This summer the Court of Appeal handed down its latest decision on the patentability of software in the UK, while a new standard licence was published for open source developers.

‘Open source’ software is software the programming code of which is made freely available for users. In contrast, the programming code of 'closed source' software, on the Microsoft model, is never seen by the user and is strictly confidential to the developer.

Champions of open source argue that it has clear advantages over closed source. It is generally free to download. Users can pick over the underlying code and correct the bugs that always crop up in software. Users can add to and expand the software functionality, in which case they will make their additions available as free resources. The availability of open source software has mushroomed, and it is tipped by some to become the dominant mode for consumer software.

Practitioners need to be very careful, though, not to make the mistake of thinking that open source software is rights-free (more on that further on).

If the open source movement is the group-hug of the IT world, software patents are its 600lb gorillas. Opponents to software patents argue that far from encouraging innovation and investment, they stifle it. Examples such as the well-known Amazon ‘1-click’ patent are cited to show how patents might be used to impair the development of new applications and to tie up the internet (it was claimed that the 1-click patent could be used to stop consumers making purchases online with a single click; although most of the 1-click claims were not, in fact, granted).

Some of the criticism is probably fair. The sheer speed and scale of the global advance of software means that it can be hard to tell whether an application for a software patent is novel and inventive, with the result that it is more likely that a software patent will be bad. But other critics go further, and challenge the notion that patents should be granted for new and inventive software at all. In 2007 the European Software Directive, a relatively innocuous piece of draft legislation designed to harmonise software patenting practice across the European Union, was abandoned following a campaign of opposition that bordered on the hysterical.

Software patents have flourished in spite of this opposition. In the US, software is treated no differently to any other medium for the purposes of patent grant. If the software invention ticks the boxes as being new, not obvious, and sufficiently described in the patent application, then a patent should be granted for it. In the UK and Europe, the position is different. The Patents Act 1977, derived from the European Patents Convention 1973, includes ‘computer programs’ in a list of items not capable of being patented (section 2(c); article 52(2)(c)). However, the 1977 act and the convention go on to say that ‘the foregoing provision shall prevent anything from being treated as an invention for the purposes of this act only to the extent that a patent or application for a patent relates to that thing as such’.

Those two words ‘as such’ indicate that there is not a blanket ban on patenting inventions that involve computer programs. But just when is a software invention just an unpatentable computer program, and when is it something more?

Software patents first came to be considered by the UK Appellate Courts two years ago, in a pair of cases known as Aerotel/Macrossan ([2006] EWCA Civ 1371; and see [2007] Gazette, 29 March). The Court of Appeal formulated a four-step approach for the UK:The most significant of these steps is probably the third: that if the ‘contribution’ is nothing more than a computer program, then a patent cannot be granted. On the basis of that step, post-Aerotel/Macrossan the UK Intellectual Property Office (UKIPO) instigated a policy of refusing to grant patents for any invention that was contained in software.

  • Construe the patent. What is the monopoly being claimed?;
  • Identify the actual contribution of the invention. Eg, what problem does the invention overcome, and how?;
  • Ask whether the contribution falls solely within the computer program exclusion. If it does, then a patent should not be granted;
  • Check whether the contribution is technical in nature. If not, then a patent should not be granted.

The approach caused a storm of protest and appeals from UKIPO decisions, and last summer the Court of Appeal had a second run at the question, in the case of Symbian Limited v Comptroller General of Patents [2008] EWCA Civ 1066. The case involved dynamic link libraries (DLLs). DLLs are a common and very useful tool, comprising a means of storing functions that are common to a number of different applications, so that those functions only need to be stored once. When an application is loaded and run, the relevant functions stored on the DLL are selected and linked. A problem could arise, though, when additional functions were added to the DLL independent of each other, because these could be allocated the same identifying number in the DLL, causing incompatibility and malfunction. Symbian's patent application described a way of dealing with the problem without losing the advantages of the DLL system. However, the UKIPO rejected Symbian’s application on the grounds that the contribution was a computer program, and so fell solely within the computer program exclusion.

The Court of Appeal reaffirmed the four-step approach of Aerotel/Macrossan, but said that the UKIPO had construed the patent incorrectly. It identified the contribution as not solely an excluded computer program, because the computer program has the knock-on effect of the computer working better. The contribution was, if you like, the technical effect that the software had on the running of the computer. The patent would therefore be granted.

This decision will undoubtedly give rise to an increase in patents granted for software in the UK. Developers will simply make sure to emphasise the ‘knock-on’ or technical effect of their software in their patent applications.

So, after two Court of Appeal decisions and over a dozen appeals from the UKIPO, is the law on software patents finally settled? Not a bit of it. Just days after the Symbian judgment was handed down, it was announced that the President of the European Patent Office (EPO) had made a referral to the Enlarged Board of Appeal of the EPO of many of the key questions around the validity of software patents. The decision of the Enlarged Board will be fairly authoritative: it will constitute the test which the EPO will apply when considering applications for European Patents for software, and the UK Courts are obliged to have judicial regard to it (section 91 of the Patents Act 1977). The Enlarged Board will probably hand down a decision in within two years. The referral is to be welcomed – but still, it is bizarre that these issues are still in a state of flux so many years after the software revolution began.

Let us head back to the other end of the software development spectrum, open source software and copyright.

Far from abandoning their rights in the software, open source developers usually expressly assert their copyright in it, and furthermore impose license terms on anyone using the software. The key term is usually that any development of the open source code, or any programs into which the open source code is incorporated, must be licensed on the same open source licence terms. These terms include an obligation to make the source code freely available to third parties. The intention is to ensure that the benefits of the open source code are passed on, through third-party developers making the code in their developments similarly available to use, amend and share.

The mistaken belief that third parties can do what they like with open source programs is a common problem in software development. Programmers commissioned to develop a particular program sometimes take shortcuts, by incorporating an open source program to provide a particular function that they need, rather than developing that function from scratch. Or, a developer adds to an existing open source program, and then decides they would rather sell the program as closed source rather than making it available on open source terms. These are dangerous practices. If a program incorporates code from an open source code site, then unless it is distributed on the same licence terms, it may constitute a breach of the original licence, and entitle the original developer to an injunction and damages. Over the summer last year, the US Federal Court held that the original developer could restrain distribution of the developed program (Robert Jacobsen v (1) Matthew Katzer (2) Kamind Associates Inc, 13 August 2008), though the district court in the same case this month refused to grant an injunction on the facts. Furthermore, when a software development business comes to be sold, the owners may be asked to give warranties as to whether their software contains open source. If they cannot, the whole deal can be jeopardised.

By far the most widely used open source license is the General Public License (GPL). This summer saw the first update to the GPL for 16 years, GPL v3.

The key effects of GPL v3 are much as outlined above, but it is intended to have the following additional effects over GPL v2:

  • To prevent so-called ‘Tivoisation’, the practice of using open source software in hardware that is rigged up to stop users accessing the software that is running. That will be a breach of GPL v3;
  • To neutralise the effects of the Digital Millennium Copyright Act (DMCA) and the European Copyright Directive, in relation to open source code used in digital rights management (DRM) encryptions. DRM are software-based protection measures imposed by the copyright owner on the use or copying of files, and the DMCA and directive make it a crime to write or share software that can break DRM. GPL v3 states that DRM software that uses licensed open source code will not count as a ‘protection measure’ for the purposes of the DMCA and directive, and so it will not be a crime to break it;
  • GPL v3 includes a licence to use any patents owned by copyright holders that fall within the scope of the software;
  • Termination provisions for breach are automatic, instead of needing notice from the copyright owner.

Open source code issued under GPL v2 will not automatically convert to v3, but many open source developers will make such a conversion. For practitioners the key point is that, whenever you are advising in the IT field, you should always be mindful of whether open source code might have been incorporated into software which your clients are buying, selling or licensing – and the proper way to treat it at law.