Re: V5R2 --> V5R3 CPYFRMIMPF and CPYTOIMPF

I found that a very frustrating part of the V5R2 to V5R3 move.

1 - It seemed that the change did not affect CPYTOSTMF. Granted
CPYTTOIMPF gives you more functionality. However one workaround is to
first do a CPYTOSTMF then take the output from that and run CPYTOIMPF.

2 - Make sure you start using the right CCSIDs etc on your files. I did a
CHGSYSVAL on QCCSID to 37 during the middle of the day with no negative
ramifications.

3 - What Pete said.

4 - Bypass using either command. Instead use the IFS APIs as documented
by Scott Klement. Benefits include:
- Getting around IBM's V5R3 attempt to thwart productivity.
- They actually run faster than those commands.
- Better granularity of control. For example, if the 52,763rd row of the
file you are running CPYFRMIMPF fails because there is an entry in the
column for item number that is bigger than your field size you can put it
on an edit report to handle as an exception. Used to have this error ALL
THE TIME!

Run some of the jobs performing CPY...IMPF under debug. See IBM use a
flurry of SQL commands under the covers.

I really, really, really, like SQL. However, judging by some of the
implementations I've seen coming out of IBM, I sure wouldn't contract one
of their developers to come and advise me on SQL performance tuning.

Rob Berendt