Dan Sugalski wrote:
>
> At 05:08 PM 8/15/2001 -0500, Patrick Spinler wrote:
>
> >Anyway, thanks for the answer. It _almost_ works. Try the following
> >script:
> >
> > use VMS::Stdio "getname";
> > open IN, "<mpa0:" || die $^E;
> > print getname(*IN), "\n";
> > system ("show device /full mpa");
> > sleep (5);
> >
> >I get the result "MPA0:[].;". Yet, as the "show device" command
> >displays, my real device is something like MPA8: or MPA30:.
> >
> >Sigh. I'll just have to put the perl in a DCL wrapper.
>
> Maybe. Try stripping off the "[].;" stuff and see what $ENV{MPA0} and
> $ENV{MPA0:} look like.
No luck. And actually, I found something even freakier. Specifically,
I get way different results on vms 7.1-2 decc rev 5.5-002 from vms 7.2-1
decc 6.2-008.
(Same perl installation, rev 5.6.1. This is on a heterogenous cluster)
On vms 7.2, opening mpa0 gets me a clone of the device, as I'd expect.
However, there is no $ENV{MPA*} variable created, and "getname(*IN)"
just returns MPA0:.
On vms 7.1, the open appears to succeed (ie - the die statement isn't
called), but no clone device appears, and the getname call returns
undefined.
Running the following simple little dcl script gives me identical
results on both nodes (device opening clones device & returns handle to
clone)
$ v=f$verify(1)
$ on error then goto err_exit
$
$ open/read mpa_in mpa0:
$ show log mpa_in
$ show dev /full mpa_in
$
$ err_exit:
$ status=$status
$ v=f$verify(v)
$ exit 'status
Any ideas ?
Thanks,
-- Pat
--
This message does not represent the policies or positions
of the Mayo Foundation or its subsidiaries.
Patrick Spinler email: [EMAIL PROTECTED]
Mayo Foundation phone: 507/284-9485
