(Forwarding from main Hercules group to the H390-DOS/VS group)

Seymour J Metz wrote:

AFAIK, everything from DOS/VSE through VSE/SP can exploit ECPS:VSE.
Neither VSE/ESA nor Z/VSE can.

DOS/VSE is free; AF, SSX, SP et al are licensed.

Well that's what I'm looking for! A very early version of DOS/VSE that supports ECPS:VSE.

If anyone can manage to locate one, PLEASE CONTACT ME!

Thanks.

--
"Fish" (David B. Trout)
Software Development Laboratories

mail: fish@...

Is the printer 00E defined within the DOS/VSE supervisor?
Has the printer been defined to POWER?

If neither of these things have occurred you will need to change source
code and reassemble the affected program without a 3270 for source
modifications. During this era of DOS, source modifications using a
3270 could be done through use of a program product that ran within
CICS. Both the product and CICS were proprietary for charge products.
I am not aware that either are available today.

Obviously 00E would need to be added to the Hercules configuration
file.

Working with DOS/VSE will be a challenge as Mike described. At least
getting started. This was the era where a lot of work was done with
punched cards.

Harold Grovesteen

开云体育

I googled it,but what I found does not seem to work.? I also thought of 'ADD 00E PRT1' to IJSYSRES.SYSLIB($IPLE45.PROC);
but I have not figured out how to edit the proc yet.


Rahim

Thanks for the explanation Mark. I guess it’s another “I wish IBM would release this to public domain”.?

I have used the KICKS program that’s included with VM/370 6PExt. It’s kinda neat but I’d rather use native DOS/VS CICS. ?Oh well. . .

开云体育

Hi Mark,

Thanks for the clarification, this was more or less what I expected was the case.

I have also had a look at KICKS which look quite interesting and porting this to DOS/VS could be an interesting challenge that I might take up. Hopefully it's written in assembler which happen to be my favourite programming language. I will also take a look a your suggestions liek "BEST".

BR,

Steen

There were a number of CICS "clones" or "TP monitors" back in the day.? WESTI, TaskMaster, one from GE? (IIRC, cannot recall the name right now ...), among others.

It would be great if someone could find a copy of any of those, but I think they all disappeared as those companies got out of that business, since CICS gradually came to dominate the market for "on-line transaction processing" so chances of finding any of that stuff remains slim.

There was one that was made available (open source or public domain) "FTPM" that ran under the Siemens BS2000 MVS "clone" OS, IIRC, that was made available on one of the old Hercules mailing lists, but it was all written in German, so that made it more difficult for anyone wanting to try to "port" it from BS2000 to the real MVS 3.8J or DOS/VS, etc.

There are also some early TSO-like terminal monitors that ran under OS/360, such as BEST from Baylor U., that can be found on some of the old SHARE SPLA tapes available via CBTTape.org ...? Perhaps someone could adapt some of them to run under DOS/VS, as part of an effort to re-create some CICS "clone" on DOS/VS?

And there is a modern CICS "clone" called KIKS that runs under VM/CMS or MVS/TSO, but it relies on VM or TSO doing all the multi-user multi-tasking stuff, hence why I mentioned BEST and similar programs above.

Hope that helps,

Mark

The problem is, CICS/OS and CICS/DOS, and CICS/OS/VS and CICS/DOS/VS, were always "pay for" licensed program products.? So, even if someone could find an old enough copy of CICS/DOS or CICS/DOS/VS that could run on the public domain DOS/VS Rel. 34 system, it would not be legal to do so.

However, all is not "lost" -- there is the KIKS "clone" of CICS that currently runs under MVS/TSO and VM/CMS.? It would take some effort, but someone could port that code to run under DOS/VS, to create a "clone" of CICS for DOS.?? Note that I am not suggesting that such a project would be trivial, by any means.?? The KIKS clone currently relies on VM/CMS or MVS/TSO to provide the support for multiple users and terminals.?? Someone would need to write code to handle that, much as the original CICS did.?? There are other examples of public domain code that shows how to create a multi-user teleprocessing environment, e.g. using BTAM or TCAM, such as the "BEST" system from Baylor that was a kind of "TSO clone" back in the day and ran under OS/360, even before TSO became available.? You can find some of this software on some of the old SHARE SPLA Tapes that are available from CBTTape.org.

I hope that helps.

Hi Steen and welcome to the group. ?I’m encouraged by your request and would second your motion for CICS. I’ve been dabbling with COBOL, Assembler, and RPG code and I too have experience with macro and command level CICS. I hope you receive good responses on this as I am interested also.?

Jim Snellen

Fantastic feedback as always.

Thanks everyone.?

Jim

(piggyback)

Harold Grovesteen wrote:

Jim Snellen wrote:

[...]

This is the JCL to assemble the CDMOD. What am I missing
to get it to IJSYSRL?

Note, you do not have to catalogue the relocatable version of
the xxMOD, CD in this case. You can also assemble it with the
DTFxx macro in the same assembly.

Catalog the module only if you plan to use it in other programs.
That saves repeated increase in assembly size.

In the past, I would assemble the xxMOD in the same program.
After I was sure the parameters were correct and the program ran
correctly, I would then catalog the module if that made sense.

Agree 100%. Very wise words.

As far as getting the results of a CDMOD assembly into the system relo library (IJSYSRL), I've personally always used the POWER JECL "DISP=I" (insert) technique:


* $$ JOB JNM=QSORT,USER='FISH',CLASS=0,DISP=D
* $$ LST CLASS=A,DISP=D
(1) * $$ PUN CLASS=A,DISP=I
// JOB ASSEMBLE AND CATALR QSORT
(2) // OPTION NOLOG,DECK,NOLIST
(3) // ASSGN SYSLST,IGN
// EXEC ASSEMBLY
(4) PUNCH '* $$ JOB JNM=QSORT,CLASS=0,DISP=D'
PUNCH '* $$ LST CLASS=A,DISP=D'
PUNCH '// JOB CATALR QSORT'
PUNCH '// EXEC MAINT'
PUNCH ' CATALR QSORT'
END
/*
(5) // RESET SYSLST
(6) // OPTION LOG,DECK,XREF,NORLD,LIST
// EXEC ASSEMBLY
QSORT CSECT ...

...QSORT source code...

END ,
/*
// OPTION NOLOG,DECK,NOLIST
// ASSGN SYSLST,IGN
(7) // EXEC ASSEMBLY
PUNCH '/*'
PUNCH '/&&'
PUNCH '* $$ EOJ'
END
/*
/&
* $$ EOJ


1. The "DISP=I" punch disposition causes the job's punch output to be inserted into the reader queue instead. The job's generated punch output is of course not inserted into the reader until the POWER job ends.

2. In the first and last steps, you use specify the DECK and NOLIST options since you're not interested in the assembly's listing, only its punch output.

3. Again, you're not interested in the first and last step's listing, so you TEMPORARILY ignore syslst.

4. The PUNCH statements in the first/last assembly steps are used to punch the JECL/JCL needed in the subsequently "inserted" POWER job's jobstream to catalog the results of the second step's assembly. In this example case we're cataloging the object code for the QSORT routine, but in your case it would be your CDMOD assembly.

5. "Undo" the temporary ignoring of syslst, thereby restoring the default permanent syslst assignment, since for the next assembly we actually want to see the assembly listing.

6. Here we specify the DECK and LIST options, because we're interested in both the assembly's punch output as well as its listing too.

7. The last assembly step, like the first one, is used to punch the resulting catalog jobstream's ending JCL/JECL.

Submit this job and VOILA! Immediately after it runs, another POWER job will magically run to catalog the result of the previous job's output into the relocatable library.

I used the same technique to assemble and catalog macros into the E. source statement library.

But I would strongly heed Harold's advice beforehand to ensure your CDMOD is absolutely correct and the way you want it before submitting the job to catalog it into the system relo library.

Good luck! Hope that helps!

--
"Fish" (David B. Trout)
Software Development Laboratories

mail: fish@...

On Mon, 2021-01-11 at 17:59 -0800, Jim Snellen wrote:

Hey guys,
I'm developing an Assembler program to read a card file and write
tape records. On first good assembly, I got an EXTRN on module
IJCFZIWO. I found another topic that resembled my issue (Missing
module?) in which a I discovered a hint. In that topic, it was
suggested to remove WORKA=YES from the DTFCD and voila, I got a clean
assembly. Upon execution, I receive a data exception (?) which
appears to be on the OPEN statement.
In the meantime, I thought I would generate the module using the
CDMOD macro but couldn't remember the proper JCL to CATALR it.
Again, I'm trying to remember stuff but can only recall fragments.

1. 0S03I PROGRAM CHECK INTERRUPTION - HEX LOCATION 40009A -
CONDITION CODE 0 - DATA EXCEPTION
40009A - 400078 (pgm load address) = 22. 22 bytes into program
is OPEN statement.
Is this approach correct? It's been awhile since I read a
dump!!
2. CARDIN DTFCD DEVADDR=SYSIPT,
X
IOAREA1=IOARIN1,
X
BLKSIZE=80,
X
DEVICE=2540,
X
EOFADDR=CARDEOF,
X
TYPEFLE=INPUT

* WORKA=YES commented out!
Is this DTF proper?
3. IJCFZIWO - I seem to recall a manual which decodes the letter

s FZIWO or ZIWO and indicate what
the proper statements are supposed to be.
What is the name of that manual and does anyone have an
electronic copy?
4. * $$ JOB JNM=CDMOD,CLASS=0,DISP=D

* $$ LST CLASS=A,DISP=D,JSEP=0

// JOB CDMOD ASSEMBLE IJCFZIWO CARD MODULE

// OPTION LINK,DECK

// EXEC ASSEMBLY

CDMOD DEVICE=2540,
X
TYPEFLE=INPUT,
X
WORKA=YES,
X
SEPASMB=YES


END

/*

// EXEC LNKEDT

/*

/&

* $$ EOJ
This is the JCL to assemble the CDMOD. What am I missing to get
it to IJSYSRL? Thanks.
Jim

Note, you do not have to catalogue the relocatable version of the
xxMOD, CD in this case. You can also assemble it with the DTFxx macro
in the same assembly.

Catalog the module only if you plan to use it in other programs. That
saves repeated increase in assembly size.

In the past, I would assemble the xxMOD in the same program. After I
was sure the parameters were correct and the program ran correctly, I
would then catalog the module if that made sense.

Never mind on item 3.? I found LIOCS modules?the COBOL Compiler and Library Installation Reference Material, SC33-6469-4, pg 10.

Never mind on item 4.? I found an example in the System Control Statements, GC33-5376-5, pg 64.

Hey guys,
I'm developing an Assembler program to read a card file and write tape records.? On first good assembly, I got an EXTRN on module IJCFZIWO.? I found another topic that resembled my issue (Missing module?) in which a I discovered a hint.? In that topic, it was suggested to remove WORKA=YES from the DTFCD and voila, I got a clean assembly.? Upon execution, I receive a data exception (?) which appears to be on the OPEN statement.?
In the meantime, I thought I would generate the module using the CDMOD macro but couldn't remember the proper JCL to CATALR it.? Again, I'm trying to remember stuff but can only recall fragments.

1.? 0S03I PROGRAM CHECK INTERRUPTION - HEX LOCATION 40009A - CONDITION CODE 0 - DATA EXCEPTION
? ? ?40009A - 400078 (pgm load address) = 22.? 22 bytes into program is OPEN statement.
? ? ?Is this approach correct?? It's been awhile since I read a dump!!
2.? CARDIN? ?DTFCD DEVADDR=SYSIPT,? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?X?

Funny you mention rewriting to Assembler Harold, that’s what I did and the DTFIS macro assembled just fine. Thanks for the feedback.
Now I’m gonna look at the System Control Statements to refresh my memory on ISAM JCL.

Best regards,
Jim