Re: The target for a numeric operation is too small!!!!!

[John Hall <jhall@xxxxxxxxxxx>]

Anton Gombkötö wrote:
> 
> >
> >All he is asking for is the ability to minimize the damage!  The program
> >should HALT but you should be able to continue processing once you have
> >identified the problem.
> 
> Really? Minimize the damage?

Yes Really - If there is one bad record out of 1,000,000 I can fix it
quickly.  If I need to restore 3 dozen files & re-apply the journals
from last night to the present that can take hours.  And when I finish I
then go in and fix the one record that was bad to begin with !!!!

> 
> Many years ago, when i was responsible for running such jobs "bombing" with
> division by zero (there were some :-),
> i had to find the cause, restore to a saved status and re-run the job. With
> save files nowadays, this should be much faster and easier.
> 
>  From experience: the consequences of such errors are seldom investigated
> deep enough when there is an I option, and correcting that, some time
> later, is much more work & pain. And users are willing too easily to enter
> n1 to n2 times "I" before they finally cancel, which is also not minimizing
> the problem.
> 
> Possibly better than ignoring might be starting a debug session and guiding
> the program execution through the wild, providing proper values. I didn't
> try it now, but i do think that it is possible to start the debug, set a
> break point, enter a "G" and do the rest in the debugger. (I bet that this
> program (still) has it's observabilities... :-)
> Pretty the same work as correcting it.
> 
> So i still think that it is quite good and
> Mit freundlichen Grüssen / Best regards

Um, just what processing do you think follows when you do a G while
processing a full procedural file?
Based on a key read from a primary file ?  What if other records are
being updated ?   If "I" was available I could go into the debugger &
correct the value and continue processing.  G does not permit this in
many cases.  

If this is only a report then it really doesn't matter - just cancel and
correct the data or program and then Retry from the calling program. 

Understand I don't want the system to assume or hide anything - I merely
want it to allow me to correct a data / programming error when it is
detected.  

John Hall
+---
| This is the RPG/400 Mailing List!
| To submit a new message, send your mail to RPG400-L@midrange.com.
| To subscribe to this list send email to RPG400-L-SUB@midrange.com.
| To unsubscribe from this list send email to RPG400-L-UNSUB@midrange.com.
| Questions should be directed to the list owner/operator: david@midrange.com
+---