If you want TSO/E REXX to wait only a certain duration after having called 
another program (ISPEXEC), there is no such function. REXX waits for the 
called program to return control. Sounds like you want ISPF to have an option 
that when a panel is displayed it would 'timeout' if the user did not press an 
interrupt key (ATTN). Not the normal behavior for an interactive application. 
Similar to how some other functions, CONSOLE() and SOCKET() come to mind, 
have their own settings for how long they can have control before returning 
control back to the caller.

ISPF was designed to be interactive, so it waits for the user to interact, not 
optionally interact. Just like CICS displaying a map and waiting for the user 
to 
do something, then running the transaction without letting the user start 
another transaction or press a PFKEY. In CICS it could ugly if users tried 
running three transactions at the same time that all displayed a different map 
to be used for data entry. Same with ISPF. The CONTROL LOCK option seems 
to be trying to match the windows 'progress bar' method of keeping the user 
occupied while doing things that take awhile and let them know the system is 
fine, just has a lot to do. That is how ISRDDN search uses it, and I use it to 
display which member my PDSEDIT code is currently executing on. Since you 
have just the one active TCB doing your work there is no ISPF reason to let 
them press any keys while your Rexx code is doing your work. They could not 
swap to another screen because you did not spin off a disconnected process 
under another TCB. You can't scroll around because your code is still 
executing. Generically it makes no sense to let the user change the display if 
your program is going to display a different panel. ISPF is not a multi-tasking 
operating system, only one TCB is active at a time, when control returns to 
them, they can swap to another ISPF session/screen and do something else, 
there are no swap but leave the session running in the background options. It 
stops running when they swap. Although I have wanted the ability to scroll 
down a long data set list while something else was chugging away. And having 
to press the attention key to break out to my multi-session monitor so I can 
swap logical terminal sessions is dangerous if I hit it twice, it sends the 
attention to my rexx code.


Ask John L. to open an ISPF Requirement for a TIMEOUT parameter on 
ISPEXEC SELECT and DISPLAY PANEL().


It would be nice to have some code running under session 1, swap to session 
2 and leave session 1 running while continuing to do work under session 2. 
The lack of that functionality is what led users to have multiple tso userids 
and being able to logon with the same userid on multiple systems in a 
JESPLEX. And more recelty asking why not use the same userid more than 
once on the same system.





On Fri, 19 Sep 2008 14:28:25 -0400, Barkow, Eileen 
<[EMAIL PROTECTED]> wrote:

>Dave,
>thank you very much for your synopsis on this.
>we were trying to find a way whereby the process can be interrupted.
> The programmer did manage to do this with an assembler program that
>waits and accepts an interrupt from the attention key - if none is
>received the panel is jus reinvoked with the time updated.
>
>unfortuneatly, I do not see a WAIT facility in TSO REXX - there is one
>in Netview- which would have eliminated the need for an assembler
>program.
>
>you are correct about the attention key, but that can be set with
>emulators.
>

----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Reply via email to