Arrgh, I couldn't resist to check this control.stop() method again. There I found that actually TWO instances of the Control class might exist although a display just requested ONE. In the case of the ITime control, I see that the timeout of the control gets twice bound to gobject.timeout_add. Then, one of the controls seems to shutdown properly while the other still tries to do its job, i.e. updating the display widget that doesnt exist anymore.
I still dont know exactly what happens when a control is requested (scripting/Script, method __script_get_control() ) by the sandbox. Maybe the ControlWrapper does something strange while not creating a single instance but two at the same time?
Arrgh, I couldn't resist to check this control.stop() method again. There I found that actually TWO instances of the Control class might exist although a display just requested ONE. In the case of the ITime control, I see that the timeout of the control gets twice bound to gobject. timeout_ add. Then, one of the controls seems to shutdown properly while the other still tries to do its job, i.e. updating the display widget that doesnt exist anymore. get_control( ) ) by the sandbox. Maybe the ControlWrapper does something strange while not creating a single instance but two at the same time?
I still dont know exactly what happens when a control is requested (scripting/Script, method __script_