the patentability of software', and any other parts that could not be incorporated into that be put in a 'software' section in the main patent article.
As to the template, it's just a little formal and intimidating, that's all.
--BarryC (talk) 22:03, 5 August 2008 (UTC)
Thanks for your support. I think it makes sense to rename the article to Patentability of software. I am now planning to do so. Any objections from other contributors? --Edcolins (talk) 18:42, 6 August 2008 (UTC)
While, from a neutrality and clarity point of view, your proposed title is acceptable to me, I cannot see any possible way that it can be said to comply with Wikipedia:Naming conventions. Don't think changing it's a good plan. Sorry. GDallimore (Talk) 07:42, 7 August 2008 (UTC)
I don't see there's necessarily a problem with WP:NC. While the phrase "software patent" might get more hits on Google, "patentability of software" is arguably a slightly different topic, and a more encyclopedic one for this encyclopedia; and appropriately renders to secondary importance the charged semantic discussion of what should or should not be considered a "software patent". Jheald (talk) 08:58, 7 August 2008 (UTC)
Well, I won't complain if the page is moved... GDallimore (Talk) 09:26, 8 August 2008 (UTC)

Archive 1