>>>I brought the data to my system and failed the same way the user failed.
>>>I reindexed and packed and all was OK on my system.
>>>Hmmm
>>>After several futile retries at the client, I sent my corrected .dbf and .cdx to the client and all is OK.
>
>SMB or other SNAFU at the client?
>
>You might consider changing the reindex to something that creates a matching shadow dbf, appends from the target then indexes the new, deletes the old and renames to end up with a brand new table. Or, if in a dbc then create the shadow copy, ZAP the source and append from the shadow. If there's a sleeper snafu in the table, this will pick it up...

No SMB.
This is a terminal server application. The server environment has not changed, but there was a major virus and malware attack at the TS client site.
All equipment there was compromised and had to be replaced.
AFAIK the server was not affected, but this points otherwise, doesn't it?