I agree. One possible reason against it though is that 'double' by itself is also used as a type denoting 'double precision floating point number'