There’s a blog entry over on the FOSS patents blog from software patent campaigner Florian Müller in which he comments on the German Federal Court of Justice’s (Bundesgerichtshof) “Dynamische Dokumentengenerierung” (=dynamic document generation) decision in a Siemens application on a method for producing a document display. Florian compares the German decision with the pending US Supreme Court decision on business-method patents (In re Bilski) and he comments that the decision means that “all software ideas are now potentially patentable as long as they are innovative from a purely formal point of view.” He seems to think that this represents a reversal of the trend in Germany “in which in which software patents were previously only considered valid under relatively strict criteria”.
The case is interesting since it is one of several cases in which the Federal Patent Court (Bundespatentgericht) has given applicants leave to appeal in order to clarify the law in Germany on computer-implemented patents. Last year the same German Federal Court reviewed their current case law in the “Steuerungseinrichtung für Untersuchungsmodalitäten” (commented in detail on German Wikipedia at this link. The Federal Patent Court is clearly trying to understand the boundaries of the exclusion of patents for computer programs, as codified in Sec. 1, Para. 2, No 3, of the German Patent Act – latest copy in English found here.
Both the new Siemens decision and the older one from 2009 follow the earlier case law. The test for whether a computer-implemented invention is patentable is to consider the technical problem which the invention attempts to solve. The court confirms its 1991 case law (“Seitenpuffer“) that an invention is technical when the invention relates to the functioning of a computer system and allows the interworking of the computer system’s elements. However, the court’s latest decision clarifies that that does not mean that all computer-implemented inventions are per se patentable. On the contrary the court emphasizes that merely using a computer program does not confer patentability. It is the recognition of the solution of a technical problem which is solved by a programmed computer which leads to a potential patent.
This is very similar to the European Patent Office’s approach on patents for computer-implemented inventions. – but differs somewhat from that offered by the UK courts in their Macrosan/Aerotel case.
The Federal Court of Justice goes on to state that once the technical problem has been identified, the test is to determine whether this technical solution is novel and involves an inventive step. It’s not just a purely formal examination. In this case the Court remitted the case back to the Federal Patent Court for further examination and review to see whether the criteria of novelty and inventive step were present.
This decision certainly is not a landmark ruling setting out a major change in patent law for computer-implemented inventions as suggested by the FOSS patents blog and picked up by Peter Zura’s 271 blog here and the techrights blog here. It’s rather a continuation of a long line of thinking by the German courts.
Software Patent, Germany, Computer-Implemented Inventions, Siemens
We reported some months ago on the German case concerning the relationship between patents and standards for recordable CDs. In that case, the German Federal High Court decided that the so-called “FRAND defence” could apply to patents relevant to standards. In other words, it is possible for a company implementing a standard to obtain a licence as long as it sets aside a reasonable licence fee.
The Dutch Court in the Hague explicitly rejected this defence and noted that its decision was different to that drawn by the German Federal High Court. The Dutch Court noted that there was no concept of “compulsory licence under cartel law”. Dutch law (like German law) has a specific provision in its patent code for the granting of compulsory licences. Such licences only come into effect when they have been granted by a court. The Dutch Court states that this is reasonable as nobody would otherwise understand the conditions under which the licence had not been granted.
There’s an implicit criticism of the defendant in the case that it had apparently never attempted earlier to take a licence to the technology from Philips. The Court clearly stated the defendant should have attempted earlier (on starting sale of the goods or becoming aware of the need) to obtain a licence under FRAND conditions and, if this attempt had not been successfuly, the defendant could have asked for a compulsory licence to be granted. The Court points out that at least a preliminary licence could have been granted in summary proceedings.
The decision highlights two issues of European Patent Law.
Firstly the need for a uniform patent litigation system to harmonise the decisions of the different national courts. The different decisions in the German and Dutch courts are based on different legal traditions and codes of civil procedure (and contractual law). The Dutch have traditionally granted complusory licences more readily than the German courts and so it’s not surprising that the Court in the Hague suggested this route, whereas the German Federal High Court suggested an alternative option.
Secondly the need for clarity under which licenses to patents required to implement standards can be granted. Most Standards Setting Orgnisations have a clear policy on intellectual property – licences will be granted on Fair Reasonable and Non-Discriminatory (FRAND) conditions. There are, however, difference on how to interpret these conditions – there’s little or no case law to enable a market entrant to assess its risks and It’s clearly impractical in a single European market to have to go to multiple courts to obtain compulsory licences (as the Dutch want) or post multiple bonds (as suggested in Germany).
The case has also been discussed on the EPLawblog here
CD-R, Patent licence, Standards, Orange Book, Philips