What is "float_t" for?

Should I be using floa开发者_运维问答t_t instead of float?

They're additional types requires by the C standard. They're intended to be "the implementation's most efficient types at least as wide as float and double, respectively." The question of what you should be doing is a matter of opinion, I suppose, but I've never seen them used in any Cocoa code (and probably not even any C code) if that means anything to you.

No, you should be using CGFloat.