OK, so this was a shop standard rather than any actual requirement.  So not 
"cannot be defined to RDO" but "chose not to be allowed to be defined in RDO by 
management decision".

That kind of thing is a deliberate decision to live with the consequences for 
some perceived benefit.  To each their own.

Peter

-----Original Message-----
From: IBM Mainframe Discussion List <[email protected]> On Behalf Of 
Steve Thompson
Sent: Sunday, April 9, 2023 4:14 PM
To: [email protected]
Subject: Re: Cobol calling module with non alphanumeric no longer allowed???

In the case of a common routine(s) that is (are) statically linked for Batch 
and CICS environments. I didn't design the system the customer had, I just had 
to be aware of it when I was working on Compile support tools.

So I thought that this aspect should be considered in this conversation.

And, yes, if these routines changed, the causing the common
copybook(s) to be changed, that would require a recompile and linkedit of the 
batch and CICS sides, QA testing, and then promotion back into production. So 
that had to be coordinated across a few application groups for each program 
that used one of those "dual" routine programs (dual from the idea that they 
were used in both environments).

Steve Thompson

On 4/9/2023 2:50 PM, Farley, Peter wrote:
> The only CICS routines that MUST be statically linked are the DFH**** 
> subroutines that are called as the result of coding CICS commands like EXEC 
> CICS LINK or EXEC SQL SELECT.  AFAIK, any user-written program can be defined 
> to RDO including those with national characters ($#@) in the name.
>
> I am curious, in what case(s) do you believe that a callable subroutine could 
> NOT be definable in RDO?
>
> Unless it is for very high frequency user programs where maximum performance 
> is critical, statically linked functional subroutines (under CICS or not) are 
> a maintenance nightmare, because when (not if) those subroutines need to 
> change, whether for new function or for defect repair, EVERY program that 
> statically linked them must be re-linked and (more costly and more 
> critically) MUST be re-tested for quality assurance.
>
> And even for high performance requirements, the EXEC CICS LINK (or better, a 
> plain COBOL CALL variable-name, even under CICS) has a fairly efficient path 
> to execute already-loaded subroutines, and if they are being used frequently 
> they are almost surely already loaded in memory.
>
> Peter
>
> -----Original Message-----
> From: IBM Mainframe Discussion List <[email protected]> On 
> Behalf Of Steve Thompson
> Sent: Sunday, April 9, 2023 2:16 PM
> To: [email protected]
> Subject: Re: Cobol calling module with non alphanumeric no longer allowed???
>
> The other reason for the literal calls is because one wants the load module 
> to have the subroutines statically linked that the COBOL program is using. In 
> this way there is no LOAD done during execution time because COBOL "knows" 
> that these are statically linked.
>
> There are a few reasons to want this behavior.
>
> The major one that comes to mind is you have to have static linkage for 
> running in a CICS environment because the subroutines are not CICS definable 
> (RDO) and so have to be statically linked.
>
> Steve Thompson
>
> On 4/8/2023 1:50 AM, Farley, Peter wrote:
>> Not true for non-static calls.  We are past COBOL 5 (V6.2 at the moment) and 
>> "CALL variable USING . . . " where "variable" has any of the "national" 
>> characters ($#@) works every time.  We have multiple dynamically called 
>> utility subroutines with those characters in the program name.
>>
>> Why in the world are you using literal calls?  Or are you using the DYNAM 
>> option to convert literal calls to dynamic ones?  If so, bite the bullet - 
>> convert them to "CALL variable" and you are done.
>>
>> The only legitimate case I have seen for using literal CALL's is when you 
>> are using nested subroutine programs in the same source file as the calling 
>> program.
>>
>> Peter
>>
>> -----Original Message-----
>> From: IBM Mainframe Discussion List <[email protected]> On 
>> Behalf Of Frank Swarbrick
>> Sent: Friday, April 7, 2023 6:07 PM
>> To: [email protected]
>> Subject: Cobol calling module with non alphanumeric no longer allowed???
>>
>> I've tried calling modules (that exist!) with both '@' and '#' signs in them 
>> and Enterprise COBOL 5+ does not allow this.  COBOL 4 allowed this.  Is 
>> there any good reason why this is the case?
--

This message and any attachments are intended only for the use of the addressee 
and may contain information that is privileged and confidential. If the reader 
of the message is not the intended recipient or an authorized representative of 
the intended recipient, you are hereby notified that any dissemination of this 
communication is strictly prohibited. If you have received this communication 
in error, please notify us immediately by e-mail and delete the message and any 
attachments from your system.


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

Reply via email to