Patent law, or any other law protecting someone's inventions from being stolen, would seem like a good idea, but even if it is, how can it be implemented? The idea is to stop someone from taking an idea, a method, a newly invented or discovered technique, and stop them from using it without compensating the originator of the invention or the researcher who came upon it, who no doubt spent considerable effort to find or create this thing or method. This cause might be noble, there is nothing inherently true in the idea that one should be compensated for a discovery, however it came about, only a deep historic sense of right to compensation that may well be founded on nothing.
I don't call out the sense of right to compensation, I have no strong opinions on whether it is right or wrong to claim you deserve something for your effort, but I do have a strong opinion on who, if anyone should be compensated. I believe that anyone who invents something, comes across something, discovers a new thing (new to them) should be allowed to use that thing and be one of the people to which compensation must be paid, if at all. This belief drives a stake through the heart of the patent system as is stands, as the Patent system does not reward the people who invent. The patent system does not reward everyone who discovers new things. The patent system rewards only the first person to claim they invented it, per territory.
This gulf between what seems right and what is, is insurmountable, irreparable, and points to a basic flaw in the expectation of compensation for invention and discovery.
I find that thinking about things in terms of infinity often helps to bring a more stable conclusion. If the world were infinitely large, and there was only one patent office, what would happen? Apart from the patent office being immediately overrun, the chance that your invention can be patented drops to zero. In an infinite world, someone else has always invented it before you.
In fact, this is pretty much the situation we are now in. So many trivial patents are applied for by so many, that if you go out of your way to invent something new, you're not going to be first. Not even because you didn't file for patent first, but because someone will invalidate your claim through a selection of other trivial patents claiming that they make your patent a trivial extension of the suite of common sense they are presenting. It matters not that you may have been the irate to make the link, only that you didn't have enough money to defend your case.
The other failing is when you really truly invent something new, but it is overridden by someone else who also thought of the new technique, and this happens all the time. It even happened with calculus. In this case, the loudest will win.
If the aim of patent law is to inspire people to invent, then I'm not sure they could have chosen a much worse way of doing it.
At the base of this argument is a reference to the reason why the current system cannot work. Why should a first to invent be given any advantage? Why should a happy accident be rewarded? If anyone could have come up with the inventions, as can be seen in the car of multiple inventors, then why should the invention stand for anything? If the invention is trivial, then how can you claim it cost you enough effort to deserve s reward?
If you can't answer these questions with any certainty, then patent law is a house built on sand.