On 04/29/09 12:13, Glenn Knickerbocker wrote:
Paul Gilmartin wrote:
I'm mystified. What went wrong with F 80?
Try typing 80 blanks (or just one) on the command line in an interactive
trace on CMS and you'll see. When every line is 80 bytes long, there's
never a null line to tell REXX to go on to the next statement of your
exec.
Ah, so my suggestion was exactly wrong.
I'm surprised that TSO (and even more CMS) ever passes an empty
line to the program.
o To CMS EXECIO, a null line terminates input from the stack.
o By default, the z/OS C Run Time Library discards empty lines
on input.
o ISPF (and XEDIT?) editors pad empty lines to one blank when
outputting to a RECFM=V file.
o I had to start an APAR to get Rexx to pass empty lines to user
subcommand environments.
Etc. IBM abhors a vacuum.
-- gil
