Teodor Sigaev <teodor (AT) sigaev (DOT) ru> writes:
Mmm, sweet :-(. There is only one legal way to form the outer join, but
make_rels_by_joins() doesn't try it because have_relevant_joinclause()
says there is no relevant joinclause ... as indeed there is not, the
"true = true" thing having been optimized away. I guess we need a hack
to treat empty outer join conditions specially.

Right, not a problem before 8.2 because outer join order was driven by
the syntax instead of by searching for a good join order. Also, you
need at least two base relations on each side of the outer join, else
the "last ditch" case in make_rels_by_joins() finds the join.

regards, tom lane

---------------------------(end of broadcast)---------------------------
TIP 3: Have you checked our extensive FAQ?

http://www.postgresql.org/docs/faq

I wrote:
Actually, it can happen with non-empty join conditions too, if the join
condition doesn't mention the outer side; for instance, using your
example

explain SELECT * from fooview fv1
LEFT OUTER JOIN fooview fv2 on fv2.a_for_max_b < 10;

So my original thoughts of a narrow special case for "OUTER JOIN ON TRUE"
went up in smoke, and I ended up just having have_relevant_joinclause()
troll for relevant outer joins all the time. This probably isn't going
to cost enough planning time to matter, anyway.

regards, tom lane

---------------------------(end of broadcast)---------------------------
TIP 2: Don't 'kill -9' the postmaster

Thank you a lot, I was near around it

--
Teodor Sigaev E-mail: teodor (AT) sigaev (DOT) ru
WWW: http://www.sigaev.ru/

---------------------------(end of broadcast)---------------------------
TIP 7: You can help support the PostgreSQL project by donating at

http://www.postgresql.org/about/donate