I added the ability to set a flag on the numpy format handler to raise errors when a numpy array is passed in that requires a copy to be passed to ctypes. The idea is that you can use it to catch unwanted copying (during development). Found quite a few instances in OpenGLContext where I was needlessly copying data, in all the cases I found it was because I was using a 'd' instead of an 'f' array.

I see by the diffs from yesterday that Josh has been doing quite a bit of work on the darwin port. I'm thinking I'll take a 30 minute break and then boot into Windows to take another crack at that port. (I'm not up to doing C compiler wrestling this evening, but ctypes hacking is considerably easier).