07 August 2011

Patent Absurdity Becomes Absurdly Patent

Something wonderful has happened over the last few weeks: more people have woken up to the threat that patents represent to innovation.

I'd like to think that it was my call to abolish patents completely that started this, but it's more likely to have been the NPR feature that got people aware of this.

As well as NPR's own follow up, Forbes joined in with a call for software patents to be invalidated, and we even saw The Economist belatedly waking up to the reality of this intellectual monopoly.

And still they're coming. Here's Mark Cuban putting his oar in [update: and here's his solution - abolish software and process patents], while Dave Winer concluded a piece with the memorable line:

These guys [referring to Nathan Myhvold et al.] are so perfectly evil and on such a huge scale, it's as if they were out of a DC Comic.

Finally, we even had Dilbert on the subject.

So, now that everyone with a brain agrees there's a problem with patents, how about really trying to solve it?

Follow me @glynmoody on Twitter and identi.ca, or on Google+


Arjen Kamphuis said...

How do we change things? The same way we have always changed them: massive civil disobedience. Laws that are patently unjust should be ignored and when enough people do that they will go away. Just like slavery or non-universal suffrage.

Glyn Moody said...

@Arjen: so how do you think that might work in this area?

ipmaster said...

Time to adopt "Use it or lose it". Patent trolling only works if sufficient patents are (a) acquired (but cannot be used as they are patent trolls, not manufacturers) and (b) warehoused until the rest of the world overtakes the patent. If a patent is not commercialised within 18 months, it should become defunct unless the patent owner can demonstrate actual use or licence activity.

Glyn Moody said...

@ipmaster: that would certainly be an eminently sensible first step...maybe even achievable...

Engine Room said...

The current state of patents has two causes:

1) patents on "procedures" incl software
2) how patent offices are set up

Re 2)
Patent offices are expected to operate at "no cost". Thus, the more patents they grant with little or no examination, the more money they make on application fees.

If a patent is disputed, the office profits again, more fees.

In essence, the burden of proof has been shifted from the applicant (formerly checked by patent offices) to the people, who now have to proof that a patent isn't valid.


Glyn Moody said...

@ER: good points, especially about the crazy incentives of for the patent offices

MarkW said...

Use it or lose it has a nice ring to it, but it doesn't really fit with patents since patents don't let you do anything other than stop other people from practicing the claims of a patent. So the only use to be made of a patent is enforcing it against others to stop them from practicing the claims, and that is the very behavior we want to stop. As Jim Bessen and Mike Meurer have pointed out in their book, Patent Failure, much of the problem lies in the shear (and overreaching) ambiguity of patents within the software field. Greater clarity in patent applications would be one step, i.e., requiring use of common software terminology in drafting claims instead of trying to hide the claimed invention in ambiguous, meaningless terms that are interpreted years later to mean something the inventors never contemplated.

Anonymous said...

The first step should be a ruling from the courts that the execution of software programs does not, and can not, constitute infringement of a granted patent; no special allowance should be afforded just because the patent itself is on a "software" method.

If somebody patents a design for a new internal combustion then software which simulates that engine's operation is not considered infringing; why then should software which "simulates" the operation of other software?

Software has been deemed "speech" by the U.S. courts (Bernstein v U.S.) and is referred to as "literary work" by the Berne Convention. The purported justification of patent monopolies is that they encourage the publication and dissemination of the principles of the discovered technology and it is therefore inappropriate to suppress the sharing of that information -- by adjudging it to be infringement -- merely because the information is expressed as "software".

Glyn Moody said...

@MarkW: yes, the Patent Failure book is a must-read in this context.

Glyn Moody said...

@Saul: really what you're saying is that software patents just don't make sense...

nextoptions said...

I can't agree with abolishing patents. If i commit my time and effort to making something useful and valuable for society I expect to be compensated and i expect some protection under the law.

Glyn Moody said...

@nextoptions: well, if you are making something useful for society that is one thing, and you should be thanked for doing so - just as we all thank Berners-Lee for inventing the Web, and then giving it away.

But that's quite separate from being rewarded financially.

If you are doing it to make money - which is fine - then *that* is its own reward. Why should society reward you with a monopoly when the market will reward you with money?

Now, you may say it is not worth your while to invent without that monopoly: and you are perfectly justified in doing so. But there are now so many people inventing stuff that this really isn't a problem for society - someone else will invent it and offer it if you don't.

Anonymous said...

@glynn moody, Yes, but then I'm also proposing a (relatively) simple solution to the mess in which we, at least the U.S., find ourselves.

The Supreme Court of the United States has declared software not subject to patentability on three separate occasions. Despite this, lower courts such as the 9th Circuit and the Federal Circuit Court of Appeals have disregarded this precedent, as have the examiners of the Office of Patents and Trademarks.

If the Supreme Court does not wish their authority to be undermined then they need to step up and recapitulate their previous rulings. There is no need for the legislative branch to write new laws; at least not until the existing laws have been properly enforced and found lacking. As long as existing law is being ignored by patent examiners and the lower courts, patent reform legislation is a premature measure.

The proper manner to address this is to assert that software does not -- indeed can not -- infringe upon any extant patents. Let the USPTO examiners grant all the patents they want on abstract software algorithms. Let foolish applicants pay $10000 a pop pursuing such patents. However, the Supreme Court should not make itself beholding to the whims of bobble-headed clerks in the patent office and they don't have to get approval from Congress to prosecute the laws that Congress has already passed.

Glyn Moody said...

@saul: but isn't that equivalent to abolishing them (fine by me, of course...)?