| From: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
|---|---|
| To: | Jeff Davis <pgsql(at)j-davis(dot)com> |
| Cc: | Greg Stark <gsstark(at)mit(dot)edu>, pgsql-hackers(at)postgresql(dot)org |
| Subject: | Re: pre-proposal: type interfaces |
| Date: | 2009-10-23 21:55:30 |
| Message-ID: | 29044.1256334930@sss.pgh.pa.us |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
Jeff Davis <pgsql(at)j-davis(dot)com> writes:
> On Fri, 2009-10-23 at 16:25 -0400, Tom Lane wrote:
>> Conceivably we could get different left-of
>> operators out of different opclasses, but if they don't behave
>> effectively the same, the user has messed up the opclasses.
> It would probably be worthwhile to make an attempt to throw a useful
> error there, but I agree it's not really a problem.
Sure, right after we solve the halting problem ;-). The point I
was trying to make is that getting different operators is not wrong.
It's only wrong if their behavior isn't consistent with what the
opclass asserts, and there's no practical way to determine that.
We have to trust the opclass maker. (This is one of the main
reasons why CREATE OPERATOR CLASS is superuser-only.)
regards, tom lane
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Josh Berkus | 2009-10-23 23:04:19 | Re: per table random-page-cost? |
| Previous Message | Jeff Davis | 2009-10-23 20:40:44 | Re: pre-proposal: type interfaces |