Re: [GNU-COBOL] New to list

["Boris Kortiak" <BKortiak@genlyte.com>]

>>> Fred Mobach <fred@mobach.nl> 10/28/99 04:20PM >>>
Boris Kortiak wrote:

> >>> Fred Mobach <fred@mobach.nl> 10/28/99 11:23AM >>>
BK-
<snip>
> Actually this is an extension I have seen mostly with Burroughs.  I have also seen a COMP-6 extension (mbpCOBOL of Germany) for unsigned BCD.  It's nice to have when you don't need a sign, but not an absolute requirement.

FM-
It's indeed not so important for new programs but try to imagine what the consequences are for converting old programs that use that kind of fields in redefined area's :

01  AREA.
     03  GROUP1.
        05  FIELD1   PIC XX.
    03  FILLER REDEFINES GROUP1.
        05  NUMFIELD1 PIC 9(4) COMP-3.

Without the sign in NUMFIELD1 both groupfields are 2 bytes long. With the sign in NUMFIELD1 the group FILLER is 3 bytes long. It's a hassle to convert this kind of constructions and you can find them anywhere :-).

BK-
I think getting to the standard first is more important.  These things can be handled later by adding dialect switches -or- extensions to allow for this.

But your absolutely right that for this to be a viable COBOL replacement it will have to account for the various dialects in use.  Don't forget that a gnu project, the source for the compiler will be available to you to extend any way you wish <g>.


--
This message was sent through the gnu-cobol mailing list.  To remove yourself
from this mailing list, send a message to majordomo@lusars.net with the
words "unsubscribe gnu-cobol" in the message body.  For more information on
the GNU COBOL project, send mail to gnu-cobol-owner@lusars.net.