Hi Calvin.

>if I remember correctly the Repair function in SDT5 did not actually do a pack even thought the documentation said it did.

I think you must mean Reindex(), since until SDT 6.0, the Repair() method didn't accept nor pass a parameter to Reindex() to tell it to pack.

>You might consider doing something to alert them to this potential problem.

I'll add that to my list of things to do for a future release. Since we temporarily open the table using low-level routines at the start of Reindex() anyway, it wouldn't take much more to see if the record count is good or not.

>I hope that in the next VFP service pack or at the latest the next release of VFP something will be done to ensure that the recordcount in the file header is not able to be corrupted or that the DBC will alert us when that happens.

I agree -- this is a huge problem, and doesn't do anything to counter the reputation out there that the DBF file structure (and by extension, VFP) is flaky.

Doug