So you think that inventors who don't practice shouldn't be compensated.
No, they should not. If you don't practice then you, and the public in general, are losing nothing from your inability to sue people who are actually making things happen. If you don't practice then you have no expectation of being compensated for your work other than hoping that someone else steps on your patent and trolling them. If your patent actually has value then you should be forced to take the risk to demonstrate that fact by putting the invention into practice -- you do not need to be successful at the practice, but you should be forced to at least make the effort.
> If you don't practice then you, and the public in general, are losing nothing from your inability to sue people who are actually making things happen.
I don't work for free, so I'm not gong to bother to disclose if I'm not compensated.
> If you don't practice then you have no expectation of being compensated for your work other than hoping that someone else steps on your patent and trolling them.
Wrong. In the current world, I can file and approach folks who might benefit, offering to license.
In your world, I can't disclose without risking them saying "great idea, thanks".
And no, no one signs NDAs with individual inventors, so that's a non-starter.
> If your patent actually has value then you should be forced to take the risk to demonstrate that fact by putting the invention into practice
The existing patent thicket around processors means that there's no way for me to practice. Plus there's the "invest millions of dollars" part (getting to tape-out is at least $1M, tape-out adds another $1M, and I still haven't done product support or any of the other things required to actually practice).
The value of my invention has nothing to do with whether I can do other things (design register files, etc) so it's unclear why my compensation should depend on doing said other things.
The motivation for you to patent is that someone else can file before you even if you invented it first. That's new in the United States and common in other jurisdictions.
You make money by licensing your inventions to companies that have means to create a product. You can use trade secrets and copyright to protect your invention. You don't need patents. However, in the defensive patent world, you should still get one in case they steal your idea.
What you'd be prevented from doing is creating an expansive definition of a market. That's not crazy talk. In the case of the Lodsys patents, Novell has filed that for prosecution history estoppel, which my reading of Groklaw might be because the original patents were filed with caveats they would only apply to a restricted market (probably set-top pay-per-view TV boxes).
I think this constraint on patents should be explicit and part of the law. i.e. a Pharmaceutical patent should declare it is for the pharmaceutical industry; if material engineering could use the same organic compound, that is sufficiently novel and non-predatory of the original patent holder. However, it's unclear in the current law whether or not the new use could pass.
> You make money by licensing your inventions to companies that have means to create a product.
That's assuming that I have something to license. Abolishing patents or requiring me to practice means that I don't.
> You can use trade secrets and copyright to protect your invention.
Neither one applies to processor features.
For example, I invented a new way to do branch prediction. Copyright would apply to any documents that I produce or to the mask that someone used to implement that feature, but a new processor, by definition, wouldn't use that same mask or even a portion thereof.
As to copyright, a processor company can implement without copying any document (and even if someone did copy, the damages are at most $250k if I can prove it, which is going to be expensive). They merely need to read whatever description I provide.
For similar reasons, trade secret is a non-starter. (You have to disclose many processor features to make them useful.)
> What you'd be prevented from doing is creating an expansive definition of a market.
Except that I'm not trying to do that. I'm trying to be compensated for inventing a new way to do branch prediction (for example).
No, they should not. If you don't practice then you, and the public in general, are losing nothing from your inability to sue people who are actually making things happen. If you don't practice then you have no expectation of being compensated for your work other than hoping that someone else steps on your patent and trolling them. If your patent actually has value then you should be forced to take the risk to demonstrate that fact by putting the invention into practice -- you do not need to be successful at the practice, but you should be forced to at least make the effort.