The search continued this week for more information on software patents and just how, if at all, open source impacts the process. In addition to the two conversations about which I've already written (Phil Marcus and Bruce Abramson), I also spoke to three intellectual property attorneys in private practice and a representative from the Software Freedom Law Center (SFLC). Not surprisingly, all had interesting things to add to the discussion.
To get the free software/open source software perspective, I called Bradley Kuhn, who serves as the SFLC's community liaison. Many times when I've covered the SFLC's activities, they've been rankled at something Microsoft or another proprietary software company was trying to do. So I was surprised when, for background on the organization's position regarding software patents, I was pointed to an amicus brief the SFLC had filed with the U.S. Supreme Court in 2006 -- on Microsoft's behalf.
In its first segment, the brief states:
SFLC has an interest in ensuring that limits are maintained on the reach of patent law...so that Free and Open Source software development is not unreasonably and unnecessarily impeded.
The writers then go on to argue that software is not patentable subject matter because "it does nothing more than execute mathematical algorithms." And the Supreme Court has held that algorithms, abstract concepts and the like, on their own, are not patentable. (There is more to the argument, of course, but that's the main point.)
We all know, however, that software patents do exist and continue to be granted. Kuhn explained:
What happened is that people began to patent what were called "business methods," and they were easily able to draft patent applications such that software algorithms could be written as business methods.Therefore they effectively snuck software patents in.
And software patents, in the SFLC's view, impede the advance of technology. He illustrated the point this way:
There was a software patent years ago on an algorithm called RSA, which is an encryption algorithm that today is used in everybody's Web browser. But that patent was generated at MIT, and MIT spun off a company that charged exorbitant fees to license the patent. I firmly believe that e-commerce on the Internet could have happened a lot more quickly had that patent not been in the way. Because it was very expensive to license and it was very difficult to do encrypted, secure transactions on the Internet without that algorithm.
Not until that patent expired in 1999, he said, did e-commerce really take off.
He also noted that the patent system is very hard on the individual software developer, small startup or any company that doesn't have a lot of resources because they have no way of knowing whether something they have written -- even completely independently, from scratch -- infringes someone else's patent until they're sued for patent infringement.