If I invest the time to work out a new sorting algorithm, through much subtlty, etc., then this is a method that is not attached to specific code, and couldn't be protected under copyright.
An interesting point. Mind you, what might take a certain person a year to figure out, might take only weeks or minutes for someone else.
I'll give you a personal example. Back in the 70s I was writing graphics code in machine language. I needed a quick way to draw a line. I spent many weeks figuring out how, and eventually "discovered" a way. Then I bought a new graphics book and found out that all I had done, was "rediscover" the many years old Bresenham Algorithm.
That's one problem with software patents. As you said, I put a lot of work into creating a very subtle algorithm. But others could do the exact same thing with either much more or much less effort. Code solutions tend to converge towards a similarity (just made that up in place of singularity).
Btw, thank goodness software patents were basically unknown back then, or we'd have had almost no personal computers. Everyone depended on using everyone else's ideas at the time, especially Apple.
-- Patenting Ideas
Another problem with many software patents, most especially the ones that Apple likes to go for, is that they often do not detail any subtle process. They're written in very general terms, and basically patent an IDEA, which is normally unpatentable.
For example, their patent on searching multiple locations. Even a casual user can come up with that idea. "Hey, why can't this search box look at both my machine and the web?" Etc.
Now, Apple's patent is on using multiple software modules, one for each type of sub-search, to which the search is passed and results are passed back. Well, duh. It's just normal object oriented programming.
So their patent is a combination of the obvious and the well known. If you put ten developers in a room and gave them the problem, all or most of them would come up with the same idea.
Ditto for their patent on recognizing phone numbers etc in text. Although wrapped up in fancy language, it's nothing special or detailed. It's just the description of the idea.
-- Patenting Gestures
Apple continues to try to patent the obvious every day, because no one else was greedy enough to. One of the worst was their attempt to patent the idea of using two fingers to "rotate" a virtual knob on a multi-touch screen... an obvious gesture. Fortunately, the USPTO denied it as they should. But the sheer fact that Apple tried is worrisome, because it shows the depths to which they'll go, plus we've seen the USPTO grant some other similar and just as ridiculous gesture patent applications like using two fingers to scroll in a window.
As you noted, perhaps one solution is to stop allowing lay juries to decide patents. (The recent Patent Reform Act unfortunately still allows that, but it also adds the option of direct challenges via the USPTO.)
Interesting discussion. Regards.