Hello, Mark.

Thanks very much for the fun. I finally got a chance to run the code. Did you notice that it seems that only the index file that is being used exclusively? IOW, this only *works* if the cdx does not already exist.

So, the table in the first data session isn't getting reflagged as exclusively used. ISEXCLUSIVE remains .F.

So, this actually makes a lot of sense. To reindex, the .DBF (literally the file on disk) doesn't need to be locked. But of course the CDX disk file does. So in session 1 when you created the first index, you "created" the CDX and it got created with exclusive use.

Does that make sense?


>Nancy,
>If you're still interested, here is code which will re-create this effect. Notice that it is divided into segments, each to be run on a different session of VFP (2 VFP's loaded at once, on the same machine, defaulted to same path) in a certain sequence.
>
>*- Run this segment on Session #1:
>CREATE TABLE tblIndex FREE (cDummy C(10))
>APPE BLAN
>REPLACE cDummy WITH "Hello"
>USE
>
>SET EXCL OFF
>USE tblIndex && Table is opened SHARED.
>
>*- Now run this segment on Session #2:
>SET EXCL OFF
>USE tblIndex && Table is now opened twice.
>
>*- Now run this segment on Session #1:
>INDEX ON cDummy TAG cDummy && Creates tag even though opened shared & in use.
>
>*- Now run this segment on Session #2:
>BROWSE && This demonstrates that the table is still useable.
>USE
>USE tblIndex && Here's where you finally get an error.
>
>
>Have fun!
>Mark