>I think this type of problem is usually associated with users shutting down machines without properly exiting the application, or the application itself terminating due to a crash.
>

in this case thats one of the few causes we can rule out. All users are out of the system by the time the evening cleanup takes place and the relevant index is always rebuilt during that process.

>If you get the problem during the application, it's possible that it's being caused by the network or disk hardware - the index file being corrupted.

yes, thats what Ed Rauh has given me chapter and verse about.
>
>Just some top-of-the-head sugestions.

Thanks

Harry