>Thanks, Sergey.
>
>I was conjuring up queries that in my mind were complex and nested or whatever and I didn't understand the rules about the immediate parent correlation. I don't have that particular query since it was experimental and I lost it. In VFP9 SP0 the query returned a cartesian set and VFP9 SP1 produced the error. And the error is correct based on the docs.
>
>Thanks for putting me on the right track, though.

You're welcome.