Categories: patent

Open Source Software and the Patent System

As we’ve highlighted in the past, we believe that the software patent system is challenging for open source software development.

Because of the short innovation cycle and continuous iteration of software development, long patent terms impede the rapidly iterative processes that sustain the pace of software innovation. The “FUD” (fear, uncertainty and doubt) caused by software patents in a rapidly innovating space can cause everyone, particularly open source projects, to be frustrated (for an example, one of many, look at Google’s and Microsoft’s open codec attempts to see this in action).

The problem has several aspects. First, overworked patent offices have lead to the issuance of many non-innovative (and therefore at least partially invalid) software patents (some commentators have suggested this may be 38% of software patents).  The number of patent applications only continues to grow, exacerbating this problem. Secondly, patent examiners who are strapped for time and resources mainly look to filings in the patent office itself as evidence of whether a patent claim is novel.  Unfortunately, this makes it hard to adequately evaluate new software patent applications and identify relevant prior art — especially when the prior art is located in open source repositories and wikis, outside of patent office filings. This landscape has allowed software patent applicants to successfully claim software functions without worrying about open source precedents that would otherwise invalidate their patents, and sometimes even without any meaningful limitation to a particular system or purpose.  Finally, once a patent is issued, it is very hard to invalidate. It can be at least two to four times the cost to invalidate a bad patent, even in the case where prior art is clear.

The threat posed by the growing pervasiveness of these types of overbroad and vague software patents is the shroud of FUD they cast over emerging and innovative technologies. It can feel impossible to know whether you are infringing someone else’s software patent, which can slow or frustrate innovation.  Aggressive patent litigation and settlement strategies have also created an atmosphere of FUD for the purpose of damaging open alternatives to proprietary products. Additionally, patent trolls have added uncertainty to open development by aggressively suing based on patent portfolios that allegedly cover foundational software technology. It is sadly ironic that much of the increasing costs of software patent issues are being borne by innovators themselves, including those in the open source community — the very individuals the patent system was supposed to incentivize.

Many of us (including the EFF, Engine Advocacy, Open Invention Network, and, to name a few) have addressed the issue in various ways. Each of these tactics have led to changes in the way we think about the issue as well as breaking away at various parts of the problem. As Mozilla, we need to do something different that leverages our position in the ecosystem as an open innovator. We have been thinking about this issue and, over the next few weeks, we’ll be working towards a Mozilla contribution to addressing the software patent problem. If you have thoughts about things we could do, we’d love to hear from you as well.