Created attachment 177336 [details] [review]
patch to fix the issue

Under the following circumstances banshee will crash with an X error:

- KDE 4
and
- x86_64
and
- "Notification Area Icon" extension enabled

How to reproduce:
- start banshee

Backtrace:

Gdk-ERROR **: The program 'Banshee' received an X Window System error.
This probably reflects a bug in the program.
The error was 'BadMatch (invalid parameter attributes)'.
  (Details: serial 3066 error_code 8 request_code 2 minor_code 0)
  (Note to programmers: normally, X errors are reported asynchronously;
   that is, you will receive the error a while after causing it.
   To debug your program, run it with the --sync command line
   option to change this behavior. You can then get a meaningful
   backtrace from your debugger if you break on the gdk_x_error() function.)
aborting...
Stacktrace:

  at (wrapper managed-to-native) X11NotificationArea.gdk_x11_get_server_time (intptr) <0x0005b>
  at (wrapper managed-to-native) X11NotificationArea.gdk_x11_get_server_time (intptr) <0x0005b>
  at X11NotificationArea.SendManagerMessage (X11NotificationArea/SystemTrayMessage,intptr,uint,uint,uint) <0x000af>
  at X11NotificationArea.SendDockRequest () <0x00037>
  at X11NotificationArea.OnRealized () <0x00037>

Notes:
- the issue can be prevented if banshee is started with: 
XLIB_SKIP_ARGB_VISUALS=1 banshee

- similar to https://bugzilla.gnome.org/show_bug.cgi?id=620461 

- Fedora bug report: https://bugzilla.redhat.com/show_bug.cgi?id=664636

Suggested fix:
- it looks like that the check whether the notification area is an rgba visual in src/Extensions/Banshee.NotificationArea/Banshee.NotificationArea/X11NotificationArea.cs: 
--------------------------
       // TODO the proper check is (visual->red_prec + visual->blue_prec + visual->green_prec < visual->depth)
       // but this seems to work
       visual_is_rgba = visual != null && visual == Screen.RgbaVisual;
--------------------------
is not sufficient

- on x86_64 systems running kde the visual of the tray area seems not to be identical with the visual returned via gdk_screen_get_rgba_visual (it seems that there is more than one rgba visual in the system ...)

- I have implemented the algorithm already suggested in the "todo" of the file: don't compare the native GdkVisual objects but actually check for the content of the used visual. Please see the attached patch against git master which applies to stable-1.8 as well. It is only roughly tested, but so far it looks good.

- If there will be another 1.8.x release, please fix it in stable-1.8 and master, otherwise master should be sufficient. Thank you very much!