At 11:29 AM 8/20/2001 -0400, Dan Sugalski wrote:
>At 10:23 AM 8/20/2001 -0500, Craig A. Berry wrote:
>>Thanks, Carl, that's exactly the sort of stuff I was looking for. I'll see
>>if I can work up a patch to VMS::System that uses this to figure out whether
>>we can request the environmental stuff from $getsyi.
>
>You want VMS::System? (Or the other VMS modules?) You're certainly welcome to it. :)
I'm more interested in renting than buying at the moment, and some of this
stuff is twisty enough that it wouldn't hurt to have one person patching and
another reviewing and applying. (Yes, I've found a couple other latent bugs
unrelated to the current discussion ;-) ).
>FWIW, and I'm coming in really late here, I don't think the problem is actually in
>the system service routine, but I'm not 100% sure. The original error looks familiar
>and may well be caused by something else. (Like a bug in VMS::System) Those four
>status routines are supposed to be universally available in 7.2...
VMS::System has the exact same problem f$getsyi() does, except that
f$getsyi() can't give you a hash with all the items in one fell swoop. It's
not so bad to be told you can't get a particular item, but to be told you
can't have your list of all items because a handful of them aren't doable on
your system isn't nice. It might be considered a bug in the $getsyi service
that it blindly depends on the cpu routines to feed it data, not recognizing
that some cpu routines can't even say "no can do" intelligibly (if that is
indeed a fair description of what's happening).
