which of three sizes of floating point types should be used when extended precision is required
By Jimmie Wolff at May 10 2020