For the half-year to 30 June 2014, the IPKat's regular team is supplemented by contributions from guest bloggers Alberto Bellan, Darren Meale and Nadia Zegze.

Two of our regular Kats are currently on blogging sabbaticals. They are David Brophy and Catherine Lee.

Thursday, 30 April 2009

Don't change European software patent rules, say CIPA

A press release from the Chartered Institute of Patent Attorneys (CIPA) in the UK today affirms its position on computer software patents in the light of the European Patent Office's Enlarge Board referral in Case G3/08 (see earlier post here). The deadline for submitting amicus briefs is today. Says CIPA:

"... the Chartered Institute of Patent Attorneys has urged the EPO’s Enlarged Board of Appeal to reject calls to make radical changes to the rules on patenting computer programs. Changes to established practice would cause the kind of confusion that exists in the USA ...

Chairman of CIPA’s Computer Technology Committee, Simon Davies, says that any major changes to the established practice “would create confusion in the computer industry throughout Europe.” [Quite true -- but failure to change current practice appears to have created a present state of confusion too. Which is better? Our current, known, confusion, or future unknown confusion?]

“Although computer programs as such are not inventions, under EPO rules,” Davies explains, “this does not necessarily exclude patent protection for an inventive product or process which is implemented by a computer program. Over many years, the EPO Boards of Appeal have developed a carefully formulated approach in which patents are granted for a technical solution to a technical problem. This has now become well-established practice before the EPO in general.” [Is it agreed that the formula is one which is susceptible of reliably consistent application to any given set of facts? Does it need (i) to be left completely alone, (ii) to be gently fine-tuned or (iii) to be supplemented with further clarificatory guidance?]

According to CIPA, the situation at the EPO is in stark contrast with difficulties in the USA. For example, the State Street Bank decision from the US Court of Appeals for the Federal Circuit caused a big increase in the range of subject matter for which a patent could be obtained. However, the recent Bilski decision from the same court represents a potentially significant contraction. These two decisions have left US patent law in a state of great uncertainty [This may be true -- though some US patent experts disagree -- but it doesn't really address the problems in Europe].

There have also been problems in the UK, where the Intellectual Property Office has issued four different Practice Notices relating to computer-implemented inventions in the last five years. CIPA is not in favour of such frequent changes, as patents have a lifetime of up to twenty years. [Is there something missing here? It looks as though perhaps CIPA would be more in favour of frequent changes if the maximum patent term were different]

CIPA’s submission to the EPO states that ‘the current approach provides a sensibly balanced and legally coherent treatment of excluded subject matter that benefits both applicants and third parties. This approach should continue.’
The IPKat appreciates the need to limit change in the interests of stability and predictability, particularly when each change is followed by an empirical stage in which patent applicants and litigants have to pay for the privilege of finding out what those changes mean in practice. But Merpel wonders whether the crucial criterion is not that of either stability or predictability, but of whether the present system works so well that no change is justified in terms of functional utility.

"If it aint broke(n), don't fix it" here and here

44 comments:

Anonymous said...

Is CIPA aware of the story of Keith Beresfords bestselling "Patenting Software udner the EPC"? The first edition appeared in 2000. The second edition was announced several years ago - but the author himself told me that the law was in so much turmoil that he felt unable to write it - until perhaps better times.

Perhaps the happy few among the CIPA members developed a special skill to guess what the EPO will decide, but the law is far for clear.

And the law consists of just magic formula's, like the "further technical effect", originating from a word-play, rather than from the balance of interest that must eventually govern all situations where statutes are unclear and other sources (like the legislative history) do not help either.

Anyway, why is CIPA interfering at all? It is an institute of patent agents. Their interest is questionable: they make a business my drafting as many patents as possible, and to some extent, they prosper by unclear rules! Perhaps this is not a nice comment, but it is just too obvious ...

Klaus said...

Indeed, nothing is broken. The system works so well. No problems. Patent agents are happy. After all that is the very purpose of the patent system: To increase the revenue streams of patent agents and get fat examination institutions.

Anonymous said...

As someone who was involved in the CIPA response, a few comments if I may.

"Present state of confusion". Most people who work with the law on a day-to-day basis (specialist patent attorneys, EPO Examiners, the Board of Appeal) don't find it confusing.

(Unfortunately the UK courts probably may not fall into this category, but since they've devised their own approach, and after two years, nobody is yet certain whether it involves 3 or 4 steps, they hardly provide a model for clear thinking).

In fact, the current approach of the EPO Boards of Appeal almost goes out of its way to be straightforward.

I suspect that Merpel is confusing two different issues: (a) whether the current practice is confusing, and (b) whether the current practice is legally correct.

It is true that there is uncertainty at a broader level, but this is inevitable. Computer programs are excluded in Article 52(2)(c), but at the same time, computing and hence computer programs are clearly part of technology, and therefore included under Article 52(1). Everyone has their own interpretation of how to resolve this, but since the EBA can't change the law, they cannot remove the inherent uncertainty.

"some US patent experts disagree" [that Bilski has left US patent law in a state of great uncertainty]. I was in the US last month (April) and spoke to many attorneys. There was widespread uncertainty. Any US patent "experts" who think otherwise are deluding themselves. The large number of dissents in Bilski doesn't help.

"It doesn't really address the problems in Europe". I think it acts as notice to Europe that making significant changes to the scope of patentable subject matter at the judicial (rather than legislative) level normally creates more problems than it solves.

"CIPA would be in favour of frequent changes if the maximum patent term were different". This is a rather idiosyncratic interpretation. The point is that as a business, you make an investment in a patent, which has a potential timescale of 20 years. It is much harder to decide whether or not to make that investment if the legal situation keeps changing every other year. (This point is obviously true in general, it is not limited to patent law).

"Merpel wonders whether the crucial criterion is not that of either stability or predictability, but of whether the present system works so well that no change is justified in terms of functional utility".

No, that is not the crucial criterion. As Merpel recognises, there is an inevitable cost in changing an established practice. The crucial criterion is therefore whether the evidence demonstrates that any change would definitely provide enough benefits to overcome this cost.

It's also easy to forget that the EBA review is one of law not of policy - all the more so since this topic has twice been considered recently at a political level (EPC2000 and the Directive). It would be profoundly undemocractic if the EBA were to be swung by policy arguments that were not accepted by elected governments.

This is also important with respect to your two commentators who question why CIPA should be making submissions at all. The ERBA referral is a legal review based on a (supposed) divergence between two decisions from the Board of Appeal. Patent attorneys are experts in patent law so it does not seem inappropriate to comment.

In contrast, many of the submissions on the EBA web-site, especially those critical of software patents, have clearly never even read the cited decisions. It's therefore difficult to understand how they could have anything helpful to say regarding resolving any differences between these decisions (which is the actual task of the EBA).

twr57 said...

Of course CIPA has an interest. Doctors have an interest in pandemics. Most people however assume that the interest is in how to prevent them rather than how to cause them.

Anyone is free to question CIPA's motives, but it would be more interesting to challenge their arguments. Cynicism is cheap.

Matt Dixon (patent attorney) said...

If patent attorneys only wanted to make more money, they would be lobbying for a removal of all restrictions on patentable subject matter. In 1996, when the State Street Bank decision opened up the range of patentable subject matter in the USA patent filings went through the roof. I think that CIPA is more interested in a reasonably fair patent system that has a chance of maintaining public respect. The current system is not perfect and is easy to criticise, but I haven't heard any workable solutions that seem better or fairer.

MaxDrei said...

Many European patent attorneys routinely opine, on the threat to their client represented by the pending apps of competitors of their client. They get paid good money for that activity, the more so if their opinions are firm. Many EPA's draft and file oppositions at the EPO, as a matter of routine. They have an interest in clear criteria for finding a claim invalid. I second Matt Dixon in challenging the assertion that interest of patent attorneys is nothing more than "The more patent applications the better".

Gerry said...

While the two Treaty articles conflict, the issue seems to fold back to TRIPS, which seems to fold back to the then obsession US Govt had regarding protecting "IP" leading to the DMCA and other great advances in suing people for abusing garage door openers. As with all international treaties, TRIPS was a political rather than an economic process and as the Treaty Articles demonstrate, doubtless failed to cover all the bases. The USA situation has evolved since then, with the DMCA being trimmed back and indeed the whole question of the patentability of software back in the frame following Bilski.

While justice may be blind it does not seem to excuse blundering around with an obsession for neatness and consistency. In this state of legal uncertainty it seems to me that a better question should be to ask why software should be patentable at all.

There are two perspectives

a) what's wrong with copyright?

If you've done something neat you get more years of protection for less legal effort

b) technical effect (or its local equivalent)

I can describe (e.g.) a novel car braking system as a transfer function and so convert it to a mathematical equation. But from the perspective of protecting significant investment in technical development I suspect the interesting bit is the shape of the actuator, braking pad material or choice, weight and structural characteristics of other mechanical bits. And few of us would be having an intellectual debate about whether these things could be patentable. If the innovation is all in the software I'd be curious to know which line of code in the software programme didn't have prior art somewhere, somehow (there's a lot of it that been written over the years). A lot of libraries of mathematical functions have existed for a very long time and I'd really like to know which classes of transfer function have not been previously described.

But the debate that is not being aired is "Why so much interest in software patents?" I'm thinking long filenames here... amd I'm looking forward to the outcome of a prior art analysis http://www.openinventionnetwork.com/press_release04_28_09.php

It seems to me that software patents are a neat way of engaging in anti-competitive practices. And if you are going to defend an alleged infringement of a software patent you seem to need a lot of money. And that makes is a "big boys" game. In the Free Software world we now have to rely on "protection" broadly equivalent to young people having to join a gang (the OIN in this case, nice though they seem to be) in inner-city London for protection from other gangs. I'm not sure that having to rely upon "benign protection" is a great place to be.

The reality for small innovators is described very neatly in another area of "IP" protection.

http://tuxmobil.org/mobilix_asterix.html describes the efforts (in 2003) of designers of a mobile platform based on GNU/Linux to use the name "MobiliX" but of course they didn't have much money.

Contrast this with France Telecom who spent three years to take the very same plantiff all the way to the ECJ regarding the very same name. (search on this website http://www.ip-talk.eu/?cat=1)

And explain to me the difference in terms other than power and money.

Gerry Gavigan
Open Source Consortium

Gerry said...

Further grist to my mill is provided here: http://www.theregister.co.uk/2009/05/02/microsoft_open_source_eu_patents/

Paragraphs I particularly liked include:

[software patentability] "opens the way for expensive and opportunistic legal actions and ambulance-chasing lawsuits of claimed patent violations, which typically favor those with the lawyers and money to win"

"According to Red Hat, software patents in the US are more than twice as likely to be the subject of a lawsuit than other patents and account for one quarter of all patent lawsuits."

Anonymous said...

@Gerry:
As I understand the current case law, a car braking system for which only the software is new could get a patent if the software improvement reflects some mechanical principle and the prior art provides no hint to use that mechanical principle in a car braking system. It is (or should be) immaterial whether that principle is implemented using mechanical parts, using hard-wired logical circuits, or using a general purpose processor and software. The point is that the inventive step lies in the application of this mechanical/technical principle to a car braking system.

However, if it is obvious to apply this principle to a car braking system, then implementing it in software would normally not give an inventive step (given that it is obvious to implement part of a car braking system in software). Implementing the known principle using mechanical parts could still give an inventive step (provided that this implementation is claimed), unless the implementation required only straightforward measures.

Gerry said...

@Anonymous May 02

Thank you, your description was better formulation of my understanding.

However, my real problem is that I am trying not to foam at the mouth about the whole idea of software patents in the first place.

I was attempting to be sympathetic to the idea that there may be circumstances involving software in which technical innovation can be protected. However given that the software can only be a combination of implementations of (textbook) transfer functions I cannot see how the software itself can ever be patentable.

I understand that any absolute statement is vulnerable to rhetorical attack. I consider this attack is possible because in the absence of a specific case we are indulging in a combination of non-popperian argumentation (non-falsifiable) and also deploying the argument of the excluded middle "if I didn't think it was not technically innovative then I wouldn't seek a patent on it.

This goes nowhere to addressing why software patents and dealing with the might is right consequences of allowing them at all.

Anonymous said...

The Microsoft brief referred to by The Register doesn't seem to be listed on the EPO website yet. Apparently Microsoft is OK with the current practice, so they essentially agree with the Philips and IBM briefs.

Note that there are other briefs that actually do propose to extend patentability. For example, the AIPPI and IZGSE briefs.

The Register also refers to Canonical's brief arguing that Art. 52 should be interpreted in a "simple" fashion. The problem is however that any "simple" approach leads to the same kind of subtleties that we see in current case law once you start looking at the boundary of patentability. And there must be a boundary, unless you exclude all inventions from patentability that potentially could be partially implemented in software.

Note that if you interpret the exclusion as "claims to a computer program (product) are excluded", this would indeed be very simple, but it would leave no way to object to method and system claims to the program in operation and to the program loaded on a computer. Such claims can easily be formulated even without mentioning the word "program". (And most national laws on contributory infringement would extend this protection again to the software on a disk or transmitted over a network.)

Anonymous said...

@Gerry:
What do you mean by "(not) allowing them at all"?

Do you accept that there can be a patent-elligible innovation that has an essential part that happens to be implementable in software? Then you have to allow at least some "software patents".

If you are only concerned with "software itself", then maybe you mean the claims to a software program on a carrier. G 3/08 might conceivably reject those claims, e.g. on the basis that a program on a disk does not actually achieve any technical effect, even though it may have the potential to produce one when executing. But I would imagine you are equally concerned with method claims that aim to protect the series of steps that form the functionality of the program and apparatus claims that essentially claim the computer plus program. My point is that those can't simply be discarded on the basis that they do or might involve a computer program.

"if I didn't think it was not technically innovative then I wouldn't seek a patent on it."

That argument will normally not suffice to show technicality ;-)

Maybe it's good to have an example. Mp3 works because it is based on a model of the human ear. The idea of throwing out frequencies that a human can't hear anyway seems technical to me (which is not the same as new or inventive), and it seems wrong to me to reject patent protection only for the reason that mp3 can be implemented in software. However, if the technical principles were already disclosed in the prior art, mp3 should be obvious (lack inventive step), regardless of the exact sequence of programming instructions used for its implementation.

I understand that a patent on mp3 can be annoying for the open source community, but if the effects are really unacceptable, this should imo be solved by amending infringement laws.

Anonymous said...

Gerry:

However, my real problem is that I am trying not to foam at the mouth about the whole idea of software patents in the first place.Well, there's your (and a few other people's) problem. As long as the OSS community finds it so difficult to approach patents in a rational way, this is likely to remain a religious argument. What is truly so offensive about "software patents".

a) what's wrong with copyright?IMHO, plenty. For starters, the original scope of copyright is the protection of "artistic and literary works". It's difficult to see where software fits there: there can't be many who read through a few million lines of code for the solace of their souls. I won't exclude the possibility that there are such people who keep source code (never mind binary) as bedside reading, and possibly they are to be found among the OSS community, but I'm sure it's a small minority.

This, in my view, is important, because, unlike patents, copyright doesn't require disclosure. This is in my opinion one of the biggest problems in the IT area: because it has relied for so long on proprietary code (that is, protection through secrecy), there's an acute lack of documentation. The whole idea of the patent system is to encourage not only invention, but above all its disclosure. Its original ethos is in fact not at all dissimilar to that of the OSS movement...Of course, you can argue that the patent system has fallen short on this account, and I wouldn't entirely disagree with you, but there are ways to remedy that which would be a bit more rational than "Software patents?! OH NOEZ!"

Also, what is protected by a patent is hairsplittingly defined by its patent claims. What is protected by copyright is a lot fuzzier (and, I should add, as a result often a lot more profitable for litigating lawyers). Sure, copyright protects you against somebody making a copy of your disk. Not so much against using your original (and inventive) idea to rewrite his own code. And I won't even address the nightmarish potential complications of shared copyright between a multitude of cooperating programmers.

However given that the software can only be a combination of implementations of (textbook) transfer functions I cannot see how the software itself can ever be patentable.Well, I'm a mechanical engineer by training, and machines can only be a combination of implementations of (textbook) mechanical elements as well. This doesn't exclude that their combination can be new and inventive.

Regulation and control has long been one of the mainstays of engineering, starting at least from Watt's centrifugal governor. At first this was implemented mechanically, then hydraulically and pneumatically, then electronically. Electronic control systems were first custom-built circuits (first with lamps, then with transistors). Now, however, it's mostly general purpose programmable processors running on software.

Since its beginning, the patent system has recognised that an inventive control method can be implemented in different ways. For this reason, it has long allowed the possibility to independently protect, besides the control system as a physical device, also the control method as such.

Now, would you argue that a patent on an utterly inventive machine control method should not be enforceable against an infringer implementing that very same control method using a programmable computer, rather than cams, valves or lamps? That would be hardly fair or logical, would it?

However, there is a further question: method patent claims are in principle only infringed when the method is carried out. This is troublesome for those whose patented control methods are implemented with software because it would mean that only the ultimate user would be infringing. And if there's a golden rule in business it's that you don't sue your customers. Hence the push to be able to have claims on "computer programs" or on "computer-readable media carrying a computer program" as devices.

Finally, there's the ultimate question: where do you draw the line? I'm aware that few people, even among the OSS community, are arguing against the patentability of, say, a braking system control method, even when it's implemented as software. On the other hand, even fewer people, even among what you'd call "pro software patent activists" argue in favour of making a business scheme patentable just because it's implemented on a computer. In between, there's a wide grey area ranging from image or sound processing through memory access management to how data are presented in user interfaces. Unfortunately, the legislators have failed to set a clear line (not least because of the frenzied lobbying on both sides around the CII Directive), so now basically the EBA is being asked to draw the line in their place. I'm not sure it'd be wise on their part to accept...

MaxDrei said...

What a fantastically lucid and helpful anonymous contribution at 09.40 hr on Monday May 04. See how the value of a blog thread depends more on what people are writing than on who they are. Would we ever have had the benefit of that clarity of expression, if this blog were to be confined to contributers who have to reveal their true identities?

Gerry said...

@09.40 May 04,

What you have not chosen to address is the role of power and money in all of this.

You introduce the concept of fairness, may I ask you to address the concept to "Mobilix" and explain what happened there?

In the circumstance you pose (deploying the rhetorical devices I already describe), until the patent office awards and the litigation upholds I have no idea if there is an innovative combination of transfer functions implemented mechanically engineering that can be patented. As I infer it, in your description, if I can demonstrate how my set of transfer functions were produced earlier and patented for something else are the same as your later set, used for something else in a differnt domain (say resistor-capacitor-inductor rather than mass-spring-damper) then my patent trumps your patent application, and that doesn't seems "fair or logical" either.

Not that I appreciated that fairness or logicality were relevant.

It doesn't seem relevant that discovery required millions of lines of source code to be examined. Either I did it before you and published or patented or I didn't.

There may well be a golden rule about not suing your customers, but I have no idea what that has got to do with the rule of law, nor why the rule of law requires a workaround to fit your needs.

It seems to me that the rule of law has to meet the needs of society, and the current use of software patents seems to work against society, promoting instead anti-competitive behaviour, which can often only be addressed by groups such as Free Software. It was their preparedness to read millions of lines of source code, conduct large studies into prior art, or pursue competition cases to the CFI that has proved to be the most effective defence against "enclosing the commons" as large companies constantly seek to do, and other companies cannot afford to challenge.

This doesn't seem to be fair or logical either.

Liberal and democratic societies rely upon profit maximising enterprises to create wealth and enable redistribution. However the rule of law is not there to promote their interests to the exclusion of other issues.

The current situation with software patents is reminiscent of the creation of the Single Market, at which time car companies were pushing to make pattern parts illegal, luckily for the consumer/citizen they failed.

The ubiquity of software in every day life was a circumstance not envisaged by the patent system. It should not be possible to institute a private tax on the general ability to participate in modern society. The rule of law is to protect the weak from the powerful.

Anonymous said...

Gerry,

until the patent office awards and the litigation upholds I have no idea if there is an innovative combination of transfer functions implemented mechanically engineering that can be patented. As I infer it, in your description, if I can demonstrate how my set of transfer functions were produced earlier and patented for something else are the same as your later set, used for something else in a differnt domain (say resistor-capacitor-inductor rather than mass-spring-damper) then my patent trumps your patent application, and that doesn't seems "fair or logical" either.I'm afraid that you suffer some serious misunderstandings about how the patent system works.

First of all, at least if you file at the European Patent Office first, you can get the official Search Report, with a first Written Opinion, usually within 6 months. This SR is usually quite thorough.

Secondly, patent applications are published approx. 18 months after the application, and they are searchable online (the EPO actually offers possibly the best free patent search tool for patent documents of 80+ countries). It is generally a good idea to do your own searching to start with.

Thirdly, the state of the art includes anything that was previously disclosed to the public, in any form whatsoever, not just patent applications.

The case you present is far too sketchy to determine the outcome:

If the first patent just disclosed your "set of transfer functions" in that particular form in a different field, for a different purpose, you may still get a patent for their use in your field if it's inventive. As for whether you'll be free to use it, it's a completely different matter that will depend on whether you fall under the claims of the previous patent, not just whether it is anticipated by something that just happened to be disclosed in the description.

You go on about "money and power", but, as I tried to explain in my previous post, the copyright system is just as subject, if not more, to abuse and distortion as the patent system. Of course, it isn't a good idea to file a patent on your own and then try to enforce it in court without the assistance of an (expensive) specialist lawyer. But you'll be about as hopeless acting pro se in a copyright case, let me assure you. You cite a trademark case, which is yet another form of IP, and a straightforward one compared to patents, never mind copyright. Frankly, I'm not a trademark specialist, but this seems to have been a borderline case. I must answer with another question: Why was this OS group so unreasonable as to insist on such a futile and expensive legal battle on something which arguably had no real value for them?

All the drawbacks of the patent system which the FFII and others protest again, apply just as much to "old tech" as they do to software. These are quite old debates that have gone on for a couple of centuries and which have led to the compromises that shape the current patent system. It isn't perfect, but then, nothing human is.

Anonymous said...

Gerry, notice how many times you use the word "seems". The viewpoint you have, from which everything "seems" to be polluted by "power" and "money", is evidently shared by many like-minded people, by which I mean people with a mind like yours. Then there's the community of patent law experts, academics, patent agents representing freelance inventors, patent judges, bloggers, those sort of people, who worship the "Rule of Law" and are experts in patent law, yet don't have your viewpoint. To them, it "seems" like we already have, in Europe mind, a patent law on CII that's pretty damn good already. And sorry but you can't reasonably dismiss everything members of that group write using nothing more than a hand-waving assertion that it's irredeemably polluted by "power and money".

Tell us about the Mobilix case. I thought that was trademarks.

Anonymous said...

Anonymous May 04, 2009 9:40:00 AM.

For starters, the original scope of copyright is the protection of "artistic and literary works".

The programs for computers were definied as copyrightable works in 70's. So it was almost at the moment when programs for computers have appeared (almost since the beginning of IT). Copyright has established position in IT.

"Also, what is protected by a patent is hairsplittingly defined by its patent claims. What is protected by copyright is a lot fuzzier".

Not at all. The source code is strict mathematical formulation of algorithm in semi-natural language. That was idea behind introduction of c++, java... - semi-natural description of code writen previously in assembler. It may be complex for non-programmer (not specialist) but discloses algorithm in detail. If somebody doesnt know what programming language constructions mean then he is not a programmer. Claims almost never can disclose algorithm because they are to general in formulation. If you would like to patent engine and you would not describe how its magnets works in detail then would you deserve on patent? (just you would write "magnets that rotate rotor" and thats all - something like this is happen when you try to describe complex software in a few claims - have you disclosed invention?)

"Well, I'm a mechanical engineer by training, and machines can only be a combination of implementations of (textbook) mechanical elements as well. This doesn't exclude that their combination can be new and inventive."
It is main difference between programming and machine construction. When you construct machine you have technical problems with physical parts (even of forms taken from textbooks). But when you write program you compose abstract entities - math equations. The mechanics is not similar to programming. More similar is music composition when you simply compose also abstract entities - scores. You may then play them on some device - but still the result of your work (software) is in abstract space.
ipod may decode scores and do physical results in exactly the same way as computer may decode instructions and do some physical results. Therefore software is storage independent as music. You may store it on paper, CD-ROM or paper card with printed holes. Software is also processing unit independed because you may construct mechanical processor, electronical processor or quantum processor maybe in the future (on all these devices the same software can be compiled and executed in the same way). The music of course may be 'inventive' in this (artistic) way but until now nobody want to patent scores combinations.

"Since its beginning, the patent system has recognised that an inventive control method can be implemented in different ways."

Control method is something different than data processing algorithm (equation). Process mastering requires operation on physical objects, needs timings so handles physical forces. Data processing algorithms operates on abstract data (not things).
There are some bussines methods patents (abstract ones) but these are US inventions.

"Now, would you argue that a patent on an utterly inventive machine control method should not be enforceable against an infringer implementing that very same control method using a programmable computer."

There is a difference however between data processing and machine controling as presented above. When you control machine using computer you have data processing part (software) and steering mechanism (e.g. I/O port). The point is that data processing part (equations, math) is not patentable if it does not work in conjunction with steering mechanism in novel inventive way so to achive novel inventive technical effect. So graphics card and its interface (even if partialy realized by software) may be patentable as invention. However when you display data then you merely use graphics card functions. The boundary here is slight between hardware, software supporting (steering) hardware and bussiness software using hardware (so software operating on abstract hardware and not couping with details how pixel is switched on LCD should not be patentable - the same is with any compression algorithm) therefore the problem with CII (how to define boundary between graphics card control and data processing using card, here it is necessary analysis of particular functions of software realized at different levels of API). There is also a feasibility problem because the comparison of two software products is more like comparing of two different moon rockets than two valves if it is going about complexity.
To be compatible with mechanics: if you have car you wouldn't patent twice the same car once as method and device driving forward and next time as method and device driving backward, wouldn't you? Especially if you would have that descibed in the car instruction.

"Hence the push to be able to have claims on "computer programs" or on "computer-readable media carrying a computer program" as devices."

You want to compare math to device. It is the same as you would call music (on CR-ROM or not) a device. Software is not a stream of electrons in circuits. It is also a medium representing some logic. In the same way your thought is not the same as conducting neuron in your brain. Your neurons can form thought but the same thought may be formed in the head of another person with different neurons and electron currents composition. Thought is and abstract idea and software, too.

Gerry said...

@Anonymous whoever you are

As I tried to set out at the start of all of this, in a situation where circumstances make cases, any attempt to be specific is stymied by the absences of facts, hence my use of "seems". We are all engaging in rhetoric.

My impression is that the responses to my posts are veering to the ad hominem... whatever.

The role of power and money are relevant in any political/economic system, On the one hand Lord Justice Sir James Matthew on the other Thomas Jefferson. Alternatively one might consider "who-whom?" Lenin or "How many divisions does the Pope have?" Stalin.

As for Mobilix, as I thought I had mentioned, it was a different area of "IP" protection. I was unclear about why France Telecom succeeded where the GNU/Linux group failed. I could not spot the difference in facts, merely the ability to engage expensive legal process.

And I am happy not to be anonymous

Gerry Gavigan
Open Source Consortium

Anonymous said...

Mp3 works because it is based on a model of the human ear... anyway seems technical to me.

Then you could also patent Newton law: F=m*a
It bases on mathematical model of physical object (any).
The same would be true for Einstein General Theory of Relativity which calculates forces in space time curvature and all the rest of models in physics. All these mathematical models examples of physical objects seems to me also very technical and related to real life. But nobody did it. Why? IMHO becuase mathematical model of physical object is not the same as physical object itself. It is pure math, data processing + abstract model not invention.
IMHO, the difference between invention and data processing is if device containing implemented algorithm does something more than producing pure information. This is especially for computer that has been designed as general computing device - this idea 'as such' has been patented probably on 50's of the past century. So by mere executing program you use existing from years invention.
In my opinion the value that brings mp3 is in new method of a sound wave (information) presentation. However it has not direct impact on computer disk or sound card way of working. Software that play mp3 simply uses abstract disk and sound card I/O to get data process it and send data to sound card. It does not change a way how these devices in computer works. It is data processing technique like calculation of derivatives however more complex (but it is going about nature not complexity) so simple computer usage here. If such algorithm would change a way how disk or sound card works then IMHO it would be invention. For example if you using software would convert you printer into a toaster. It would be new CII.

Anonymous said...

@Anonymous of 10:28PM
"Claims almost never can disclose algorithm because they are to general in formulation. If you would like to patent engine and you would not describe how its magnets works in detail then would you deserve on patent? (just you would write "magnets that rotate rotor" and thats all - something like this is happen when you try to describe complex software in a few claims - have you disclosed invention?)"

The claims define the scope of protection. The application (description, drawings) must disclose the claimed invention in a way that enables the skilled person to carry it out (Art. 83 EPC).

So the EPC already prohibits the grant of a patent for an invention that has not been disclosed. However, this is not a requirement of the claims, but of the application (mainly the description).

To be more precise, the requirement also has an effect on the claims. The claims should not claim more than what is made available to the public. "The patent monopoly should be justified by the technical contribution to the art", T 409/91.

"The music of course may be 'inventive' in this (artistic) way but until now nobody want to patent scores combinations."

The same is true for software. Software only contributes to inventive step to the extent that it implements a technical principle. The EPO case law does not allow patents on pure data processing that does not somehow solve a technical problem in a technical way. Those that attempt to patent musical scores will be asked to convincingly explain what technical effects are achieved. There is no difference in approach.

"The point is that data processing part (equations, math) is not patentable if it does not work in conjunction with steering mechanism in novel inventive way so to achive novel inventive technical effect."

Exactly! And you're correct that it is not easy to draw the line, but that's life... As case law stabilises more and more cases will become clear cut.

Anonymous said...

Unfortunately I don't believe that in the field of CII the border between what is patentable and what is not will become clear cut in the future. It seems to me that the dividing line is fractal like, so the closer you look at the border the more details you recognize. For this reason, I don't think that the EBoA can help in clarifying this fundamental issue and I believe that the referral will be dismissed as inadmissible, inter alia because there are no diverging decisions but only an evolving approach, completely normal in such a difficult field.

The situation is particularly difficult for patent examiners having to decide whether granting or not a patent in short time. But there are no clear cut rules that are simple to apply and I doubt there will ever be.

MaxDrei said...

In a different context admittedly, one of the learned patents judges (Robin Jacob, I suspect) said something about a "puzzle at the margin" being insufficient as a basis to find a particular claim invalid. I think that assessment serves just as well, as a response to the criticism of the evolving EPO jurisprudence on CII subject matter generally.

Anonymous said...

The border will never become clear cut. But the number of border cases should become less. For example, business method claims can hardly be called border cases anymore.

However, it would help if the various Boards start to draw one line. A decision like T 1446/04 appears to me to completely contradict the line of Board 3.5.01 that effects on the human mind are non-technical. Why was the issue of technicality of "eine bessere und schnellere Erkennbarkeit" not even mentioned?

MaxDrei said...

Sorry, that last anonymous, at 11.38 hr. Am I missing the point here. I had not realised that "effects on the human mind" and "a technical way to solve an objective technical problem" are mutually exclusive. Haven't read the case yet, but meanwhile, are there any other readers who are satisfied that the treatment of Art 56 in T1446/04 is inconsistent with the progressively sharpening line of TBA Art 56 jurisprudence?

Anonymous said...

@Anonymous of 12:51AM
"Mp3 works because it is based on a model of the human ear... anyway seems technical to me.

Then you could also patent Newton law: F=m*a
It bases on mathematical model of physical object (any)."

You must be aware that many patents are based on "forces of nature", such as the one described by Newton's law. The principle that makes a balance scale work is technical, right? Is the idea of throwing out frequencies that the human ear can't hear technical? (Don't confuse with novelty or inventive step.) Does it then matter if you implement this in hardware or software?

If you mean that a claim to mp3 should only be patentable if it includes a step of recording or reproducing the actual sound, then fine with me. Such a step might be necessary to actually achieve the technical effect in combination with the mathematical model. (However, I tend to think that a limitation to processing of sound data should suffice.)

@MaxDrei:
See e.g. T 1567/05, point 3.5:
"Although relating to technical phenomena, the stress values are mere pieces of information aimed exclusively at the human mind. It follows that also the features determining the kind of information displayed do not contribute to an inventive step."
and see point 3.4 and its reference to T 119/88, maybe even more relevant when comparing to T 1446/04. I would therefore at least have expected some discussion; it shouldn't be the case that issues that brake or make a 3.5.01-case go completely unmentioned in a 3.5.04-case. Maybe the Board did not feel obliged to go into this point because of the nature of the opposition procedure, I don't know.

Of course the "effect" of mp3 is in a way also on the human mind (or ear?). Mp3 distorts sound, but in a way that humans don't notice (too much).

Anonymous said...

I consider the test "happens in the human mind" completely irrelevant for assessing the technicality of an invention.

If I remeber correctly, this line was put forward eg in decision T 1121/02 (that case concerned a fence on which lines were drawn that induced fear in animals so that they stayed away from the fence. I think that the marks were black and yellow, similar to the tail of a bee, and that cows reacted on these signals). The board refused the application on the ground that the only "technical feature" was a painted fence and equated this to a fence on which the words "no trespassing" were drawn.

I doubt however that anybody would find a chemical shark repellent non technical, although no one knows why exactly the sharks stay away (the same is true for an ultrasonic mole repellent). I doubt that the board's decision would have been the same if the fence had flashing lights on it which made the animals stay away (or would this also be considered non technical?).
There are several inventions which rely on an effect in the mind of the observer (the cinema for example) and nobody ever considered them to be non technical.

Anonymous said...

"... works because..." just to continue

Compression algorithm applied to increase memory (disk) storage capacity may be considered here as patentable. It affects way how device working and enhances its capabilities.
However it does not exclude using the same compression algorithm in other data processing technique (so for the other purpose than storing smaller amount of data on disk) like sound wave visualisation.
IMHO using compression algorithm in conjunction with drive to increase store of drive may be patentable as "inventive device to store more data of some kind". It is the same as cache algorithm that may speed up disk access.

To be honest invention as above can be easily circumvent simply by using other algorithm or simply use none. Therefore IMHO examiners should take care about reciving detailed description of logic deciding about inventive way of device working. Generalizations not disclosing method of compression in detail should be always rejected because they do not present solution.

So IMHO in the practice as I understand it and in the context for example of storage devices some disk access algorithms, compression algorithms could be patentable if it will be proven that disk will run faster or its capacity will increase.
So IMHO by implementing new kind of data sorting algorithm you do not have invention because you do not cause that processor or memory runs in improved way (as devices) even if you solve your bussines problem faster because of lower sorting algorithm complexity. However when you develop scheduling algorithm for processor scheduling streams of instructions then you may change a performance of processor for the set of tasks so you affecting processor performance in ordered and aimed way. So for me it seems that patentable are only algorithms having aimed influence on computer part(s) capabilities (if the impact is innovative in the sense than device would get some new capability not previously descibed in device manual like can store sound data in better way).

MaxDrei said...

Thanks, Anonymous at 1.26 pm. I agree immediately that these "is it technical" Art 56 issues are substantive enough that they ought to have been discussed. But if the representative of the Opponent doesn't force them on the Board, and the Board (if it notices at all) perceives them to be on the margin, and lacks any appetite to tackle them on its own volition, then the default is that the patent survives, no?

Do you know T0012/08 from 3.2.08, on game rules and what is technical. I was wondering whether it is "technical", or a Rule of the game, that the chance of encountering a game character that is a nocturnal animal (say, a fox) is greater at night.

Anonymous said...

@MaxDrei:
T 12/08 is interesting. I might agree with the Board that the idea of using time as an ingredient for a random generator is technical. (Unfortunately, this idea is also very well-known, but apparently not to the Board ;-).)

As I understand it, the fact that the randomness was used for deciding when game characters appear was considered non-technical. The only technical aspect was the implementation of the random event generator.

@Anonymous of 1:29PM:
It is debatable whether the effect in T 1121/02 was technical or not. Are the instincts of an animal technical? Maybe they are, maybe not. But now suppose the application claimed a shop window with certain features that can be shown to attract more customers. Now the effect is on the mind of the human, and for me there is no doubt that this effect is non-technical.

Flashing lights undoubtedly are technical, but it is another question whether an alleged effect that flashing lights deter a particular animal is technical. If this effect is non-technical, the invention becomes a mere aggregation of a known fence and known flashing lights.

Your example of the cinema is a bit different. The fact that the image is projected (and the way this is done) is certainly technical. However, what is projected (i.e. the content of the image) is completely immaterial from a technical point of view. A movie with certain story features that make the cinema attract more viewers should not get a patent.

In my view, movie features that sexually arouse humans are non-technical. But the answer might be different for movie features that sexually arouse panda bears.

MaxDrei said...

I now have T1121/02 and T1567/05 and note that both are appeals from ED cases represented by UK attorney firms. As to the stripey electric fence, I would have tried arguing that the objective technical problem was that the fence posts were getting broken because the cows stumble into the fence so often. Once a fence post is down, the cows just step over the no longer functioning electric fence. The solution was the stripes, which made the cows more leary of the fence, and less likely to stumble into it. As to T1567/05, the patent application was already refused, so easy for the TBA to confirm its refusal. As to T1446/04, I confess I don't yet see the justification to maintain the patent.

MaxDrei said...

Does anybody remember the "blue squash ball" case? The evidence proved that a player picks up the flight of a blue ball faster than that of the conventional black squash ball. Is that effect, in the brain of the human player, one that is "technical"?

Anonymous said...

So it seems that any software is CII because you may always argue that it causes less power consumpion of computer.

Anonymous said...

But even if in practice the stripes make the cows less likely to stumble into it, the deterring effect remains a crucial link. If this link is considered non-technical, any other effect further in the chain should not be able to restore technicality.

To me, deterring animals is sufficient for a technical effect, provided that the deterrence relies on some "technical" principle. Of course the anonymous of 1:29PM has a good point that sometimes the exact principle that makes an invention work is simply not (yet) known. In T 1121/02 the exact principle was also not known, but in the view of the Board the fact that it was an effect on the animal's brain was sufficient to conclude that it was non-technical. Interesting decision.

In any case, T 1121/02 nicely shows that technicality issues are not restricted to computer-implemented inventions.

Anonymous said...

@MaxDrei:
I did not know the "blue squash ball" case (discussed e.g. here).

It seems to me that a 3.5.01-minded Board would not easily accept the effect as technical. Maybe the argument could succeed if the effect could be shown to rely on the working of the human eye instead of the brain...

Somewhat related is T 619/02, e.g. point 2.3.2. According to the applicant the claimed method relied on subconscious perceptual associations. Board 3.4.02:
"Thus, the "mechanism" underlying the selection procedure of the claimed invention leaves the domain of "technicity" and enters the domain of subjectivity inherent to human perception, i.e. of what - at least presently - lacks an objective causal description to the extent proper to mechanisms of a technical nature. This conclusion is not altered by the fact that human perception phenomena may well be governed by complex neurological processes in the human brain that in turn are ultimately governed by chemical and physical processes. The perceptual associations in the human mind, and in particular those underlying the claimed method, generally depend on personal factors (cultural background, gender, age, past experiences, capacity to evoke dormant meanings and emotions, perception subjectivity, etc.) and vary, for the same person, according to the actual circumstances of the moment. Thus, irrespective of the degree of repeatability or reproducibility of the claimed method itself, the mechanism underlying the selection method, i.e. that which is taking place in the mind of the test person, does not belong to the kind of mechanisms that can be reproduced and repeated under the same or analogous conditions to lead consistently to the same or similar results with the degree of objective verification and reliability that is generally attributed to mechanisms of a technical nature."

However, the facts in T 619/02 are a bit different, since it is pretty clear that subjective choices are being made there. The improved visibility of a blue squash ball might instead be a fully objective fact.

@Anonymous of 4:37PM:
Less power consumption compared to what? It's always possible to come up with arguments, but they should also be convincing.

Anonymous said...

"Less power consumption compared to what? It's always possible to come up with arguments, but they should also be convincing."

Simply to naive version of software doing the same thing with greater complexity. E.g. I may patent quicksort because it execution consumes less power than execution of bubblesort.

Anonymous said...

"E.g. I may patent quicksort because it execution consumes less power than execution of bubblesort."

The argument fails already when a general purpose computer is taken as a starting point for the inventive step assessment. Compared to a general purpose computer, a computer (quick-)sorting an address list sorts an address list. Sorting an address list is not technical. The computer sorting the address list therefore lacks inventive step.

In general, I think it is unlikely that a technical effect can be based on the comparison of one algorithm to another algorithm. The effect should already be there before you do the comparison. The comparison is there for finding out which effects are caused (or improved) by the features distinguishing the invention from the prior art.

Of course you can say that any program, when executing on a computer, consumes energy. This is true, but it is not a "further" technical effect, since this is one of the normal physical effects of a program running on a computer.

Anonymous said...

"further" technical effect

Even if it consumes _less_ (improvement) energy because smaler number of procesor cycles?

So the cache managements system for disk invention not depends on type of algorithm used for cache management?

MaxDrei said...

The blue squash ball case featured by Foley and Hoffmann is based on a real case in the English court, where a professional squash racquets player gave affidavit evidence (on behalf of Applicant Dunlop) that a blue ball delivered a superior technical effect. The case was remitted to the Patent Office with a finding that the claim was directed to matter that was inherently patentable. Would the TBA have done the same? Who knows. It would surely have given the witness testimony scant weight.

Anonymous said...

My two contributions in this discussion: that the frontier on what is technical is fractal like (ie it becomes more and more complex the more you think on it) and the second that the test "happens in the mind" is irrelevant for deciding if something is technical or not.

I believe that the discussion showed that the subject is extremely complex and that it is nearly impossible to draw clear cut lines. The "is it technical" issue is far more general than software or CII and arises with bussiness methods, games and even in traditional domains like mechanics (eg fences). An interesting case of Board 3402 is T 619/02, in which the effect of odours for improving remembering a certain brand was claimed.

The test "happens in the mind" is in my view irrelevant, since in some cases a technical effect may be associated to it while in others not (my example of the cinema was aiming at the persistence of the image in the eye/mind which creates the illusion of a fluent motion while in fact only a succesion of still images is projected). A test which gives a different answer depending on the case is thus irrelevant if not plainly the wrong test.

What about the hypothetic case that by playing a Bach cantata to milk cows their milk production is increased? Apart of the issue of credibility (which has nothing to do with technicality) would this be seen as technical? I cannot see why specific sound waves should be treated differently than chemical compounds (eg an hormon for increasing milk production).

It seems that what is seen as technical depends a lot of the narrative skills of the parties.

MaxDrei said...

If the "narrative skill" of the parties will determine the boundary of what is or is not "technical" then we in Europe can all be grateful for the sheer number of opposition cases at the EPO, where cases get decided, by narrative force, on a level "preponderance of evidence" playing field. If "technical" is the optimal test, it will deliver, bit by bit, an ever-sharper line between what is patentable and what is not. Conversely, if and when the line starts to get fuzzier, DG3 will then need a better test. So far so good, can we say?

Anonymous said...

@Anonymous of 8:26AM:
"(my example of the cinema was aiming at the persistence of the image in the eye/mind which creates the illusion of a fluent motion while in fact only a succesion of still images is projected)."

My apologies for completely missing that point. I agree that making use of the illusion effect is technical. Wikipedia tells me that in 1912, "persistence of vision" was shown by Max Wertheimer to be brain centered and not eye centered. So it is definitely an effect on the brain.

"A test which gives a different answer depending on the case is thus irrelevant if not plainly the wrong test."

Absolutely.

You have convinced me that an effect that relies on how the brain functions may still be technical. So if the "effect on the mind"-test is to be retained, it should be refined. If subjective factors are involved, I would still be very hesitant to call the effect technical. However, maybe it should then be argued that the effect is not achieved over the whole claimed scope.

Anyway, I still believe that a case by case analysis can lead to the development of usable tests for (non-)technicality, as long as the Boards are willing to clearly state why they follow certain cases and deviate from others (and not simply ignore the issue as in T 1446/04, although certainly the nature of an opposition appeal may have been the reason there). At least the answer to "is it technical" is not dependent on the filing date (ignoring T 1227/05, r. 3.4.2).

Anonymous said...

@Anonymous of 8:01PM:
T 1227/05 has this to say about the "it runs faster / consumes less energy / uses less memory"-argument:
"3.2.5 In this context the board notes that the above conclusion cannot be drawn from the mere observation that a claimed method runs faster than a "conceivable" reference method (see point IV(d) above). As it is always possible to conceive of a slower reference method, a mere speed comparison is not a suitable criterion for distinguishing between technical and non-technical procedural steps. If, for example, a sequence of auction steps leads to price determination more quickly than some other auction method, that does not necessarily imply that the auction steps contribute to the technical character of the method (see T 258/03)."
In T 258/03, the Board reasoned that a change of the business scheme that reduces the need for data transmission was not a technical solution of a technical problem, but a circumvention. Maybe the same works here: purely algorithmic changes do not solve but circumvent...

Maybe another way to look at it is to realise that speed / power consumption / memory usage all belong to the normal technical effects of a program running on a computer, so are not "further" technical effects. (But you have to be careful... using less power by executing the HALT instruction is probably a further technical effect, but of course not inventive.) As long as "improved" speed / power consumption / memory usage is only achieved through purely algorithmic changes, this would not improve a "further" technical effect.

So if two algorithms for a cache management system result in essentially the same system from a "technical" point of view, except that one algorithm is O(n^2) and the other is O(n), then I am fairly confident that this difference is non-technical. On the other hand, if the improved algorithm improves caching behaviour (which also happens to result in reduced energy consumption), the difference is probably technical.

Anonymous said...

The anon of May 05, 2009 1:29PM had truly remarkable foresight:
"I consider the test "happens in the human mind" completely irrelevant for assessing the technicality of an invention.
(...)
There are several inventions which rely on an effect in the mind of the observer (the cinema for example) and nobody ever considered them to be non technical."

According to the recent decision T 1749/06:
"4.2.3 The board is further of the view that the test "happens in the brain of the viewer", invoked by the examining division, is not useful for deciding whether a feature contributes to the technical character of a claim or not. Inventions such as the cinematograph are based on an effect which only "happens in the brain of the viewer", namely that the projection of a rapid succession of still images on a screen creates the illusion of fluent motion. Although the illusion of perceiving a real action is only created in the viewer's brain, nobody would seriously contest that the cinematograph is an invention based on technical features."

One would almost get the impression that the Board reads this blog.

Personally I still have some doubts, essentially based on what the Board then adds:
"It may be argued that the apparatus of the cinematograph comprises technical elements with the ultimate purpose to create the illusion of motion in the viewer's brain. However, in the present case the dark and light stripes are also technical elements which contribute to creating the three-dimensional illusion."

If we assume for the moment that visual illusions are non-technical, it still seems likely that a cinema projector that implements that illusion involved an inventive step at the time of its development. So the feeling that the cinematograph must have been patentable does not necessitate that visual illusions must be considered technical. On the other hand, if creating a visual 3d illusion using a particular pattern (the subject of T 1749/06) is considered non-technical, then a technical implementation of this visual illusion using technical tools such as a known mobile phone display will normally be obvious.

Subscribe to the IPKat's posts by email here

Just pop your email address into the box and click 'Subscribe':