CGFloat is 32-bit on 32-bit architectures and 64-bit on 64-bit architectures for historical reasons. Rather than having it alias either Float (32-bit) or Double (64-bit), introduce a distinct struc...