GENERAL > SOFTWARE

FAQ - Patents on computer-implemented inventions

("SOFTWARE PATENTS")

Q1: Can software be patented?

A1: In short, the answer is twofold:

• NO, software seen as text in a linguistical sense cannot be subject-matter of a patent claim.
• YES, the functional ideas behind the text can be patented if they are of a technical nature, novel, inventive and industrially applicable.

In particular, software is a "legal hybrid" and has a somewhat ambiguous and janus-like nature:

• On the one hand, a computer program is simply a certain sort of text written in a "programming language" in accordance with a particular formal grammar.
• On the other hand, the primary source of value of a computer program is its behaviour or functionality, not its text.

Q2: What about the relationship between copyright and patent law?

A2: There are two fields of law, namely copyright law and patent law, which are mapped to the aforementioned different aspects of computer software, respectively:

• From a copyright-centred point of view, a computer program is seen as a text.
• From a patent-centred point of view, a computer program is seen as an expression of functionality or of dynamic semantics described by operational and/or denotational semantics, i.e. it is viewed as an expression of run-time behaviour.

Q3: What is the relationship between the software code, on the one hand, and the patent law, on the other hand?

A3: The run-time behaviour of a certain computer program cannot be deduced from its text ("code") alone. The dynamic semantics of the text largely depends on the processor which is assigned to execute the program. In fact, text and behaviour are largely independent from each other [01]:

• One and the same functionality can be expressed by different texts. For example, the texts of computer programs implementing the RSA asymmetric cryptography algorithm can look quite different when written in C++ or Java. Even when only a single programming language is used, many different texts representing RSA are conceivable.
• One and the same text can cause many different behaviours if run on different processors. For example, a program carefully written in C++ and implementing said RSA algorithm will perform RSA if run on a C++ processor, e.g. a general purpose computer equipped with a C++ compiler. If the same C++ text is run on another processor, say, on a general purpose computer equipped with a JAVA interpreter or with a PERL interpreter, other things will happen (usually an error message will be generated).

So, because the behaviour of a computer program is largely independent from its text, we recognise that the code text of a computer program taken alone is always insufficient for discussions of a computer program behaviour in the context of patent law:

• It would be a rather silly idea to include the text of the code of some computer software into a patent claim. Doing so would not unambiguously determine a functional behaviour - unless the matching processor is also specified in the claim language. However, this would be a strange practice. If a piece of software comprises an invention, the normal way of patenting the same would be to express the inventive functionality thereof in words and expressions of natural language. Of course, this natural language might be - or should be - an expert's language, not that of laymen.
• When it comes to patent litigation after a patent has been granted, evidence might have to be taken that a certain piece of computer program code text sold by an alleged infringer in fact reproduces the functional behaviour as defined by the claim. This can be done only by watching the behaviour after having matched the captured computer program code text with the appropriate processor. In terms of proper application of technical concepts, no text taken as such can ever directliy, i.e. more than in a contributary manner, infringe a patent claim. Hence, dealing with the pure text of the code of a computer program (be it source code or binary code) should - in terms of proper understanding of the underlying concepts - never be considered to be be more than an act of contributory infringement because of for proper determination of the infringing functional behaviour the assignment of a matching processor is always necessary. However, no case law has come to be known up to now clarifying any positions with regard to these matters.

Q4: Why all that public controversy about "Software Patents"?

A4: When searching the Internet using search words like "Software Patents" or the like, a huge amount of sites will show up debating the effects and justification of patents on computer-implemented inventions. The reason for this seems to have an economic as well as a pure political component:

• If a programmer does not "cheat" by simply copying code from an existing body of code written by anybody else, there is for sure no problem with regard to copyright infringement. Hence, most programmers feel comfortable with copyright law: If they are honest to themselves, there is no risk of accidential copyright infringement.
• With regard to patents, the situation is different. Infringing a third party's patent does not necessarily mean deliberately making use of a patented invention. Even if the programmer inadvertently makes use of some pratented invention there is a problem with patent infringement. Programming without having in mind the risk of potential patent infringement is in fact always a matter of negligence.
• No doubt, if a programmer sits down and starts programming, it might well happen that the code he is writing, when executed on an appropriate computer, may cause infringement of a patent (or even a bunch of patents) of a third party.
• Normally, producing software code is done by utilising a structured approach: In a first step an abstract solution for the problem to be solved is derived, and in a second step a code text is noted down implementing said abstract solution. During the first step, the risk of patent infringement should be carefully considered.
• Many larger companies making software-based goods e.g. in the mobile phone industry have set up an IP department providing adequate services to the programmers actually coding the software which is then exploited commercially. However, in particular some smaller companies or freelancer programmers may suffer difficulties in economic terms when attempting to allocate sufficient ressources to take care of patent infringement problems.
• In traditional industries, the shareholders' funds necessary to start a company have been quite huge due to required inventments into large industrial facilities and machinery. In such cases it is in general feasible to set up a well-stuffed legal department taking care, inter alia, of patent law matters. On the software market, to the contrary, there are a many smaller companies (SMEs) and even freelancers which are much less funded. In such cases, any duty to extensively deal with legal matters may appear to be an unbearable burden. However, there is no reason perceivable why SMEs should be exempted from observance of IP law.
• During recent years, a movement of individuals and NGOs has put forward a more fundamental criticism of the patent system. One of the more profiled figures of this scene is Richard Stallman. Prominent organisations supporting this sort of criticism are the Free Software Foundation as well as the Eurolinux coalition. The major goal of these activities seems to be a fundamental change of patent law so that no programmer on earth has ever to worry about patents. In effect, a small circle of activists supported by a huge number of supporters appears to be very busy in lobbying against the patent system as it stands now.

Q5: What is the most important purpose of the patent system today?

A5: Patents are instruments to secure fair market values for non-obvious technical problem solutions ("inventions"). This is also valid regarding such inventions for which the realisation requires the usage of a programmed computer or computer network:

• Patents are closely related to the concept of a "market" which in turn makes sense only within capitalistic economic structures. Therefore, lobbying against modern global capitalism can also be camouflaged as a battle for the rights of programmers against the patent system. So far it would perhaps not to be deemed frivolous to say that the contemporary anti-patent movement also has some aspects directing to a more general critics of global capitalism.
• No majorities are visible in politics for such destruction of the current patent system. Moreover, there are international treaties preventing single nations from doing so on a national basis, in particular the TRIPS agreement in the context of the World Trade Organisation ("WTO")

Q6: Why is the concept of "computer-implemented inventions" necessary?

A6: There are no patents claiming computer software in terms of a fabric of code. What really happens is that there are patents covering a device or a method identified by some defined functionality. A programmed computer running a certain software exhibiting the same functionality may then well constitute a patent infringement:

• Hence, albeit software as such cannot be subject-matter of a patent claim, running a computer with software may fall within the scope of a patent.
• Software is not the core of any patentable invention but means for its technical realisation.
• The concept of "Software Patents" has been proven highly effective for populist campaigns of the anti-patent critics but is is in fact grossly misleading. The truth is that there are numerous patented inventions which are implemented by means of a computer, and the term "computer-implemented invention" makes this clear.

Q7: How could theoretically the effects of the patent law on the software business be eliminated?

A7: The ultimate goal of the lobbying anti-patent activists, namely that programming and using of a computer can under no circumstances constitute a direct or contributory patent infringement, would be achievable only if patenting of each and every invention dealing with the processing of signals would be banned irrespective of whether it utilises a computer or some other dedicated circuitry:

• This conclusion is based on the universal interchangeability of software-controlled computers and dedicated circuitry.
• The usual Von Neumann universal computer is a finite realisation of the abstract Turing Machine, and it has been proven that each and every computable function can be computed by a Turing Machine. Hence, a dedicated circuitry having an output signal being a function sof some input signal can always be emulated by a Turing Machine.
• Therefore, only a full amputation of all potentially computer-implementable signal processing inventions would help to make Mr. Stallman and his fellows happy.
• This would effectively mean to abolish the patent system altogether with the exception of chemistry and mechanics.

Q8: What about the relationship between Open Source Software ("OSS") and the patent system?

A8: Distribution or use of OSS in a commercial environment may constitute an infringement. Insofar the producers, distributors and users of OSS are well advised to take care of the effects of the patent system:

• Nevertheless and irrespective of ideological brawls, OSS is an important building block for safe e-commerce in particular and for a workable information society in general. Hence, all due care should be exercised to avoid destruction of the OSS scene by means of patent law.
• Any problems caused in the context of patents in the field of IT, in particular in conjunction with OSS, might be solved by carefully redesigning the effects of granted patents, preserving freedom of expression when software code is exchanged over the internet as long as the computer program code isn't matched with the desired processor for production operation.

Literature:

[01] Pamela Samuelson, Randall Davis, Mitchell D. Kapor, and J.H. Reichmann: "A Manifesto Concerning The Legal Protection of Computer Programs"; Columbia Law Review, Vol. 94 (1994), pages 2308 to 2431.

THE DISPUTE ON "SOFTWARE PATENTS"

EXTERNAL LINKS (OFFSITE):

Please read the Disclaimer & About This Website (Pflichtangaben gemäss TDG) section. Feel free to contact PA Axel H Horns via e-mail horns@ipjur.com. BEWARE: DO NOT SEND CONFIDENTIAL INFORMATION UNENCRYPTED VIA E-MAIL. USE OF ENCRYPTION SOFTWARE IS HIGHLY RECOMMENDED. PA AXEL H HORNS IS PROVIDING SUPPORT FOR ENCRYPTED E-MAIL MESSAGES USING PGP OR PGP COMPATIBLE FORMATS. THE PGP PUBLIC KEY FOR PA AXEL H HORNS IS AVAILABLE HERE. THE GnuPG PUBLIC KEY FOR PA AXEL H HORNS IS AVAILABLE HERE. Dipl.-Phys. Axel H Horns is a Patentanwalt (German Patent Attorney), European Patent Attorney as well as European Trade Mark Attorney. In particular, he is Member of:
	VPP