Rod Taylor <rbt(at)sitesell(dot)com> writes:
> It seems that with more than a few thousand structures in pg_class, tab
> completion seems to take quite a long time in the en_US locale.
> Simply switching the where clauses around gets a substantial performance
Hm. There's no guarantee that the planner will evaluate the WHERE
clauses in the same order you write them, but it does seem to work that
way for these particular queries at the moment, so we may as well write
the clauses in the best order. You're right that the foo_is_visible()
family of functions are pretty slow...
regards, tom lane
In response to
pgsql-hackers by date
|Next:||From: Bruce Momjian||Date: 2004-12-24 15:46:27|
|Subject: Re: RC2 and open issues|
|Previous:||From: Alvaro Herrera||Date: 2004-12-23 21:58:37|
|Subject: Re: CreateFunctionStmt struct|