SPUG: OT: Re: PERL technical interview
sthoenna at efn.org
Mon Jan 23 13:29:17 PST 2006
On Mon, Jan 23, 2006 at 02:30:14PM -0600, Jeff Almeida wrote:
> *d'oh!* now you know why i don't code all that much in c
> *embarrassed grin*
> leave it to me to confuse shell-true with actual-true on a monday
Gah. I didn't notice either, just translated bin/true to EXIT_SUCCESS.
That's a good reason to use EXIT_SUCCESS/EXIT_FAILURE :)
I changed the signature of main() because I remember there being
at least one compiler that would give an error if the usual parameters
weren't specified. Don't remember what compiler/platform that was,
or if envp was needed or not, now.
WRT exit values, I think it was the CCS/C compiler on MPE/V (and maybe
MPE/XL aka MPEiX) that supported MPE's concept of some return values
indicating a warning rather than an error, so that EXIT_FAILURE needed
to be high enough to indicate an actual failure. I think the HP C
compiler has always translated any non-zero into a failure, though.
More information about the spug-list