T.R | Title | User | Personal Name | Date | Lines |
---|
519.1 | | CSC32::BUTTERWORTH | Gun Control is a steady hand. | Wed Dec 14 1994 18:28 | 15 |
| James,
Is your customer trying to send an Eventlist manually or does he
have a filter thats trying to do it?
If it's a filter, then the user-data field for the "Multi-Line Window"
action within the filter should read
LAT_08002B13B7FB 0 0 LAT
i.e., the general format of the userdata field is
nodename server-number screen-number transport
Regs,
Dan
|
519.2 | | WELSWS::GRIFFINP | Either that wallpaper goes, or I do... | Thu Dec 22 1994 11:09 | 17 |
| Dan
I'm the "engineer" referred to in .0. After speaking to James, we decided it
would probably be less confusing if I replied directly.
The eventlist is being started either manually, or automatically when the C3
interface is started (this is one of the startup options).
I haven't tried it via a filter, but I would imagine it would work because
you can specify the node parameter format ie. "LAT_08002B13B7FB". The problem
only seems to happen when the "=_LTAxxx" is appended to the node parameter.
Cheers
Paul Griffin
LAN Services, MCS
|
519.3 | | OPG::PHILIP | And through the square window... | Thu Dec 22 1994 11:33 | 10 |
| Paul,
Can you please install the MUP, I put a fix in for a very similar
problem to this.
If you still have the problem after trying the MUP, we can
investigate further.
Cheers,
Phil
|
519.4 | Still a problem.... | WELSWS::GRIFFINP | Either that wallpaper goes, or I do... | Thu Feb 02 1995 15:05 | 8 |
|
The customer has now installed the MUP (ECO1) and they still have the
problem.
Any suggestions...??
Cheers
Paul
|
519.5 | Whoops! - wrong MUP | WELSWS::GRIFFINP | Either that wallpaper goes, or I do... | Thu Feb 02 1995 15:40 | 6 |
|
I've just realised there is a later MUP - I will get them to install this
as well.
Cheers
Paul
|
519.6 | It's changed slightly...! | WELSWS::GRIFFINP | Either that wallpaper goes, or I do... | Thu Apr 13 1995 18:29 | 20 |
| The customer eventually got the MUP installed and it has got rid of this
problem.
However....
Unfortunately another problem has taken it's place. Whereas before PCM was
complaining about the _LTAxxx suffix for the display node parameter, the message
now is:-
Unable to open the requested display (_WSAxx:).
where xx initially is the number of the original display, plus one. This display
doesn't exist. Each time you try to start the eventlist subsequently, this
number is incremented by one. Removing the _LTAxxx portion of the node
parameter has no effect. I have managed to reproduce the problem in-house, but
unfortunately I haven't had the opportunity to look at it closely.
Any help would be appreciated
Paul
|
519.7 | Any news? | 42204::GRIFFINP | Either that wallpaper goes, or I do... | Fri Apr 28 1995 13:05 | 10 |
| Hi
Just wondering if there is an update to the previous note (I've
been looking around the notesfile and found a lat fix mentioned in
598.3, but this was for VAX and the customer has it running on an
AXP).
Regards
Paul
|
519.8 | | YOSSAM::PHILIP | And through the square window... | Fri Apr 28 1995 13:09 | 9 |
| Paul,
All known/reported problems around displaying to VXT devices have
been fixed in the V1.6 kit. Please try this kit (watch note 2.last
for the announcement) and let us know if your customers problem
persists.
Cheers,
Phil
|
519.9 | Original problem back (but not quite) | WELSWS::GRIFFINP | Either that wallpaper goes, or I do... | Tue May 02 1995 17:00 | 17 |
| Phil
Thanks for your reply. I haven't been able to put it on the customers system
yet but I have upgraded our local system.
Unfortunately there is still a problem - it appears to have reverted to the
original problem (in .0) ie. not liking the "=LTAxxxx" value in the display
node name.
There is a slight difference though - the problem only occurs if you try to
start the eventlist from within the C3 interface (or automatically when C3
starts). If you use the CONSOLE EVENTLIST command directly from DCL it
works OK??
Cheers
Paul
|
519.10 | | YOSSAM::PHILIP | And through the square window... | Tue May 02 1995 17:05 | 9 |
| Paul,
As you can start the eventlist OK from DCL, it could be the mechanism
that the C3 is using to invoke it.
Dave, are you reading this thread, can you shed some light on this one?
Cheers,
Phil
|
519.11 | Console$debug C3 | ZENDIA::DBIGELOW | Innovate, Integrate, Evaporate | Tue May 02 1995 19:27 | 15 |
| Try turning on debug using the following command:
define CONSOLE$DEBUG C3
Start the C3 and then manually start the event list from the pulldown
menu. Look in the window that started teh C3 and you'll see that
command that is used to spawn the event list window. It should look
similar to:
C3: Spawn command = CONSOLE EVENTLIST/NODE=MADWON/SERVER=0/SCREEN=0/
TRANSPORT=TCPIP/C3
Dave
|
519.12 | CONSOLE$DEBUG ouptut | WELSWS::GRIFFINP | Either that wallpaper goes, or I do... | Wed May 10 1995 11:41 | 29 |
| I defined CONSOLE$DEBUG as you suggested - this is the output I got:-
C3: Console Event list spawn command:
C3 create CONSOLE EVENTLIST -
/NODE=LAT_08002B2B45FD=_LTA5045 -
/SERVER=0 -
/SCREEN=0 -
/TRANSPORT=LAT
/C3
%DCL-W-PARMDEL, invalid parameter delimiter - check use of special characters
C3: Spawn command = CONSOLE EVENTLIST/NODE=LAT_08002B2B45FD=_LTA5045/SERVER=0/SC
REEN=0/TRANSPORT=LAT/C3
\=_LTA5045\
C3: Spawn command length = 83
C3: Spawn command status = 1
C3: Spawn command pidadr(7f934d08) = 00000261
C3: Event list pid = 00000261
There was more output than this for the C3 startup but this is specific to
the eventlist startup. If you issue that command manually it doen't like it
either.
Cheers
Paul
|
519.13 | | ZENDIA::DBIGELOW | Innovate, Integrate, Evaporate | Wed May 10 1995 15:55 | 16 |
| Paul,
Well, this redefining of the node name is definately the problem. I've
looked at our wsa_get code which decomposes the decw$display and all
looks to be OK. I'm just curious on how the nodename gets redefined
from LAT_08002B2B45FD to _LTA5045. Is this something special that is
being done on at this site?
I can parse the 'LAT_08002B2B45FD=_LTA5045' string and
remove the '=' which theoretically should fix the problem.
Can we test this fix in-house or is this problem at a customer's site.
Dave
|
519.14 | | CSC32::BUTTERWORTH | Gun Control is a steady hand. | Wed May 10 1995 19:46 | 31 |
| >Paul,
> Well, this redefining of the node name is definately the problem.
>I've looked at our wsa_get code which decomposes the decw$display and
>all looks to be OK. I'm just curious on how the nodename gets redefined
>from LAT_08002B2B45FD to _LTA5045. Is this something special that
>is being done on at this site?
Dave,
I noticed this name change somewhere around Motif 1.1 or so. They
started tacking the LTA device name onto the end of the nodename. I
assume they had some reason for wanting to know the name of the LAT
port created. This is whats screwingyou up. The other option here is to
enclose the LAT_08002B2B45FD=_LTA5045 in quotes. I reproduced this a
while back when we had the original report of this problem. At that
time the eventlist code didn't have a big enough buffer to hold the
string LAT_08002B2B45FD=_LTA5045 when it was encased in quotes. Phil
fixed this by increasing the buffer space but somewhere in there the
quotes got deleted. I remember them being there when I looked at this
originally. I think the easisest thing to do is to have the C3 encase
the nodename string in quotes. Eventlist is coded to handle this.
Regs,
Dan
I can parse the 'LAT_08002B2B45FD=_LTA5045' string and
remove the '=' which theoretically should fix the problem.
Can we test this fix in-house or is this problem at a customer's
site.
Dave
|
519.15 | OK | ZENDIA::DBIGELOW | Innovate, Integrate, Evaporate | Wed May 10 1995 20:20 | 3 |
| Thanks Dan. I'll implement your work-around.
Dave
|
519.16 | MUP for PCM V1.5? | 42730::BAKERP | | Mon Jun 12 1995 17:51 | 13 |
| Hi,
I have just discovered that this is also causing problems at the
customer site where I am working. Up until now all of our work has been
done from a workstation. VXTs are now starting to be used.
Is there any chance of a MUP for PCM v1.5, as I would like to fix this
quickly. I cannot install field test software on what is essentially a
live system.
Regards,
Paul Baker (London S.I)
|
519.17 | | OPG::PHILIP | And through the square window... | Mon Jun 12 1995 17:57 | 9 |
| Paul,
Do you have V1.5 or V1.5a installed? We did some work for VXT problems
in V1.5a that may fix yours. Failing that, you will have to wait for
V1.6 as we dont have the resources (people) to work on several releases
at the same time.
Cheers,
Phil
|