RE: Compiling to V7R1 from a V7R3 Test LPAR fails

Sue--

Thanks for the info. I added it to our V7R3 test bed and it worked! Our programmers have compiled some programs, and so far they're working as intended.

I added the change to our CRTCMDDFT program for the future.

The strange thing is we've never had that option selected on our V7R1 machine, and everything seems to work fine...

Paul E Musselman
PaulMmn@xxxxxxxxxxxxxxxxxxxx

.

-----Original Message-----
From: MIDRANGE-L [mailto:midrange-l-bounces@xxxxxxxxxxxx] On Behalf Of Sue Romano
Sent: Thursday, October 20, 2016 2:57 PM
To: midrange-l@xxxxxxxxxxxx
Subject: Re: Compiling to V7R1 from a V7R3 Test LPAR fails

I think you are missing the RPGPPOPT(*LVL2) parameter on your precompile
to handle the conditional /IF directives that are in the include file. I
need to specify that option on both 7.1 and 7.3 releases when trying to
precompile your source. Maybe you missed some command customizations
during your upgrade?

Sue Romano
IBM i SQL Development

Oh, what fun!

We're migrating from V7R1M0 to V7R3M0 and have created a test LPAR from

our secondary system. It's a Full-System Flash Copy.

We flash copy the LPAR, IPL, then run through the items in the 7.1 to 7.3

instruction manual: Load some stray PTFs, apply PTFs *PERM (with >the IPL
to make them so). Then we run the upgrade: Image Catalog loaded with the
new OS DVDs, Cumulative package C6085730, HIPERs, GROUPS, >etc.

The upgrade runs; some old products don't have upgrades available (ie

Office Vision and WebSphere 5). INZSYS completes normally. CHKPRDOPT

*OPSYS shows everything looks OK.

Then our programmers started poking at things, and ran into a problem

compiling, using CRTSQLRPGI (from the STRPDM interface in the test >LPAR).
This creates an intermediate source, and then compiles with CRTBNDRPG.
The program is being compiled using TGTRLS(V7R1M0). This is >because our
production LPAR won't be upgraded for a few months, but our second system
will be upgraded by the end of October.

The program fails to compile under V7R3. Apparently, it doesn't copy in

the SQLCA code that's supposed to be included automatically. The >library
list includes the our production source and the programmer's source
libraries.

The program compiles fine in the secondary system under V7R1.

We have a ticket open with IBM, but sometimes this group is quicker! (;

Anyone have any thoughts?

Thanks for any light you can shine on this fun!

--Paul E Musselman