RE: Record commitment and SVCPGM question

I noticed something similar in service programs that update or write to
files. Since we do a lot of interactive updating, I routinely close the file
after each iteration of a write/update subprocedure. Once the file closes,
the change is in the file. This practice does not slow interactive jobs
noticeably, but I haven't benchmarked it in batch jobs. And I get rid of the
level 10 message on module compiles. 

Perhaps we should look into the POST opcode, as Booth suggests, though it
looks like it would take more time than merely closing the file. When I
wrote about this practice a few months ago, at least some members of the
list thought this practice was unnecessary. Perhaps other list members will
once again share with us why this isn't the best way to handle the problem.

HTH,
Roger Mackie

-----Original Message-----
From: George Kinney [mailto:GKinney@xxxxxxxxxxxxxxxx]
Sent: Monday, October 27, 2003 12:49
To: 'midrange-l@xxxxxxxxxxxx'
Subject: Record commitment and SVCPGM question

<snip>

What bothers me is that I can't see the records that are being written until
after the SVCPGM terminates.  Is this normal? I am not using any commitment
control that I am aware of.