From: "Tom Lane" <[EMAIL PROTECTED]>
> Yeah, this is a fairly obvious thing to want to do with a user-written
> aggregate. It does not work in released versions, because the planner
> does not notice that the inner SELECT's output ordering matches what
> the GROUP BY needs, and so it inserts an a
"Julian Scarfe" <[EMAIL PROTECTED]> writes:
> OK, I know relying on ordering in an aggregate is sinful, but I don't know
> if it's mortal or venial.
> ...
> SELECT c.fir_ident, c.fir_indicator, create_path (c.node) AS fir_edge
>INTO fir_e
>FROM
>(SELECT fir_ident, fir_indica
OK, I know relying on ordering in an aggregate is sinful, but I don't know
if it's mortal or venial.
Long explanation, please bear with me, here's the background:
---
CREATE TABLE "foo" (
"a" point,
"b" int
);
INSERT INTO ...
SELECT * FROM foo;
a| b
-+---
(1,1)
