X
Tech

Bruce Perens lashes out at absurdity and injustice in software patents process

Among my take-aways is that Perens is quite correct that the wheels of justice grind slowly (if at all) in many instances where open source stands to gain and large commercial vendors do not.
Written by Dana Gardner, Contributor
It started out as a "state of the union" address for open source, but what I really got today was a legalistic lesson from Bruce Perens, open source maven, consultant, and author of the Prentice Hall Open Source Series that takes his name. It's telling that legal minutiae dominated a discussion of open source at LinuxWorld Boston.

Perens, who has 21 books published in his series under an open source-like license (a PHP book was freely downloaded 830,000 times over the past year), says that the U.S. patent system is patently broken, especially for software. Check out the notes from his presentation.

Among my take-aways is that Perens is quite correct that the wheels of justice grind slowly (if at all) in many instances where open source stands to gain and large commercial vendors do not. And that open source coders seem to have few friends when lawsuits, frivolous or otherwise, come their way. I'm shocked, shocked that big money and the power of public companies with their lobbyists have a role in how bureaucracies prioritize their regulatory and legislative reform zeal.

But that doesn't make it right. It does mean that the messy circumstances of how the legal machine bumps into the software machine can be highly unproductive to all but a few.

Such a hit to productivity might a Microsoft-led patent infringement assault to stem or halt further open source development/deployment, for example. Such a threat, real or perceived, would have a deleterious effect on many enterprises, ISVs, SaaS providers, communities, and vendors who rely on open source efforts to augment their TCO, development efforts, market reach, and innovation. Imagine the SCO suit (Perens called SCO a proxy of Microsoft) and RIM/NTP affair on a larger order of magnitude. The RIM settlement shows, Perens says, that software patent intrigue can work on its own to disrupt without court findings: "RIM paid to end the FUD," not to give up on seeking legal vindication, said Perens.

Perens says the U.S. needs legislation to fix the software patents litigation merry-go-round. Meanwhile, he calls for the enforcement of perjury protection in the applications for patents. The last time a patent perjury case was tried: 1974. He said recent claims for patents, including those from Microsoft, amount to knowingly making claims with no rationale ... in other words, to make knowingly false claims, and therefore commit criminal perjury.

And what of enterprises who buy and use open source products and services? Perhaps they should band together and fund legal defenses against frivolous patent cases against open source projects? They should. What's more, some fairly powerful vendors are working quietly on behalf of enterprises to offer open source legal meltdown insurance, of sorts. IBM and four others (Red Hat, Novell, Sony, and Phillips) last fall created the Open Invention Network to offer a broad backstop of support against lawsuits directed at Linux.

But what of other open source projects and components? "Open will win in the long term," IBM's Scott Handy, vice president of worldwide Linux and open source, told me Wednesday. "We bet our business on open, and we don't want anyone to own 'open'."

indeed, IBM has 10 initiatives under way, says Handy, to protect against open source legal assaults from those who would game the patents and intellectual property deployment processes.

More needs to be done, says Perens. "There is no sense of peril for those that try and game the system for software patents," he says. The patent examiners' jobs are impossible. Lawyers trolling for suits take advantage of that. "We need a real measure of triviality in the presentation of patent claims. Being obvious should count in triviality," said Perens.

That seems obvious to me.
Editorial standards