RE: Working with IFS Files

Grizzly,

You're right, using api's does give you a granularity of control.  For 
example, if it dies on row 23 you can handle the exception better.  Bob, 
no slam, but moving from IBM's api to your's probably does not give the 
granularity of control wanted.

We've found out that custom writing something can often be faster than 
IBM's CPYTOIMPF.

With the fun and games IBM played with V5R3 and CPYTOIMPF/CPYFRMIMPF it 
left a bad taste in our mouths.  And I am concerned that any of IBM's 
published workarounds to get V5R3 to work like earlier releases may not be 
carried over to future releases.

You may want to consider not using "a big honking character field" to 
cover all anomalies from a particular column of CPYTOIMPF but instead to 
use a variable length character field.  Sounds like an ideal application 
for it.

Rob Berendt