>>I notice there are several tools out there to fix corrupted tables. Anybody have any suggestions as to which one works best in VFP 5.0? Do any of them fix a corrupted database in VFP 5.0? Are any of them written in code that we can run from within our application, allowing us to automatically fix the tables without any user interaction? Any help would be appreciated. Thanks
>
>Hi Gene,
>
>About two weeks ago I had the same thought when a user from a distant state reported me a typical error that is consequence of a table corruption. Then I imagined the system being capable of detecting those errors and fixing them automatically. I began to have a look over some tools (specially those available in the files section) that deal with these things.
>
>But, as I understood, when there's a table corruption usually you've got also some damaged records. So it's not only a matter of fixing the table header - that only makes the table USEable again. And what about the damaged records? Someone should at least open and browse the table to have a look at it... And what about referential integrity? Things seem to be more complex, don't you think?
>
>However, one day we will surely reach the point of developing "self-curable" systems...
>
>Anyway I would be glad to know your conclusions and also if you find out something interesting on this issue...
>
>Regards,
>
>Renato

Hello Renato,

You have many interesting points that I had thought of also. In my experience though I have found that at least 95% of the time the only problem is the table header. And the other 5% there are only a few records corrupted so it usually takes a user looking up something to actually find the corrupted records. And when you do find them you usually have to have the user reenter the data, or go to a backup to retrieve the records. So, what I thought I would do would be to fix the problem automatically but also have some way to notify the system administrator that the fix had occurred. At least this way you are not locking the users out of the system. Its not a perfect solution but I think it will work. Guess we will find out after I implement it <grin>. Thanks for your comments,

Genes