The VAX floating-point representation is different from that
on HP 9000 computers. You will have to change any programs dependent
on the characteristics of VAX floating point. In particular, this
difference could expose errors in the code that happen to work acceptably
on the VAX. These errors include mismatched function return types
(float on one
side, double
on the other), and passing the address of a double
instead of a float
to scanf. The
VAX representation of a float
differs in the number of bits in the exponent, as well as the mantissa.
Therefore, mismatched types can cause a vastly different answer
on HP 9000 computers.