Yesterday, Red Hat had something to say to the United States Patent and Trademark Office (USPTO) about granting software patents. In all fairness, the USPTO asked for public input on how it should use the Supreme Court's Bilski decision to guide it when granting new patents. Red Hat has a long history of answering the call for public comment on this cause.
Red Hat's legal team essentially said the patent office now has the excuse to become uber strict when issuing new software patents. Although the July decision didn't abolish software patents (as some hoped it would), it could still be used to Just Say No to most of them. That's what Rob Tiller figures. He's Red Hat's vice president and assistant general counsel, IP.
"There was a reaction of disappointment and frustration from the free software community [over the Bilski decision]," Tiller told me on the phone today. "But now with the chance to think about it more, certainly the question was not answered in a way that is terrible. There is still the possibility that the patent office apply the decision in a way that is productive and will improve situation."
Improving the situation -- if not outright fixing it -- would be a good thing for the USPTO to do. To paraphrase Red Hat's position paper, there are hundreds of thousands of software patents, with tens of thousand more granted each year. Many are so vague that it's impossible to ensure that a new piece of code doesn't infringe on one of them, somehow. This in turn places a big fat bullseye on the back of all software developers, as infringement lawsuits cost millions of dollars to defend, let alone actual damages or injunctions. Patents are meant to encourage innovation, but software patents have become weapons that discourage it.
Free software advocates, and others, had hoped that the Bilski decision would mean an end to "business process" patents, which includes all software patents. The ruling fell far short of that. They had wanted the court to affirm a lower-court ruling that the "machine or transformation" test was the sole test of patentability. This [generally] means a process must use a particular machine specifically invented for an unconventional process [an/or] one that transforms an article from one thing or state to another.[Revised to address comment.] Most software wouldn't pass that test.
Instead, the judges said that the machine transformation was an "important" test of patentability, but not the only one. The court offered a wishy-washy decision that offered little clarity as what was, and was not, a patentable "process."
And in there is a small but significant silver lining. If the USPTO merely includes the machine test, and makes it an important factor, most software patent applications won't hold. Here's the advice that Red Hat is offering to the USPTO:
"Red Hat submits that the Interim Guidance should be revised to recognize that software patents will ordinarily fail to satisfy the requirements of 35 U.S.C. Section 101 as interpreted in Bilski and prior Supreme Court cases. Software is essentially nothing more than a set of mathematical algorithms expressed in a particular programming or machine language. As the Bilski Court recognized, mathematical algorithms, by themselves, are abstract ideas that are not patentable."
Congress still needs to pass patent reform. But there's reason to hope that the USPTO will use Bilski to stop the crazy business-as-usual. As Tiller says, "In the Supreme Court ruling, the Court said the applied machine transformation test wasn't the sole test, but it was an important test. That's a much better situation than we were at before."
Like this? Here's more:
- All of today's open source news and blogs
- Linux servers, Windows XP desktops: a happy marriage?
- Linux guru offers advice on MeeGo tablets, Linux security and virtualization
- Does the Linux desktop matter?
- Burning Man is the proving point for Earth friendly, open source Linux-based cell tower
Follow all Open Source Subnet blog posts on Twitter @OSSubnet