I'd rather transition to a system where if two people try to file for the same thing within a few months or a year of each other, neither one gets it because it's probably pretty obvious.
Like naughty children, either you share nicely or nobody's having any cake!
:)
Independent invention should be a sign that the idea isn't worth patenting, that it was "in the air" at the time.
They don't, unless people use it and the patent holder wants to charge for licensing.
I don't have facts at hand, but I imagine that like any other thing (starting a business, writing a book, creating an App) 99.9% of patents result in no monetary gain whatsoever. But if we're going to go all in on the property aspects of patents, then I think that morally and ethically, we have to give an independent inventor a share of the property. Anything else just looks arbitrary, and leads to disrespect of patents, "intellectual property" and The Law in general.
It still must be an enabling disclosure, though. That is, it has to teach the public how to make and use the invention.
Because I think a whole load of really obvious rubbish gets patented at the moment, but I do think the foundations of the patent system and the utility of patents ought to be salvaged.
Applications are only published to the public either 18 months after the earliest priority date or after the patent has been granted.
The only other time you might know what they were up to is if they made a public disclosure about the technology. But as that has implications for their own filing strategy, they're not likely to do that in a way that left you an opportunity.
--edit-- this would be because patents (in my mind) are the exchange of protection/exclusivity from the state in return for disclosure to the commons of an idea that might never otherwise become known or publicised.
If loads of folks are coming up with the same idea and racing to patent it, that points (to me) to a situation that patents aren't meant to address - giving exclusivity to someone based simply on them being first of many. If many can come up with it, then the deal is lopsided and the exclusivity shouldn't be granted. IMHO.
More broadly, I think the problems we have with the patent system can be seen as being rooted in the fact that there isn't anyone fully empowered to represent the public interest in this negotiation. The PTO is supposed to do this, in theory, but the examiners are overworked and have the wrong incentive structure.
And, there's a key problem: in a dispute over the obviousness of a claimed invention, the burden of proof is on the examiner to show that it's obvious -- which can be very difficult, particularly when the applicant can reply that anything seems obvious in hindsight. I propose that instead, the burden of proof should be on the applicant to provide objective evidence of nonobviousness.
Such evidence could take various forms: showing that others have tried and failed to solve the same problem; publication in a peer-reviewed journal or conference proceedings; showing that the need for a solution has been long and keenly felt.
Slightly different subject, but I think this the best argument for why software patents, at least in their current form, are bad.
With a useful patent system, the patent library should be a treasure trove of information. People with difficult problems to solve should be wandering through the archives to see what's out there. If the patent is still in force, the license fees should be well worth it, since it solves a hard problem!
Have you even heard of a programmer doing such a thing? I certainly haven't. In fact, the general advice I hear is the exact opposite: actively avoid looking at patents, because you're liable for more damages if you infringe willfully than if you do so by accident.