22 July 2010

What Do You Think About Software Patents?

Despite the fact that most major patent offices around the world have been granting patents for at least some classes of computer-implemented inventions for over two decades, the patentability of software remains controversial and continues to receive regular attention from courts, legislators, and other interested parties.

As reported here last week, the New Zealand government appears to be pressing ahead with a move to significantly limit the scope of patentable software.  This year, both the US Supreme Court (in Bilski v Kappos) and the European Patent Office Enlarged Board of Appeal (in G 0003/08) have been given cause to reconsider the issue, and in both cases elected to preserve the status quo, much to the consternation of software patent opponents.

We would like to find out a bit more about our readers' positions on this issue, and your participation by completing the brief survey below would be much appreciated.

2 comments:

Philip said...

The basic problem I have with existing software patents is that there doesn't seem to be enough novelty or uniqueness to most of them. I think that most patents don't pass the non-obviousness test. I'm not sure how that would get judged, but IMO merely applying an existing algorithm/process to solve a new problem isn't patentable.

Dr Mark Summerfield said...

Thanks, Philip, for your comment. I am sure that many would agree with you. In any given case, however, deciding whether the application of a particular algorithm to a new problem qualifies as "merely" obvious, or was in fact inventive, may not be so straightforward.

To give an example, Cooley and Tukey published their algorithm for computing a fast Fourier transform (FFT) in 1965 (although apparently it was known to Gauss in 1805). At around the same time, Chang published a paper setting out the principles of Orthogonal Frequency Division Multiplexing (OFDM) in the Bell System Technical Journal.

However, it was not until 1971 that Weinstein and Ebert published their paper on the application of the FFT to OFDM. With the benefit of hindsight this might seem an obvious combination of the two pre-existing ideas, but personally I think it is an inspired synthesis that has ultimately led to practical ADSL, Wi-Fi and broadband cellular communications technologies, among others things.

There is an element of subjectivity in the obviousness analysis that affects all technologies. At the moment it appears more problematic in relation to computer-implemented inventions because patent offices and courts are less experienced in dealing with these technologies, and the relevant prior art is harder to search.

I believe that the situation is improving, and I am hopeful that it will continue to do so.

Post a Comment