Re: Hooking display file I/O

[Dave McKenzie <davemck@xxxxxxxxxxxxx>]

Yes, I got that point.

My point was that while you expect that only your test job will be affected 
because you're changing only its own PCO and static storage, in fact if you 
start a group job from that test job, the group job's PCO will get the 
UPDATED pointer which points to the SEPT copy in the test job, not QINSEPT in 
QSYS as new jobs normally do.  Then if you end the test job, even normally, 
the group job will crash.

Of course, the group job is also "your" job--it's not a job run by someone in 
the Accounts Payable department, so it's not a catastrophe.

My point was that another job CAN be affected (albeit still "your" job), even 
though you would expect that your changes would be confined to the test job.

Naturally, I understand that if you put the technique into production, you 
would take steps to avoid this pitfall.

--Dave


On Sunday 05 August 2001 20:36, Leif Svalgaard wrote:
> From: Dave McKenzie <davemck@zbiggroup.com>
>
> > I remember once when I did something similar, I had trouble with group
>
> jobs.
>
> you are missing the point.
> The point is to make a local copy of the SEPT for testing
> purposes. If you screw up, only your own job is affected.
> Once everything works, it is a different matter.
+---
| This is the MI Programmers Mailing List!
| To submit a new message, send your mail to MI400@midrange.com.
| To subscribe to this list send email to MI400-SUB@midrange.com.
| To unsubscribe from this list send email to MI400-UNSUB@midrange.com.
| Questions should be directed to the list owner/operator: dr2@cssas400.com
+---