I'm glad you where able to fix the problem.

>Thanks Sergey, in fact I was hoping you would notice my post and you did.
>
>>The second byte of the table header is the year of the last update. Both 02 and 03 are valid table types

0x02   FoxBASE
>>0x03   FoxBASE+/Dbase III plus, no memo

Switching 1 and 2 bytes in this case may not make table header invalid but my change table type. You'll have to come up with another way to protect your tables.
>
>It did not happen last year but this year created hell.
>
>>Pack creates new temp. table, copies record there, dletes original table and renames temp. table back to the name of original table. It doesn't correct header, it creates new correct one.
>
>This explains my observation as at a clients place I configured my app to keep dbfs unlocked as well as pack on close and the problem went away.
>
>Thanks, you may not be able to guess how problematic this was turning out to be, I had visions that I might have to leave this field of work <g> and your help is really appreciated. And because it was happening with my main app, more users were affected.
>