It may come as a bit of a surprise when the C floating-point types (float, double), fail to capture a particular value accurately. Certainly double, which is usually stored as a 64-bit value, is far better than the old float, which is only 32 bits, but problems do arise. For example: float n