>>If I write "BUG" this is a prove sufficient for me.
>
>If you're doing this for yourself only, why are you bothering to post it?

Because you are alone that you say that this is not a bug,
I always have trust that somebody else understands that this is a bug,
because the names of the fields of the result are not deterministic.

Enough that someone looks at the source C++ of VFP and the bug it will be evident.

I make you an example:
before VFP8 the SELECT had not a restrictive GROUP BY,
this could be useful in some rare case, but it could return
of the non deterministic results.
If in VFP7 I had told yourself that it was a bug, you would certainly have told me
what a feature that existed from FP0.0 was.
It is sure what in VFP8 someone wakes up and he has understood that it was better
to systematize the matter.

I know that VFP is dead, and SEDNA is only a way to move his soul in NET.,
but if it doesn't require me more than a few minutes I write the same so that someone,
can avoid some problems.