I'm guessing they'll take this opportunity to somehow find a way to extend it even further. Perhaps that Velvin fellow was onto something with regards to prior art being required to run on the same CPU to be considered valid?

If they want to make prior art architecture-dependent, I'm all for it, as long as they make infringement similarly architecture-dependent. Then we can all just switch architecture to 64-bit ARM and get rid of all existing software patents, and port as much as possible to establish a broad base of unpatentable prior art on that architecture.

I don't think you understand the who-whom here.

Requiring prior art to be strictly identical is good for patent lawyers, so that is more and more the rule. Hogan actually would have had a point, if he were interpreting CAFC rulings instead of serving on a jury.

Requiring patents to be interpreted narrowly would be good for inventors, so that is out of the question. The patent bar will see to it, higher courts be damned.

I understand the lengths the Federal Circuit will go to in supporting patents, I just think it would be difficult even for them to require prior art to be tied to a particular architecture but allow a patent to claim the use of any general purpose processor. I wouldn't be particularly surprised if they tried to go that way, but it seems like that would make it even more likely that the Supreme Court would overrule.