On Thu, Mar 29, 2012 at 12:31 AM, John Cowan <[email protected]> wrote:

> Stefan Edwards scripsit:
>
> > I guess what I'm driving at is for the standard to explain how to treat
> > items that fail the above instance of
> > translation into appropriate values, in some standard way.
>
> I don't see how: the issue seems about as implementation-dependent
> as anything could be.  What approach should the standard provide
> for converting Plan 9 error strings into the far more common Windows/Posix
> numbers from 1 to 255?  (I assume that APE has a reverse convention,
> but I can't find any documentation of it online.)
>

"For any value that does not map to an operating system acceptable exit
value, and is not a boolean,
it is the recommendation of this report to treat it as a true conditional,
for purposes of  creating an exit value."

Which seems pretty hairy to me. As I said, it was only a thought, rather
than a deeply held belief, but it seems
reasonable to have a standards-compliant method for dealing with objects
that cannot be mapped.


> --
> "Repeat this until 'update-mounts -v' shows no updates.         John Cowan
> You may well have to log in to particular machines, hunt down
> [email protected]
> people who still have processes running, and kill them."
>



-- 
====
Q. How many Prolog programmers does it take to change a lightbulb?
A. No.
_______________________________________________
Scheme-reports mailing list
[email protected]
http://lists.scheme-reports.org/cgi-bin/mailman/listinfo/scheme-reports

Reply via email to