To: "Jim C. Nasby" <jim@nasby.net>
cc: Josh Berkus <josh@agliodbs.com>, pgsql-hackers@postgresql.org,
	pgsql-performance@postgresql.org
Subject: Re: [HACKERS] Hints proposal 
In-reply-to: <20061013161614.GT28647@nasby.net> 
References: <20061012151439.GT28647@nasby.net> <452E6FFE.7010504@agliodbs.com>
	<20061012192410.GU28647@nasby.net>
	<200610121358.22739.josh@agliodbs.com>
	<20061013161614.GT28647@nasby.net>
Comments: In-reply-to "Jim C. Nasby" <jim@nasby.net>
	message dated "Fri, 13 Oct 2006 11:16:14 -0500"
Date: Fri, 13 Oct 2006 12:36:04 -0400
Message-ID: <25352.1160757364@sss.pgh.pa.us>
From: Tom Lane <tgl@sss.pgh.pa.us>

"Jim C. Nasby" <jim@nasby.net> writes:
> I completely agree that it's much better *in the long run* to improve
> the planner and the statistics system so that we don't need hints. But
> there's been no plan put forward for how to do that, which means we also
> have no idea when some of these problems will be resolved.

You keep arguing on the assumption that the planner is static and
there's no one working on it.  That is false --- although this thread
is certainly wasting a lot of time that could have been used more
productively ;-).

I also dispute your assumption that hints of the style you propose
will be easier to implement or maintain than the sort of
statistical-assumption tweaking that's been counter-proposed.  Just for
starters, how are you going to get those hints through the parser and
rewriter?  That's going to take an entire boatload of very ugly code
that isn't needed at all in a saner design.

			regards, tom lane