Claudio Valderrama C. wrote:

> >
> > >Roger: the problem is not the optimizer, but the DSQL layer.

I wrote:

> > When the engine parses and compiles the
> > blr that DSQL generates, it does the same sort of thing.

Claudio Valderrama C. wrote:

>The engine didn't detect those ambiguities before I wrote the convoluted
>code in the DSQL layer.

I was wrong. Claudio is absolutely correct that the ambiguity checking
is done in DSQL and that all this ruckus has been about that. The engine
never sees an ambiguous field reference - it can't be expressed in BLR.

Regards,

Ann