Compute the number of representable double-precision floating-point values that separate two double-precision floating-point numbers along the real number line.