C/C++:

   printf("%lf\n", 1000000 * 8.2);
   8200000.000000

Here, printf() is doing the necessary rounding to hide the fact that 8.2 isn't exactly representable. It's interesting that other languages don't do something similar when outputting a float to text.