On Wed, 17 Dec 2025 17:59:14 +0100, Radoslaw Skorupka <[email protected]> 
wrote:

>It is not me who used binder improperly. Or messed order of activities.

It was the developer who messed up. They created the SMP/e definitions and you 
don't have any control over this. UCLIN (not ++x) is your only control to SMP/e 
definitions. It was not my intention to blame you.

>However I dare to disagree. It is matter of SMP/E logic. 
>Same tool (binder), two different SMP/E
>command sets which *should* yield same result. But the results differ.

The goal of SMP/e is to "yield the same result" but it does not make this 
guarantee. As developers, we use SMP/e to help us make this guarantee. SMP/e 
has utilities and each developer must use these utilities so the results will 
always be the same regardless of how many FMID's and PTF's. 

Consider the SMP/e Unix utilities (specifically scripts). SMP/e has no control 
over scripts and cannot make any guarantee the developer coded the script 
correctly.

Likewise, SMP/e does not control nor limit how I use the binder. I should never 
use CALL instead of NCAL but it's my choice. Same with INCLUDES inside of 
INCLUDES. I can have multiple CSECTS in a single object. And much more. If I 
use any of these features, it is my responsibility to convey this correctly to 
SMP/e to guarantee the same results no matter how the customer applies PTFs, 
APARs & FMIDs.

You've mistaken SMP/e convenience tools (designed to make developer's lives 
much easier) as SMP/e binder commands. They are very basic and minimalistic. 
Consider a ++MOD (object file) that contains multiple CSECTS. JCLIN does not 
see the CSECTS in the ++MOD because all binder compatible formats are allowed. 
Since JCLIN for INCLUDE statements build a MOD entry with a matching CSECT, it 
is up to the developer to tell SMP/e about the other CSECTS.

----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN

Reply via email to