Of course there are folks using patented technology without permission of the owner who nevertheless did not intend to infringe a patent -- that's exactly what's happening here with Mindcraft!
For the sake of argument, imagine I can prove that I came up with an idea for something on my own. I'm completely unaware that it has previously been discovered and patented. I did not even think it was an idea that could be patented. Did I intend to infringe that patent?
That's like arguing that you thought it was legal to kill someone using radioactive isotopes and therefore you had no illegal intent. Patent infringement is defined by the corpus of protected inventions and being unaware of the locus of infringing actions is not a defence against infringement.
If you make sufficient disclosure part of the patent system then this is the only way to do things really. If you don't require it then patents are worthless for the [general populous of the] state.
It's completely different. For starters, there's the "reasonable person" test. No reasonable person is going to think that murder is legal.
Intentionally infringing a patent is punished much more harshly, and it's impossible to keep up with the millions of patents that are published, so being intentionally unaware is actually a decent legal defence.
And yes, this does mean that patents are generally pretty useless for software.
Also, I'm not a lawyer, so correct me if I'm wrong, but isn't it also patent infringement if the following happens:
A patent is filed for technique X by company A,
Company B use technique X having created it independently,
The patent is issued to Company A,
Company B is now infringing upon Company A's patent.
Company B could not possibly have committed patent infringement intentionally, because the patent they infringed didn't even exist when they went about implementing their product.
The key issue of disclosure - your 3rd step - happens at A-publication which is usually quite some time prior to the patent being granted but still can be a couple of years after the filing date. (this is the source of many of the 'I can't believe they granted this patent' articles because the A-document is printed from the submission prior to substantive examination).
I don't know about the US situation but in the UK Section 64 of the Patents Act (http://www.ipo.gov.uk/pro-types/pro-patent/p-law/p-manual/p-... and http://www.legislation.gov.uk/ukpga/1977/37/section/64) allows a person to continue activity which was started prior to the grant of a patent in secret or where sufficient preparations for that [infringing] activity were made prior to the application/grant in good faith. If the activity wasn't in secret then it forms part of the prior art and invalidates, or forces amendment of, the patent in question.
If there isn't an equivalent US clause I'd be surprised: 35 U.S.C. 273 "Defense to infringement based on earlier inventor." looks like a candidate.
IANAL, but I believe that's only true if B was using it before A filed. The order described was "A files, B re-invents, A's patent is issued". It may have some weight in arguing that the invention was obvious, but my understanding is that this is harder than invalidating a patent based on prior art.
For the sake of argument, imagine I can prove that I came up with an idea for something on my own. I'm completely unaware that it has previously been discovered and patented. I did not even think it was an idea that could be patented. Did I intend to infringe that patent?