« More Finger-Pointing in the E-Book Price Wars |
| Because There's a Copyright Angle to Everything »
September 14, 2012
Can Looking Backward Help Fix the Software Patent Mess?
Here's a big thought for you: maybe we already have the laws we need to untangle (some of) the current mess around software patents. This is the thrust of an upcoming paper by Professor Mark Lemley (Stanford) which is previewed in today's InfoWorld column by Simon Phipps
To begin with, you should know that Phipps is himself a strong proponent of open source development (his column is called "Open Sources", for example) and a recent opponent of software patents in part because they interfere with open-source development. So when he says there might be something good in the legislation we already have around software patenting my eyebrows tend to go up.
According to Phipps, the working paper titled "Software Patents and the Return of Functional Claiming" from Lemley (SSRN link here) will propose that the problem with software patents is what lawyers tend to call "claim construction." That is, the problem is not with what's being patented (software) but with the things to which the patent is supposed to grant a monopoly. Badly constructed (overbroad) claims attempt to protect not the (software) machine, nor the steps (algorithm) it performs but the result of executing those steps on that machine - the goal.
This is immediately interesting to me, as it neatly sidesteps my number-one objection to abolishing software patents - the Church-Turing hypothesis issue. In fact, what Lemley appears to be saying is exactly what I'd expect Church-Turing to say: you can patent a machine but in doing so you're not protecting that machine's products.
Lemley's argument (and I haven't read the paper yet, so take this for what it's worth) is that the USPTO is not correctly applying the already existing Patent Act of 1952 and if in fact the Office would do that, then it would reject these patents with overbroad claims, or at least force patent applicants to redraw their claims more narrowly. That wouldn't immediately solve everything, not least of which are the ongoing problems with disgracefully shoddy lack of prior art inclusion in software patents and the pervasive confusion over sections 101 vs 103 of the law. But it would be a huge step forward.
One final caveat to keep in mind here: Lemley's thesis is in a "working paper". That means it hasn't yet been subjected to the kinds of critical review that more formal publication might require. I suspect we'll see people who believe that Lemley is misinterpreting the 1952 Act. Still, it's a blessing to have fresh ideas for improving the situation, particularly ideas that don't require radical reforms or rafts of new legislation.
+ TrackBacks (0) | Category: Big Thoughts
POST A COMMENT
- RELATED ENTRIES
- A Smaller Piece of a Much Bigger Market
- Jim Hines' Writing Income
- Cory Doctorow Would Like You to Know About Downpour
- Too Many Words, Not Enough Readers
- Yeah, No Thanks Toast & Butter
- Apple I Reaches CAFC
- Macmillan Pretends It Can Plug Analog Hole
- Pomplamoose is Still Making It