Invention is defined as "A new device, method, or process developed from study and experimentation".
I'm going to suggest that there are only two rational schools of thought for patents: Either everything should be patentable, or nothing should be patentable. Everyone else either hasn't thought it through, or is dillusional or is being hypocritical.
Historically, patents were awarded for mechanical innovation and invention: a better steam engine for example. I've never talked to anyone about patents who is pro-patent but doesn't think mechanical devices should be patentable. So, Mr. James Watt, an engineer by training, worked to improve a mechanical device and received a patent, and most people think this is quite reasonable.
But what did Watt actually do? He didn't invent or discover either steam or metal. He wasn't the first metal worker, discovering how to shape and work metal. In fact he didn't even come up with the idea of a "steam engine".
What he did do was come up with a way for the steam engine to be more efficient. What he did do was improve on an existing design. What he did do was invent a mechanical process.
So, if we fast forward to today, and a businessman, learned in the ways of business, suggests that he has a novel business process. He doesn't invent the computer, or the internet. He isn't even the first to sell stuff on the internet. What he does do is improve on an existing process. What he does do is come up with a way for online shopping to be more efficient. What he does do is invent a business process: Customers can buy something with a single click. So, how is this less worthy of being protected by patent law?
In another city, a self-taught programmer is working on something new. He doesn't invent or discover colors or mathmatics. He isn't the first computer graphic artist, discovering how to draw shapes on a computer. In fact he doesn't even come up with the idea of a "compressed digital images".
What he does do is come up with a way for the compression to be more efficient. What he does do is improve on an existing design. What he does is to invent a computer process: a way to compress digital images. Again, what has Watt done that the clever programmer didn't do?
In all these cases, someone has designed and refined a process for accomplishing some task: a manner of accomplishing something. There is no fundamental difference between "a way to move a piston using steam" and "a way to serve the customer using a single click" or "a way to compress digital images".
To grant the mechanical engineer protection from competition, while throwing the businessman or the programmer to the wolves is either hypocritical or condescending.
Does the engineer need protection to cause innovation while the programmer and businessman don't?
So, why does this camp exist? Those that insist that patents are good for mechanical devices, but software and business process patents are bad?
Either patents encourage and foster innovation or they do not. If they do, then they should equally encourage the programmer and businessman to make new and better programs and business methods as they do the engineer to make a better engine.
By extension, if software patents are bad for software development, then it stands to reason that all patents are bad. And for the same reasons.
For those that are new here, I sit firmly in the camp that all patents are bad, and the only acceptable "fix" is to begin to dismantle the patent system.
One can only hope that one day history will look back on patents as just another failed experiment, along with communism: Looks great on paper, but it's counter-productive and doesn't work in implementation.