The reason I volunteered my Macs, is because I KNOW the new one has never had any kind of music software…the other, none has been used since upgrading the OS to 10.6–but, I checked, and Logic9/Cubase4 are still installed.
Not meant to sound harsh…at all…understand that I have a unique perspective of actually being an IT guy by trade on a large scale for many years–and actually, if Steiny were the MUCH larger company I worked for–I would be the one assigned to this…if they care. While managing data and network security day in and out, I was the one the IT management would come to when the client group and server group and any third party vendors couldn’t track down the root cause of an issue. I realize, in retrospect, you not knowing me from Adam…my posts could be read very wrong…anyway, suffice to say, root cause analysis in computers is a thing I do (did actually) professionally.
And, the hardest part were the end users’ assumption that they knew what the issue was…or that it was “random”…random isn’t what computers do. There is always an explanation…just a matter of whether it’s within our grasp knowledge wise. Random may happen…but, happening over and over…multiple people…nothing random about that–just not yet making the connection.
Process explorer (on Windows) would be my step if I couldn’t reproduce it on a clean machine. It will tell you what the handles to files or RAM the hung process has open…and related child processes.
Also for the record…not implying people need to install everything on a clean machine. Of course that’s not normal…it’s not the solution. It’s a step in isolating the issue. I could almost bet money a clean machine won’t show the issue–THAT would be caught in any testing…but, here’s what we have…current config that has the issue…and (guess/assumption) clean install of the same OS/app on the hardware that won’t. Which means the answer is something done in between to the system.
It existing on Mac and PC means its not a “bug” in Steiny’s code…as neither the OS nor executable code is shared…the basic functionality would likely be, though…meaning “on exit, close any open processes, issue X disconnect command to ASIO driver, give OS control back to midi ports, terminate code”—obviously horribly oversimplified, but the key would be to be able to see which command is going unanswered by the recipent process…thus causing it to never finish it’s exit code.
Sorry if anything I posted was taken wrong…I ReRead it assuming I was “some guy on the net” and realized it had a whole different vibe if not knowing what I did for a living before being called away by musical pursuits…it was not intended to accuse anyone of breaking their stuff, or let Steiny “off the hook”…just to explain its not likely a simple “code bug”.
If Steiny would like to contact me, I’ll be glad to provide references if they need me to track this down. But, to those whove contacted them–have they asked you to do anything? Given you any kind of debugger to run? Switch to run Cubase with to enable some form of logging? Any XML files they had you send?
