Would be patent-trolls searching through commit logs for ideas to patent sounds like a scare-story on a local news station, and the goods news is that there aren’t any cut-and-dry cases of this happening yet. However, Tech Dirt‘s Mike Masnick has found two suspicious instances of patents that may be “borrowing” technology from open source projects. In one case, IBM made the case that it had “improved” on prior art and was granted a patent. Is this something open source developers should be worried about?
The first case involves an algorithm patented by Tandberg . Jason Garrett-Glaser, an x264 developer, initially claimed that Tandberg copied an algorithm he created step by step out of his commit code. He’s since posted an update saying Tanberg claims to have come up with the algorithm on its own. Garrett-Glaser admits that it’s believable, since the algorithm is “way too obvious to be patented” in his opinion. Tandberg also claims that the algorithm it patented isn’t exactly the same as Garrett-Glaser’s. Garrett-Glaser isn’t convinced.
In the second case, IBM patented a Windows library very similar to the open source project HeapCheck. The USPTO rejected the IBM’s patent application three times because it was too similar to HeapCheck, but eventually IBM made the case that it had improved on the original project and was granted a patent.
Has this sort of thing happened before? Is this something open source developers should be worried about? How can developers stop patent trolls from stealing their work?
