No, thats clearly not desirable. For some reason, the code in the UI
looking at the response from the exec
to see if the execution has been queued was modified to accept "rror" as
well, so that error messages in the startup
process are recognized as indicating that the exec was queued, which it
wasn't. Who knows why. Anyway, clearly
wrong, so I've removed it.
As to whether the queuing (how do you spell that?) of the execution is
deswirable, that was originally put in there way back in the
PVS days so that multiple UI's could share a single server by politely
waiting for one another to be done. Shouldn't
ever be happening now. I can't find the code that emits the message
("Execution has been queued") that the UI looks for, but I left
the code to recognize it in the UI just in case.
Greg
Randall Hopper <[EMAIL PROTECTED]>@opendx.watson.ibm.com on 05/21/2001
09:12:20 AM
Please respond to [email protected]
Sent by: [EMAIL PROTECTED]
To: [email protected]
cc:
Subject: [opendx-dev] "Connection to server <host> has been queued"
In recent CVS versions of DX, I've noticed that this dxui startup
message is emitted when the spawned dxexec child/children fail (usually
because another dxui/dxexec is running, or dxexec can't allocate the
requested amount of memory (e.g. >= 1526MB on our Onyxen; need a 64-bit
version for larger arenas I suppose) for some other reason.
After this message, dxui starts up in a disconnected state
(Connection->Start Server is enabled).
What does this message mean? Is this the desired behavior?
To me the message implies that the exec might be started when the
other is shut down, but it doesn't. Are two error situations possibly
being reported with a message describing one?
Randall
--
Randall Hopper (mailto:[EMAIL PROTECTED])
Lockheed Martin Operation Support
EPA Scientific Visualization Center
US EPA MD/24 ERC-1A; RTP, NC 27711
