| From: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
|---|---|
| To: | Theo Kramer <theo(at)flame(dot)co(dot)za> |
| Cc: | hackers(at)postgreSQL(dot)org |
| Subject: | Re: [HACKERS] Index scan? |
| Date: | 1999-08-13 17:50:43 |
| Message-ID: | 5415.934566643@sss.pgh.pa.us |
| Views: | Whole Thread | Raw Message | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
Theo Kramer <theo(at)flame(dot)co(dot)za> writes:
> Tom Lane wrote:
>> We need to push awareness of the output ordering requirement down into
>> the code that chooses the basic plan. It's on the TODO list (or should
>> be) but I dunno when someone will get around to it.
> I can't wait :-)
I am about to do some major hacking on the planner/optimizer's
representation of path sort orders (for anyone who cares, PathOrder data
is going to be merged into the pathkeys structures). After the dust
settles, I will see what I can do with this issue --- it might be pretty
easy once the data structures are cleaned up.
Aside from the case with an ORDER BY clause, I believe the planner is
currently too dumb to exploit a pre-sorted path for GROUP BY. It
always puts in an explicit sort on the GROUP BY keys ...
regards, tom lane
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Tom Lane | 1999-08-13 17:58:33 | Re: [HACKERS] PROPOSAL: Statement for one-sided joins |
| Previous Message | Tom Lane | 1999-08-13 17:37:19 | Re: [HACKERS] Aborted Transaction During Vacuum |