Patents, Open Standards and Open Source
For starters, the EU announced a definition of Open Standards that excludes royalty licensed, patented technology, then saw its patent friendly "Directive on Computer Implemented Inventions" delayed in the nick of time.
Meanwhile, one arm of the UK government (the Patent Office) was busy promoting the EU software patents directive, while another (the Department for Education and Skills, DfES) was fighting just the kind of patent that the directive would allow.
"Then last week, one archetypal big business, IBM, made the surprise announcement that it would give free access to 500 of its patents to anyone working on an open source project. Which contrasts rather sharply with a statement of Microsoft's boss Bill Gates to the effect that anyone opposed to royalty licenced intellectual property rights was some kind of communist."
The reason the question of patents matters to any field of technology, and why it has caused such a stir recently, is that it pits two opposing views on innovation against each other. One group believes that progress in ideas or 'art' in the wide sense depends on the guarantee of a secure money stream if the idea or work proves successful. For software patents, that would mean that no-one would be prepared to stick a lot of resources in, say, developing the idea of transmitting student record information over a wireless network (the patent the DfES was challenging), unless it could be guaranteed that everyone who implemented that basic idea could be asked to pay.
In the opposite corner, people believe that the heart of progress is not the reward, but the process of transmitting ideas. That is, ideas rarely if ever pop up in a vacuum, they practically always build on the ideas of others, and are reflections of the state of the art at the time. Putting restrictions on the use of ideas at best hinders that progress, and at worst allows a small number of people to profit from the work of others. From that point of view, the company that patented transmitting student record information over a wireless network, neither invented student record systems, nor wireless networks nor the process of student registration. They merely brought the three together and were the first to file it at a patent office.
While this battle is characterised by some as the pro-Intellectual Property Rights (IPR) versus the anti-IPR crowds, reality is a little more subtle than that, particularly where the difference between copyright and patenting is concerned. Specifically, where copyright licencing is concerned with a creator's control over the expression of an idea, patent licencing is about control over the idea itself.
Few would dispute the right of a creator over the expression of an idea. Even the staunched believers in the transmission of IPR in a common pool —the 'copylefts'— rely on copyright law to make sure that their expressions are freely available to anyone via licences such as the GPL or Creative Commons.
The rights of a 'creator' over a broad idea, however, is a different matter. Once the patent has been filed and its licencing enforced, additional restrictions are imposed, whatever your relation to the patent holder or his technology. Hence the fact that no open source licence allows the inclusion of technology that is known to be subject to a restrictive, royalty bearing patent licence: the restriction of a typical patent licence is diammetrically opposed to the guarantee of the open source licence- the unrestricted flow of ideas.
Patents and Open Standards
The anti versus pro software patent debate also has major consequences for the process of setting interoperability standards. Inclusion of someone's known patent in a standard immediately locks the open source competition out of that market, and often discriminates against smaller commercial software houses and newcomers.
Since there are so many patents, and since there application can be so broad, and since the enforcement of patent licences is so uncertain, the 'big boys' tend to play a game of patent poker with each other. Either to make life easier for themselves and their competitors by agreeing deals that allow mutual use of whole patent portfolios, or else get strategic advantage by the threat of swingeing licencing conditions or the threat of litigation. A small company, with just a few patents, and no expensive packs of attack-lawyers stands little chance in that game.
Standards can amplify that inequality by practically mandating the agreement of a patent licence as a condition for entry into the community for which the standard was designed. For that reason, standards bodies such as the World Wide Web Consortium (W3C) specifically rule out the use of known, patented technology with royalty licencing from their standards. Others, such as the Motion Pictures Engineering Group (MPEG) in the International Standards Organisation (ISO), do allow patents.
The European Interoperability Framework
The EU realised in 2002 that a way had to be found to allow electronic data and services to flow freely between public administrations, enterprises and citizens across the Union. A beginning of a solution, the European Interoperability Framework, was announced a couple of months ago.
At present, the core of the framework consists mainly of a list of recommendations and principles. Most of the principles are just common sense; that the interchange of data should be secure, accessible by all people of all abilities and midful of privacy. Others are typical for the EU: the requirement for multilinguality and subsidiarity. The latter means that the framework doesn't replace national measures such as the UK's eGovernment Interoperability Framework, but sits on top of those, where needed.
The last few principles are notable, however. One is to "Assess the Benefits of Open Source Software", which are cited almost exclusively in terms of the use of open source software in setting and promoting open standards.
Even more notable is the principle of the use of open standards, and how they are defined:
The set represents an interesting political decision. On the hand, these principles exclude a number of standards that would otherwise be perfectly adequate. On the other, the European Interoperability Framework is a 'consumer' of standards with some clout of its own. It has clearly gone for using that clout to its and lots of other peoples' advantage: it puts pressure on the standard bodies to work to these principles were feasible. In turn, this could lead to a more open and free market than the more timid option of accepting any kind of standard.
- The standard is adopted and will be maintained by a not-for-profit organisation, and its ongoing development occurs on the basis of an open decision-making procedure available to all interested parties (consensus or majority decision etc.).
- The standard has been published and the standard specification document is available either freely or at a nominal charge. It must be permissible to all to copy, distribute and use it for no fee or at a nominal fee.
- The intellectual property – i.e. patents possibly present – of (parts of) the standard is made irrevocably available on a royalty-free basis.
- There are no constraints on the re-use of the standard.
EU Software Patent Directive
In sharp contrast to the principles of the Framework, the current state of another EU measure, the EU Software Patent Directive, places virtually no restriction on the patenting of any kind of idea in software.
The directive was designed to harmonise patent law with regard to software across the memberstates of the Union. After the EU parliament went through the original proposal from the Irish presidency, the scope of what could be patented in software was quite restricted. It couldn't be purely software, and it had to support a physical process- much like traditional patents on physical mechanisms.
The distinction between pure software and software designed to support a physical process is important because software is many respects closer to language than mechanical engineering. It is quite difficult to come up with a software patent that is precise in what it applies to; a bit like trying to define a Shakespearean sonnet such that it doesn't include every other sonnet some other poet might like to write. Not only is software that abstract, it is also quite complex in most any application; being able to patent any process or business logic in a computing system can quickly lead to an expensive mess that only a lawyer could love.
The EU council of ministers wanted a compromise that had virtually no restriction on what can be patented. That caused outrage among many scientists and software professionals, followed by some delay by the Dutch EU presidency and a bit of tinkering round the edges.
The result was added as an agenda item to be nodded through at a meeting of ministers of agriculture and fisheries last month. That's right, the people who deal with butter mountains and North Sea cod stocks. The Polish government thought that a bit too rich, and kicked it off the agenda. This opens the possibility of member states or the parliament to re-introduce some more substantial restrictions, but another EU presidency (Luxemburg, at the moment) could still try to pull a similar stunt.
The DfES and the Frontline patent
While the UK Patent Office was busy promoting the virtually unlimited variety of the EU patent directive, the DfES had to deal with the consequences of the kind of patent the directive would allow.
The patent in question covers "the process of exchanging student data over a wireless network using portable computers and transceivers in an educational environment". Which covers things like taking class registers using a laptop or any other device that is connected to a Student Record System (SRS) or Management Information System (MIS).
The patent was originally taken out by the owner of two small companies, Frontline and Bromcom, in 1992 to protect a proprietary system with custom terminals that managed to deal with the peak load of lots of teachers uploading lots of class registers at the same time. The product was quite successful during the nineties, but got more competition recently.
One such competitor, TASC, was sued for failing to pay patent licence royalties. TASC argued that the patent was invalid, because there was prior art: a college had demonstrated similar functionality using a file sharing architecture before the patent was filed. They did eventually settle out of court.
Given how easy, almost inevitable, it is to achieve the same functionality with a standard WiFi network and a standard browser connection to an MIS these days, the DfES got involved in the case and sought to overturn the patent. They were partially successful: the court ruled that the patent had to be amended in the light of the prior art in order to be upheld.
The DfES also claimed victory and told schools that they could use any MIS system over a WiFi network. The education secretary has since backpadled a bit on that one, probably not least because the prior art dealt with data storage, not wireless network technology.
Both the EU's own open standard principles and the robust opposition to the EU's software patent directive indicate that the widespread imposition of restrictive patent licences are not some kind of historic inevitability.
Though we may yet see more of the kind of messy and expensive lawsuits the DfES got involved in, there are signs of a less constraining trend from unexpected quarters: big business.
IBM has been the largest filer of patent applications in the US for a long time, with 3,248 patents granted in 2004 alone. Though it didn't actually do anything very much with that stockpile initially, it set up a licencing programme in the nineties that brings in close to a billion dollars a year today.
Yet last week, it announced the availability of 500 of its patents to open source projects under much the same conditions as specified by the EU open standard principles: irrevocably available on a royalty-free basis.
The paltry number of patents from IBM's great horde may make the move look like a cynical marketing exercise, but there might be more to it than that. By making its IPR available in this way, IBM can spread its technologies far wider and deeper than it could by restrictive licencing, thereby stopping rivals' technologies from getting a foothold. That way, it can drive more demand for what is IBM's real money spinner these days- services.
It has already tried similar moves with open sourcing the copyright on software packages such as the Cloudscape database technology, the Eclipse software development environment and large chunks of the Linux operating system. Given the pretty large impact of Eclipse and Linux in particular, they might be on to something.
If IBM's little experiment pays off, the patent move could turn out to be another step to an IT market that is not so much dominated by restricting access to IPR, but by the quality and depth of service provided.
The European Interoperability Framework document is available from the IDABC site of the European Commission, with a press release on the flosspols.org site.
More on the ongoing EU software patent directive debate is available from, amongst others, The Register. The main opponents of the directive are the Foundation for a Free Information Infrastructure.
The DfES and the Frontline patent case is outlined in the Education Guardian and The Register.
More about IBM's patent move is available from the New York Times (registration required), with a lively discussion about it on Slashdot.