Our application is written in C++ and used on Windows XP. On some client machines with only a C: drive, an error pop ups when the application starts:
There is no disk in the drive. Please insert a disk into drive "D"
If they hit "Continue" or insert a CD (even an empty one!) and press "Try Again", everything works fine.
Someone suggested that this might be related to compiling on drive D: (our build machine uses drive D: for compilation). Has anyone encountered this issue?
Try compiling it on C: and see if the error still exists. You can use Process Monitor from www.syinternals.com to see which file/dir the program is trying to access. Hope this helps!
Oh yes, we used to have errors like these ... unfortunately i can't tell you what the fix was but it was something utterly stupid, that much i remember. I could still find the comment in our CDebugStackWalk class, so it may have to do with stack unrolling ... somewhere, somehow...
Try running the program through Dependency Walker, and see if there's any reference to D: in the binary's dependencies. This might give you a hint to which library (if any) is being problematic. Or it might give you nothing, but it's free and very quick to do, so I'd recommend it.
It would certainly be a good idea to find out what's trying to access the D drive and fix it. But it's possible to suppress this behaviour, if desired, with a call to SetErrorMode using the SEM_FAILCRITICALERRORS flag. It may even help you to identify the problem, because the error will be sent directly to the application instead of being handled with a system dialog.