[Search for users] [Overall Top Noters] [List of all Conferences] [Download this site]

Conference noted::hackers_v1

Title:-={ H A C K E R S }=-
Notice:Write locked - see NOTED::HACKERS
Moderator:DIEHRD::MORRIS
Created:Thu Feb 20 1986
Last Modified:Mon Aug 03 1992
Last Successful Update:Fri Jun 06 1997
Number of topics:680
Total number of notes:5456

50.0. "Terminals/mailboxes" by GLIVET::DIAMOND () Thu Aug 23 1984 12:42

Maybe someone out here can tell me...

How do you deassign a mailbox from a terminal that you previously assigned in 
supervisor mode?  I've got a program that sets up broadcasts to go to this 
mailbox even after image exit (hence the supervisor mode).  Then when I go
back later to delete the mailbox mailbox, I first deassign the terminal
(again, in supervisor mode).  But the mailbox won't go away because the 
terminal never deassigns and the mailbox is still assigned to it.

Other symptoms are that I get "device active" errors when trying to use SET 
HOST, SPAWN/WAIT, or WHAT.

Any suggestions as to what I may not be doing to correctly deassign the 
terminal?

Dave
T.RTitleUserPersonal
Name
DateLines
50.1ROYCE::KENNEDYThu Aug 23 1984 22:2412
I think that your problem is that the mailbox/terminal connection
isn't being broken. You could experiment by saying SET /NOBRDCSTMBX
to DCL. The assignment may be removed by a SETMODE QIO to the terminal
with TT2$M_BRDCSTMBX off (See chapter 9 in the I/O Users Guide Vol 1).
The mailbox should then be deletable.

The reason why WHAT was giving an error is that it uses a broadcast 
mailbox and only one may exist per terminal at a time.

Good Luck

Hugh.
50.2TURTLE::GILBERTThu Aug 23 1984 19:2815
Is there any good reason why only one mailbox may be associated with a terminal
at a time?  You could either put in a Y-valve to direct associated messages to
all the mailboxes, or a switch to direct them to the desired mailboxes, or a
combination of both (if this were handled by LIB$ASN_WTH_MBX, which is used by
(almost?) all the mentioned utilities, the problem would gracefully disappear.

					- Gilbert

For what it's worth, I ran into this problem after trying:

	$ SPAWN/NOWAIT SET HOST node

Which is a really neat hack that can be used like spawn/attach on a single node.
It should should suffice until someone hacks RTPAD to do this more cleanly.
You'll probably want to define some DCL symbols as "WAIT 00:02:00".
50.3GLIVET::DIAMONDThu Aug 23 1984 19:5911
When I try to remove the mailbox from the terminal, I also set brdcstmbx and
broadcast flags back the way they should be, off and on respectively.  What I
don't understand is how to deassign the set.  I'm just so confused...
                                                                     .
                                                                      .
                                                                        .
     
                                                                           .
                                                                              .

Dave
50.4HARE::STANThu Aug 23 1984 21:132
What has this got to do with hacking? It sounds like a serious
problem.  Use the VMSNOTES note file if you want a serious answer.
50.5GLIVET::DIAMONDThu Aug 23 1984 23:514
I've got queries all over the place...  What I'm looking for is not a certain
place to find certain answers, but certain people who might have my answers.

Dave
50.6GLIVET::DIAMONDFri Aug 24 1984 00:0210
Aha, he said!!  What I failed to remember is that when you assign or deassign 
anything, the device is given a special something to keep for the rest of its 
life.  And that something is a channel number.  What idiot moi failed to 
remember is that the deassign is based also on the channel number, so I have 
to store the channel number used for the assign and use the same one for the 
deassign.  I can't believe it took me this long to figure it out after losing 
so many hours of sleep, and thanks to those who listened to me babble about 
it.

Dave
50.7VANNOY::VANNOYTue Sep 04 1984 20:446
RE: .2 

VMS has a one mailbox per terminal restriction.  We thought we might
fix this in V4, but never got around to it.  Maybe next time around...

jake