Clinger, William D.
(University of Oregon, 1990-06-05)
Consider the problem of converting decimal scientific notation
for a number into the best binary floating point approximation
to that number, for some fixed precision. This
problem cannot be solved using arithmetic of ...