Quote:
Originally Posted by tubemonkey
The patent system in this country is an absolute joke. We need a major overhaul of the entire system.
|
<Semi-off-topic rant>
We well may need an overhaul of the entire system. It's interesting, though, to consider the history of the patent system and how it has dealt with new fields in the past.
(Short summary for those in the tl;dr camp: Modern patent systems have
always sucked at dealing with new technical fields.
Every new technical domain has suffered from 20-50 years of atrociously awful patents, after which things settle down to a more reasonable state.)
Some examples from history:
Steam engines: The British patent office granted one of the early steam engine designers (Newcomen? Watt? I forget which...)
a patent on the crank, completely ignoring a few thousand years of prior art. Rather than being stymied by this, his main rival (the other one of the two above) invented the Sun-and-planet gear to work around the patent. This is a fine example of an egregiously horrible patent that should never have been granted in the first place.
Aircraft: The US patent office (and those of a number of other countries) granted a patent on "the wing" to the Wright brothers. The Wrights really did invent a crap-load of important new stuff: wing-warping (now implemented as ailerons), the wind tunnel, and tons more. And they deserved their patents on those things. But "the wing" had plenty of prior art, from early gliders back to da Vinci and even earlier. The patent office(s) gave them an absurdly broad patent -- another patent that should never have been granted.
Computer Displays: The "cursor XOR patent" covered using the exclusive-or operation to mix in a cursor on a display. It was innovative and useful, but clearly failed the obviousness test. See, there're only 32 logical operators total. Choosing from a total space of 32 choices is hardly innovative.
There are similar examples from early automobiles, steamships, railroads, radio, electronics, micro-electronics, chemistry, etc. As far as I can see, what happens is that the various patent offices are initially unable to hire examiners who have adequate knowledge of a new subject domain. So they grant bunches of bad patents. After a few decades, when the field is more mature, the early stupid patents have all expired and the patent office can afford to hire examiners who are well educated in the field. So the percentage of really stupid patents goes way down.
In the software and computer world I hear lots of folks argue that "we're different" and that things are changing so fast that these dumb patents are holding everything back. My usual response is to point out that
this is a phase that every new field goes through. It sucks, but it ends. After that, the patent system appears to work quite well.
It sometimes works in the early stages, too. Here are two examples:
Adobe attempted to patent the idea of a "tool palette" where you select a tool and then apply it to your drawing. In the early 1980s.
I sent a quick note to the named inventors, with a copy to their legal department and a couple of Adobe's founders pointing out that they had no excuse for being "unaware" that this technology was publicly demonstrated in 1964 in my Dad's Ph.D. thesis. After all, one of the founders in question had been one of my Dad's grad students and the other was an early employee at my Dad's company; both of them were
certainly aware of the prior art. Adobe abandoned that patent application.
One bad patent squashed!
I spent 14 years doing R&D at a start-up company that built optimizing compilers. Nearly everything I did for the first 10 years was considered a trade secret, and went unpublished. But then software patents came along. Thereafter, my discussions with management about publishing went like this:
Sr. VP: You cant publish that! It's too important to our products!
Me: Are we going to file for a patent on it?
SVP: No, too expensive.
Me: Then we'd better publish quickly, before someone else patents it out from under us and prevents us from using it in our products.
SVP: Right -- publish fast.
Just the
possibility of patenting a software innovation made the difference between "hide it from everyone" and "publish as fast as you can!"
</semi-off-topic rant>
Xenophon