I find myself back again reading the documentation to OpenGL. There are so many cool things that can be done with this stuff. At this point it is a bit like a whole new API though as the programmable pipeline is new since last time I took a serious look. Much more capability, but in many ways a completely different way of making things happen.

The issues around gimbal lock with Euler angles and transformation matrices (which is the approach I’ve used in the past) are making me seriously look at quaternions again. They’re a bit of a strange hybrid, but it seems as if they offer a means to manage rotations that may have fewer sharp edges than what I’ve used in the past.

I think I may wind up buying the paper copy of the the OpenGL Superbible as it is likely to be more pleasant to flip around in than the kindle version I’ve been reading. Also suspecting that HLSL is important enough that the OpenGL Shading Language reference may be something I want on my bookshelf. Add in the OpenCL Programming Guide that I picked up when I was job searching (thought there might be some CT related opportunities and for projective reconstruction, OpenCL seems very interesting) and all sorts of stuff should be possible.