Assuming you're not advocating all intellectual property be essentially un-patentable, then the nut of the problem seems to be in how to interpret "How about this. ... but others can patent improved versions, too".
What constitutes "improved"? An algorithm that is 1% faster? How about a web search that a user survey says givee them better results? What if I say my UI, which is functionally exact copy of Ebay's, has prettier colors and so is "improved"?
How do courts and patent offices tell the different between an substantive improvement, an implementation change which is not central to the working of the software?
If you think the current patent / legal system slows progress, think of all the little guys suing the big guys for "go away" money under ambigious "improved" definitions. Think of every company that has an existing product, which would have to check every minor improvement to their software against all the extra patented variations. I'd sure bet THAT would slow down the experimental side of incremental innovation.
I'm guessing that part of your job is to be somewhat inflammatory, but I think the very important need for software patent refore will be better served if each suggested reform includes discussing a number of likely long-tem consequences of the reform.
With Congress about to head for the beach, it’s halftime for the old patent law fight.
We do this every Congress. The software industry pushes for patent reform. The drug and medical device industries push back.
Eventually nothing is done. But there’s always entertainment, and lovely party gifts (in the form of campaign contributions) for both sides.
Alone among the folks at Smartplanet I cover the two parties in this war — healthcare and computers. I admit sympathy for computers. But that won’t solve the problem.
The real problem, it seems to me, lies in how the problem is defined. Software patents are often issued for the idea of doing something. People have patented multimedia. In drugs, in medical devices, and everywhere else, patents are based on a method of doing something.
Take the bird feeder outside my window. There are patent marks on it. The perch is weight-sensitive and there’s an electric motor inside it.
If a cute little birdie perches to feed it gets the seed. If a big nasty squirrel tries to get on the motor causes the perch to spin around and the squirrel is sent into the next yard. They only try it once, but I keep watching, hoping some entrepreneurial squirrel is going to sell tickets for the thing.
My point is that, while my bird feeder is patented, there is no patent on the idea of a bird feeder. Just on the way this one works. Birds are not starving due to this guy’s patent.
If this were software, they would be. Software doesn’t just protect code, but the idea of what the code is trying to do. So software patent holders try to hold up whole swathes of technology progress, and as we saw in the RIM case, they sometimes get away with it.
So the terms of patent fights don’t need to change. It’s just the terms under which software is patented, if it is to be patented at all, that needs to change.
How about this. You can file a patent on multimedia, or video, or blogging, but others can patent improved versions, too. Blogger can have a patent, Movable Type can have a patent, Wordpress can have a patent. Everybody can have a patent. But so can everybody else.
The same might hold for genes. You can’t patent the specific genes, just your model for finding them. I don’t want some hoser suing me in the Eastern District of Texas claiming he wants cash for my baldness. But if he creates an invention so baldness becomes a choice, I have no objection to being the last bald man in America.
In the end this makes software and genetic patents worthless, which is sort of my point. Software, in the end, is just math. Genes are, too, in a sense. They’re tools. What you do with them deserves a patent.
Now if you’ll excuse me there’s a young squirrel outside getting very, very curious….
