>>>Thanks Edward,
>>>
>>>The problem is that the application is dealing with multiple tables having several
>>>hundred thousands of records so systematically reindexing is not a valid
>>>option for me.
>>>
>>>Ever seen this kind of corruption before?
>>>
>>>Pascal
>>Ed, sorry for butting in. Pascal, CDX corruptions might be seen in variety of dummy errors, but are you sure it's a corruption and not your code ? Could you give exact portion of your code checking it ?
>>Cetin
>
>Cetin, I really saw some amazing VFP behaviour due to index corruption, however the reason of corruption itself is network/server disruption. Basically, you cannot corrupt CDX by programmatic means :).
Ed,
Sorry for misunderstanding. I wanted to mean it might not be a corruption at all. Sometimes faulty code make programmer think there is a corruption and wondered if Pascal was coding wrong.
Cetin