'Father of Method Patents' Slams CDDL
By webmink on Feb 20, 2005
It's a funny old world. Do you ever wonder who was the "architect of the landmark decision [on] the Federal Circuit ... [that] opened the door to patents for business methods"? If you'd been at LinuxWorld last week you could have met him - Steven Henry, the man who invented method patents and whose influence was the starting point for the whole software patent situation in which we find ourselves, gave a keynote. In that keynote he tried to fling FUD on the license most likely to de-claw his invention. Which was written by a company his audience probably wouldn't expect.
Now, I am very cautious about reported speech in the media, but there are several things in the anonymous Sys-Con reportage that make the jaw drop and the eyes widen in incredulity. Some of the discussion points he makes bear exploration, but in general I get the feeling he just doesn't grok the GPL. He thinks it's a license; the rest of us know that it's the manifesto for a social movement. But exploration of that will have to wait for another time and another place.
I wonder if Steve has been reading the Webmink blog? He mentions "the elephant in the room" and that's how the issue of software patents is referred to (which Steve invented so he clearly doesn't want us to ignore that lucrative elephant). I have to say I rather doubt it because if he'd read about toads and ants he'd possibly not make such gigantic errors in describing the CDDL (which is hardly a "User license", Sys-Con). He says:
While [the CDDL] grants a limited patent license, that license is lost once the original code is modified, leaving the licensee in possible jeopardy.
Let's be clear. No part of that statement is true. Let's take it clause by clause\*:
- "...grants a limited patent license..." No. If you read the license you'll see in paragraph 2.1 and 2.2 that actually anyone releasing software under this license grants a blanket license to all IP without the need for it to be enumerated or identified in any way. That's
"a world-wide, royalty-free, non-exclusive license... to use, reproduce, modify, display, perform, sublicense and distribute ... to make, have made, use, practice, sell, and offer for sale, and/or otherwise dispose of the Original Software"that applies to all intellectual property and patent rights, both real patents on things and imaginary patents on software and methods, whether the parties to the license realise it or not. Are those terms "limited" in anything but the most inaccessibly technical sense? Or maybe Steve is worried he'll lose business under paragraph 6.2, which could make software patent violation lawsuits a thing of the past? Now that would be a limitation worth having.
- "...that license is lost once the original code is modified..." No. Maybe Steve didn't read it for long enough. 2.1(d) and 2.2(d) do include a clause saying that a licensee can't extend the scope of the rights grant to patents that weren't covered by the original blanket patent grant, but that's not a surprise really, is it? Without those words the license would be an unlimited license to all patents, real or imaginary, as long as you could slip in a few lines of code that needed them. No, the license is solid as a rock, even when the code is modified.
- "...leaving the licensee in possible jeopardy." There's no jeopardy here, Steve. In fact, paragraph 6.3 even makes sure that end-user licenses survive the mutual destruction caused by a patent terrorist, and paragraph 9 tries its hardest to tie up loose ends caused by differing international jurisdictions. Face it, CDDL is the best attempt so far to neutralise the Gremlin you fed after midnight.
Bottom line - The patent grants received under CDDL are retained over time, including after modifications are made. In fact, patent grants and copyrights granted under CDDL are only lost if the license is terminated due to breach or assertion of patent infringement in Section 6 of CDDL. Steve was wrong, or at least misleading, in his comments. So while the CDDL may be "...the latest indicator of the need to address the patent issue in a constructive manner," the truth is that the CDDL also addresses that issue and constructively points the way to how to build patent-hardened open source communities, by including a blanket patent grant in the license so that the community evolves a meshed patent defence against attackers.
\* I should mention that I am not a lawyer, have never played one on TV or elsewhere, and am not empowered to make binding statements in this regard on behalf of anyone including Sun Microsystems. If you're engaging in litigation, you can't rely on anything I am saying. Still, I don't think you can rely on anything Steve says about CDDL here either, even if he's your attorney.