🚀 go-pugleaf

In <40jajf$mt0@gap.cco.caltech.edu> carl@SOL1.GPS.CALTECH.EDU (Carl J Lydick) writes:

>In article <danpop.808262674@rscernix>, Dan.Pop@mail.cern.ch (Dan Pop) writes:
>=In <40iqa1$4eg@gap.cco.caltech.edu> carl@SOL1.GPS.CALTECH.EDU (Carl J Lydick) writes:
>=
>=>43, F for 44, and ? for 45-47.  Similarly, messages 1-7 are
>=>"%SYSTEM-x-NORMAL, normal successful completion", while 0 is
>=>"%NONAME-W-NOMSG, Message number 00000000".  Note:  The C RTL, in a fit of
>=>accomodation to Unix, converts an exit status of zero to 1.
>=
>=Nope.  It does this in a fit of ANSI/ISO C compliance.  Before the ANSI
>=standard (and some time after :-) the VMS C RTL couldn't care less about
>=the Unix conventions WRT exit status codes.
>
>Hmmm.  I thought the ANSI standard required that a successful exit status be
>defined by keyword.  The exit() function in the VAX C RTL converted zero to one
>before the ANSI standard was adopted.  return, on the other hand, left the
>value untouched.

The ANSI standard guarantees that returning a value from (the initial
invocation of) main() has the same effect as calling exit() with that
value as argument.  This is what DEC C does nowadays.  ANSI also
guarantees that 0 and EXIT_SUCCESS have the same effect when passed to
exit(), but it doesn't require that EXIT_SUCCESS is defined as 0 (and
DEC C for OpenVMS doesn't define EXIT_SUCCESS as 0).

Dan
--
Dan Pop
CERN, CN Division
Email: Dan.Pop@mail.cern.ch 
Mail:  CERN - PPE, Bat. 31 R-004, CH-1211 Geneve 23, Switzerland