[webmaven]

> If the patent office (and later, the courts) would just apply the non-obviousness criteria reasonably we wouldn't have these problems.

Agreed, but given that inventors can endlessly refile with whatever minor tweaks they like, and examiners are incentivised according to the number of patents they grant, it isn't hard to wear them down, eventually.

Courts then presume a patent is valid and it takes a lot of effort to convince them otherwise, and typically it is only prior art that will do it.

[nickff]

>"examiners are incentivised according to the number of patents they grant"

This is completely wrong; their incentive is actually to reject an application. The optimal strategy for an examiner seems to be to reject an application at least twice, usually more. There is abundant literature on this topic, and I encourage you to explore it.