What, exactly, is the problem? Computer floating point
arithmetic is not exact. 0.999999999999943 is very close
to 1 (5.7E-12% deviation is negligible for most
applications).