At 04:49 PM 11/1/2000 -0500, Carl Friedberg wrote:
>My standard comment? This sounds like a *n*x*sm. Carl
Thanks for censoring how you really feel this time :-).
>The answer to your question is, obviously, they can't.
Possibly there is some magical abstraction where the various parts of a
status longword are broken down and put back together in meaningful, but
different, ways. If so, I'm well short of understanding how it works.
>Patrick, you are doing us a great service if only we could end up with a
>consistent way of getting out the relevant VMS status condition after ANY
>perl object touches something VMS...fi
exit() generally works ok, so there does appear to be something wrong with
spawned process returns. do_spawn() in vms.c appears to be where system()
is implemented (we don't use the C RTL version); this looks ok to me, thus
my suspicion of what happens afterwards in pp_sys.c.
_______________________________________________
Craig A. Berry
mailto:[EMAIL PROTECTED]
