In my baby steps on TSO, I got as far as allocating file REXX and running a stage from it, but couldn't see why it was just quietly ending with no output or messages. I finally narrowed it down to where REXX itself was trying to use the console: At the first SAY or TRACE instruction, it kills the whole pipeline with rc 20048 and no message (none successfully delivered to the console, anyway). I assume that's related to this statement in the help:
> 5. On z/OS, REXX filters run in dedicated re-entrant environments. > > Such environments cannot be merged with the TSO environment. SAY is easily enough replaced with CALLPIPE ...|CONS, but is it impossible to trace REXX stages on TSO? Particularly confusing was that I got the same response when the REXX program didn't exist at all, so I was never sure if I was successfully calling the program or not. Is there any way to distinguish between the two conditions? In tracking down the exact circumstances of the error, I tried to use the REXXVARS|RUNPIPE trick to find how deeply nested I was: 'callpipe literal rexxvars' i '|runpipe' and found that for any i>0 it also bombs with rc 20048 after issuing: BPW00233E No active EXEC environment found Is there any way to determine the nesting level on TSO without killing the pipeline? ¬R
