USPTO 708502000: A method and apparatus to perform efficient software-based integer division. The equivalent of a hardware-based integer division operation is enabled via a reciprocal multiplication operation that is facilitated by a minimum combination of multiplication (and/or add) and shift operations.
So, doing integer division by combining multiplication then bit shifting is about to be patented.
Is it... Obvious? Bleedingly.
Is it... Novel? About as novel as 4th grade math.
Is there... Prior Art? The first people to invent binary computer probably understood and used this trick, so that takes us back to Claude Shannon in, oh say 1937. And this trick was surely in use back in the 1970s as well. But if you're talking about documented uses, someone has already found a published article from 1991 that explained this technique. So that would be provable prior art from at least 15 years ago.
Being able to patent algorithms is wrong-headed, period. And Business Method patents (currently allowed) fall into the same category. If two people independently discover a series of steps to solve a problem, there is no reason the one who discovered it second should have to pay the one who discovered it first. You are not entitled to money just because you happened to be lucky and be the first to stumble across something. "Discoveries" aren't supposed to be patentable anyway - only actual inventions. Patents should only be granted on devices and aparatus. Never on ideas, algorithms, or business methods.
Patents are broken. Instead of encouraging innovation, they're preventing it. The only people getting rich off patents are patent lawyers. Nothing against patent lawyers - some of my friends, etc. But I abhor the necessity, especially when it's a transparently bad system that's hobbling the free market and being paid for out of everyone's tax dollars.