Hello.
On Tue, Sep 8, 2009 at 11:40 PM, Tom Lane<tgl@sss.pgh.pa.us> wrote:
> What will happen (at least conceptually) is that for each row
> of fdt, the whole sub-SELECT will be evaluated using that row's value
> of fdt.c1, producing a list of t2.c1 values. Then, if the fdt row's
> value of c1 appears in that list, the fdt row will be included in the
> result.
Thanks for the detailed explanation - now I see where my mistake was!
--
Sergey Samokhin