Of course, the real reason they work with pi to exactly 15 decimal places is that they're doing numerics, and therefore use a full double precision constant for

Of course, the real reason they work with pi to exactly 15 decimal places is that they're doing numerics, and therefore use a full double precision constant for pi, i.e., the number of decimal places that can be stored in a 64 bit float