10 ms·
The original didn't require infinite precision; rather, it required that the average of two different numbers be between the two or equal to the smaller one. Th
by thedufer 10y ago
The original didn't require infinite precision; rather, it required that the average of two different numbers be between the two or equal to the smaller one. This is true in integer arithmetic due to rounding down, but apparently not true for floating point.