[Bug 70028] sound kmail notification crash

[Waldo Bastian]

------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.
      
http://bugs.kde.org/show_bug.cgi?id=70028      




------- Additional Comments From bastian kde org  2004-03-23 00:16 -------
Ok, I now sort of see what is happening, but I'm not sure I understand:

Glossary:
APO: Arts::audiofilePlayObject
SAP: Arts::Synth_AMAN_PLAY
SMA: Arts::Synth_MULTI_ADD
SVC: Arts::StereoVolumeControl

#nnn is reference to line number nnn of attachment 5284

All references are to left channel only.

#417 virtualize SBU to SMA.invalue
#418 SBU connects to SMA.invalue
#451 connect port outleft to left
#452 SVC.out connects to SAP
#458 +--> resulting in SVC.out to SMA.invalue connection
#561  +--> resulting in SVC.out to SMA._invalue1 connection
#507 connect port left to left
#508 APO connects to SAP
#514 +--> resulting in APO to SMA.invalue connection
#517  +--> resulting in APO to SMA._invalue2 connection

#571 devirtualize SBU to SMA
.... +--> resulting in a lot of reshuffling
#630  +--> resulting in SVC.out to SBU connection
#636  +--> resulting in APO to SBU connection
#713 disconnect port left and left
#714 APO disconnects from SAP
#730 +--> the connection from #636 is now changed into a
          SVC.out to SBU connection, but that fails because
          such connection was already created in #630
          (this condition was caught in bug-comment #20)

The overall problem seems to be here that although both SVC.out and APO get connected to SMA.in, this results in two distinct connections to SMA._invalue1 and SMA._invalue2 and when #517 devirtualizes SBU -> SMA they are no longer recognized as being distinct.

I'm inclined to think that the SBU->SMA devirtualisation should happen later, after the APO->SAP and SVC->SAP disconnect.