Comp Type
The Comp
type is a 64-bit signed integer type that
uses the floating-point processor. The Comp
type
exists only for backward compatibility. New programs should use
Int64
instead. The Int64
type
is a true integer type and allows bitwise and shift operations,
supports the full range of the 64-bit type, and is not dependent on
the floating-point control word.
To use the full 64-bit precision of the Comp
type,
the floating-point control word must be set to extended precision.
Some Microsoft and other DLLs change the floating-point control word
to double or single precision, thereby reducing the
Comp
type to 53 or fewer bits.
Delphi’s I/O and the formatting routines in the
SysUtils
unit do not support the full range of the
Comp
type. Values near the limits of the
Comp
type are printed with less precision than is
actually stored in the Comp
variable. Most
applications will not encounter this limitation.