Re: [HACKERS] Another bug in pg_operator.h

> dg(at)informix(dot)com (David Gould) writes:
> > Also, it lets the planner generate
> > better plans. If there is a negator or commutator it can use it
> > instead of generating extra steps.
>
> Well, I did *not* go looking for links that should be there and weren't
> (except in the very special case that the reverse link existed). I
> just tried to sanity-check the existing links.
>
> I agree that it would be nice to look for missing links that should
> be added ... but that is a performance enhancement, not a bug fix,
> so I am not eager to do it this close to release. We should do
> another pass over this table after 6.4 is out the door.
>
> (Another reason I didn't try to do that is that I've got no good
> idea how to find missing links, short of brain-numbingly tedious
> hand examination. Can you suggest any automated way of finding
> missing links, or at least finding likely things to look at?)

Yes. See include/catalog/template1_check.sql. Uses SQL to do checks.

--
Bruce Momjian | http://www.op.net/~candle
maillist(at)candle(dot)pha(dot)pa(dot)us | (610) 853-3000
+ If your life is a hard drive, | 830 Blythe Avenue
+ Christ can be your backup. | Drexel Hill, Pennsylvania 19026