On 5/31/23 15:48, Phil Smith III wrote:
Gil wrote:
OpenVM fork()? It's unforgivable that OpenVM provides
something it calls "fork", but which is not.
Heh. When that came out, I got a bunch of Taco Bell sporks, drilled holes in
them so I could hang them from paper clips, and handed them out at SHARE to the
VM crowd to dangle off badges. A couple of Endicott folks privately told me
they liked them.
I'm not sure it wasn't a reasonable compromise, but I certainly understand the
distaste.
Reasonable compromise ... it would seem so.
IBM are such sticklers for the rules. What really *is*fork() and what
does Unix do on systems without virtual memory? This is a real question
given that there is a group implementing Unix for 8-bit systems. (Plus
the history that Unix originally ran on many HW platforms which lacked
the ability.) So what canfork() do in cases like that? That might be a
pattern OpenVM could follow.
CMS doesn't do virtual memory ... and that's a feature.
But ... yeah ... if you *know* that some programs will (in the source)
be lazy and just callfork() to spawn a child, and immediately exec(),
sher, makes sense. But more rigorous development would callspawn() instead.
Call me lazy.
-- R; <><
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN
