It is mentioned here that there are two ways to declare floating point types, but only one is mentioned here, what about the other one?