David Noon wrote on 07 Jan 1996 at 12:10 to Ruud Senden:

RS> Well, the RC-variable is at least meaningfull with a syntax 
RS> error.

 DN> What you are seeing here is REXX passing something it doesn't
 DN> understand to the command shell, CMD.EXE. If REXX can recognise it
 DN> but finds a syntax error, RC will not be valid.

Well, I have a document here which shows the standard handler for SIGNAL ON
SYNTAX, and it also uses the RC-variable to generate an error-message.
Also, in the appendix with errornumbers and -messages, it says "The
number can be found in the special variable RC after a recoverable
error".

I think it is an official REXXSAA-document (I am missing the titlepage, so
I don't know for sure) from an old version of REXX (which was available as
an IBM Internal Use Only package), since it describes all REXXSAA-features,
together with the installation instructions (for OS/2 1.0 and 1.1),
differences between REXXSAA and REXX370 and between REXXSAA and REXXOS2 (I
think this is an even older version of REXX).

I think many things have changed from then, but the standard SYNTAX-signal
handler listed in this document has always worked correctly in my programs,
so I think this is still correct. But the error-message isn't the same as
the current standard signal handler, for example it doesn't show the nested
procedure-calls.

 DN> These are documented somewhere (Red Books perhaps?), but coding the
 DN> REXX error message table into your program will make it rather
 DN> bloated. 

Well, I just decided not to recreate the REXX error-message, since it is so
difficult (if not impossible). I will just print the most important
information, generated with CONDITION(), SIGL and RC.

Anyway, thank you for your answers.

Regards,
  Ruud  (Internet: rsenden{at}cs.ruu.nl) 

--- timEd/2 1.10+