Abstract: A method for categorizing information regarding a failure in an application program module. The failure may be a crash, a set-up failure or an assert. For a crash, a name of an executable module where the crash occurred in the application program module, a version number of the executable module, a name of a module containing an instruction causing the crash, a version number of the module and an offset into the module with the crashing instruction are determined. This bucket information is then transmitted to a repository for storage in a database. The database may be examined to determine fixes for the bug that caused the crash.

Abstract: A method for categorizing information regarding a failure in an application program module. The failure may be a crash, a set-up failure or an assert. For a crash, a name of an executable module where the crash occurred in the application program module, a version number of the executable module, a name of a module containing an instruction causing the crash, a version number of the module and an offset into the module with the crashing instruction are determined. This bucket information is then transmitted to a repository for storage in a database. The database may be examined to determine fixes for the bug that caused the crash.

Abstract: A method for categorizing information regarding a failure in an application program module. The failure may be a crash, a set-up failure or an assert. For a crash, a name of an executable module where the crash occurred in the application program module, a version number of the executable module, a name of a module containing an instruction causing the crash, a version number of the module and an offset into the module with the crashing instruction are determined. This bucket information is then transmitted to a repository for storage in a database. The database may be examined to determine fixes for the bug that caused the crash.

Abstract: Method and system for recovering information from open files during a failure in a program module. Once a failure is detected in the program module, control passes to an exception handler that determines whether the open files have been modified. If so, the open files are verified and stored as recovery versions of the files at the time of the failure. The program module is then terminated and restarted. Upon restarting the program module, the recovery version of the file at the time of the failure is opened and displayed to the user.

Abstract: Method and system for reporting program failures. The system extracts information about a failure in a program module, such as the location of the failure, and establishes communication with a repository, such as a server. The information about the failure may be used to search the repository to determine whether the particular bug that caused the failure has already been fixed. If so, then the system assists the user in downloading the fix (such as by pointing the user to a particular website for downloading a service release). If a particular failure has not been diagnosed and corrected, then additional relevant information from the user's computer may be uploaded to the repository to provide the developers of the program module enough information about the failure to diagnose the bug and fix it (in a service release or in a later version of the program module).

Abstract: A method and system for injecting an exception into a hung program module to recover unsaved data. The method and system allows a user to determine that the program module is hung and inject an exception into the hung program module. In response to the injected exception, program code, within the program module, is activated to detect the crash and launch a failure reporting executable program that instructs the hung program module to execute its data recovery program. The file recovery program recovers the unsaved data and notifies the failure reporting executable program when it has completed the data recovery. Once the failure reporting executable program receives the notification, it will instruct the operating system to restart the program module. The restarted program module then displays a user interface with a list of the unsaved data that were recovered. The user interface also provides controls for a user to select, from the list of recovered data, the data that the program module will keep.

Abstract: A method and system for preserving document integrity between potentially partially incompatible versions of a program module by maintaining forwards compatibility and backwards compatibility. Version (n') of the program module is released after version (n). A "saved by version (n)" bit may be activated any time that version (n) saves a file. The "saved by version (n)" bit indicates to subsequently-released versions, such as version (n'), that the file has been saved by version (n) and, based upon this information, appropriate action should be taken by the subsequently-released version. A "warning from future version" bit may be activated to indicate that a feature or features of the subsequently-released version is not supported by version (n). Based upon receiving this bit, version (n) may open the file with these features disabled or may take some other appropriate action.