Hacker News new | ask | show | jobs
by ejames 5329 days ago
It is common practice for engineering to be explicitly forbidden from looking at any patent whatsoever.

A peculiarity of the U.S. patent system is that the size of the penalty is larger if the party accused of infringement is shown to have "knowingly" infringed. This is intended to punish deliberate malign behavior more severely than accidental infringement, because there is a lower penalty if you didn't realize there was a patent to violate.

However, the second order effect is that accurate knowledge of patents is a legal liability. When considering whether a party "knows" about a patent, any engineer in the company would count. If Bob reads Patent A and Sally writes code that infringes Patent A, it's difficult for their company to prove that Bob never said anything to Sally. And why even take on the burden of that proof? As a practical matter, the potential legal cost of the larger infringement penalty outweighs the potential benefit of searching beforehand to ensure you aren't infringing.

When I joined Microsoft, part of the instructions given to all new employees was that you should never view any patent, anywhere, ever, no matter what. Even if there is some reason why knowledge of patents is necessary, you should not try to search for or view those patents yourself, because if you look at a page of 50 patent search results to find the one you want, someone could argue you "know" about the other 49. This is arguably ridiculous, but when your goal is reducing legal costs, you must avoid even the possibility of ridiculous arguments, because arguing back requires time from lawyers.

Standard operating procedure is for engineers to know nothing about patents and the legal department to make all patent-related decisions. Naturally, the legal department makes those decisions with the goal of maximizing profit from lawsuits against other companies and minimizing loss from lawsuits by other companies. The process deliberately and explicitly eliminates the possibility that accurate knowledge of software and hardware will be applied to patent decisions, because the only people with that knowledge are engineers, and letting engineers know about patents increases your legal risk.

Since the Patent Office also does not have the manpower or technical resources to make accurate decisions about engineering patents, and in any event is not incentivized to reject patent claims, patents on average bear no relation to reality. They're determined entirely by what you can get away with in court.

4 comments

This is hilarious, as the whole original point of patents is to make knowledge more widely spread so that people can use that knowledge after the expiry of the patent. It seems that a point of law meant to help the little guy is instead effectively deleting the point of having patents in the first place.
I wish everyone who defends the patent status quo could read and understand this. The ideals they espouse have no resemblance to the reality you describe. Thanks for sharing your insight.
It's kind of interesting to look at it this way, and it also presents something of a way out: if all of these companies followed the same procedures (patent-clueless engineers), and all of these engineers STILL came up with the same method, it stands to reason that the method is obvious.
Can't prove a negative, though, so legally it's a nonstarter.
BTW, the America Invent Act prohibits treble damages for failing to seek opinions of counsel on patents: http://techwildcatters.com/2011/09/a-little-info-on-the-amer...