On Wed, Oct 14, 2015 at 4:08 PM, Charles Mills <charl...@mcn.org> wrote:
> Didn't IBM bring one of those missing NUMPROC options back in COBOL 5.2 or > am I confused? > > Charles > This "problem" (hardware checks invalid data & terminates), to me, to violate the "spirit and intent" (as my Colonel used to say) of COBOL. I do understand the comfort of knowing that the hardware itself will not allow you to "ADD +1 TO 'A' ". But that, to me, makes it harder on future programmers. Instead of actually _testing_ something and giving an understandable error message, you get a LE abend. Which must then be traced back to the reason, an S0C7. Which must then be tracked back back to the COBOL variable which has the bad data. And then figure out where _that_ came from. Thank God for products like AbendAid which can do most of the drudgery. This is another reason why I like production quality RDMS systems. If you define the column (attribute) as an integer, then it _will_ be an integer (or money or ???). You can't stick a character string into it. (An exception to this is the SQLite data base. I love SQLite, but it does not validate data types in any way. Dr. Hipp, the designer/programmer, did this deliberately). Also, it makes it more problematic to upgrade, as Lizette has found. And much harder to "port" to another architecture. Hum, this last might be a "good" reason to not check (increases the cost and danger of conversion, which makes it less desirable). Case in point for the above, from the misty past. This was in the 1970s on OS/VS1. The original programmer got to a "this cannot occur" situation. Rather than DISPLAY .... UPON SYSOUT and call ILBOABND (I think), he (I say he because the women programmers tended to not do this silliness) caused a deliberate S0C7. Unfortunately, he was now gone. And we had just done a DOS to VS1 conversion, so the other programmers didn't know how to debug it. I ended up (just out of college) doing it, looking at the listing, and reporting on the idiocy. This wasted about 2 hours (newbie, remember). That firmly set my opinion of not checking for valid data. Of course, the above is my _opinion_ and others will disagree. Hopefully we can disagree respectfully. -- Schrodinger's backup: The condition of any backup is unknown until a restore is attempted. Yoda of Borg, we are. Futile, resistance is, yes. Assimilated, you will be. He's about as useful as a wax frying pan. 10 to the 12th power microphones = 1 Megaphone Maranatha! <>< John McKown ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN
