fixed point arithmetic
Calculation using packed numbers
p). Is calculated with an internal accuracy of 31 or (if not sufficient) 63
decimal places. The
fractional portion is arithmetically rounded, if necessary. Decimal places before the
decimal point are never lost if a calculation is completed successfully. As a
fixed point arithmetic determines whether for numbers of type
p the decimal point is respected by operations or not. The disabling of this program attribute is obsolete.