Re: systems requiring exit?

[Thomas Dickey]

On Tue, 14 Nov 2006, Matthew Woehlke wrote:

> Thomas Dickey wrote:
> > On Mon, 13 Nov 2006, Paul Eggert wrote:
> > > If they conform to C89 there should be no problem.  In C89, returning
> > > N from 'main' is like 'exit(N)'.  This thread is about pre-C89 systems
> > > where the return value from 'main' is mangled.
> >
> > For example (reading the comments that you made earlier, perhaps you should 
> > learn something old about C89):
> >
> > > In that case, it sounds like we should suggest to people that they use
> > > the POSIX environment rather than try to port to a C-only environment.
> > > But I am surprised that a C-only environment would mess up in this
> > > way, even on mainframes.
> >
> > VMS, where an application can return EXIT_SUCCESS or EXIT_FAILURE from main 
> > - but (still complying with C89), those are not 0 and 1 respectively.
>
> Isn't that exactly what I said? :-) (I even used VMS...) You may want to 
> re-read the thread; I think the consensus was that this is a different issue.

I read it, didn't care for the response.  But since you ask:

        hmm - VAX's were introduced as large minicomputers (if you're
        thinking that everthing bigger than a PC is a mainframe, I
        suppose that's a reflection of when you picked up the term).

        VMS uses more than one value for success - typical applications
        add a mask to quiet messages on exit.  There's also a field used
        to specify a message to associate with the code.  But as noted,
        EXIT_SUCCESS and EXIT_FAILURE are defined properly for ISO C
        applications.

        It doesn't need exit.  No point in commenting if all you can say
        is "maybe".

        "has to deal" sounds like the sort of comments that Eggert makes -
        which caught my attention on this list.  A shame, when a few
        moments reading on the web would have answered the question.

quote from the referenced message:

> Hmm, mainframes? I want to say VMS also uses 'return 1' for success (or
> maybe it even needs exit; I'm not the one that has to deal with our VMS
> systems). But as I recall VMS doesn't have a POSIX shell either. Is there
> any VMS support that anyone is worried about?

--
Thomas E. Dickey
http://invisible-island.net
ftp://invisible-island.net