In a message dated 12/13/1999 1:14:49 PM Pacific Standard Time,
[log in to unmask] writes:
> Another is the (nonportable) side-effect that on many
> implementations, executing an ENDFILE on a unit will flush the associated
> buffers to the filesystem at a cost less than that of closing the unit and
> reopening it.
And I am convinced that some of the systems I have used do nothing at all
with ENDFILE, other than what is done anyway when the file closes or the
program terminates normally. The most serious consequence is that incorrect
use of ENDFILE has no consequence until the program is tried on another
system.
Tim
[log in to unmask]
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|