>> Ofcourse, this is another can of worms. To do this you would have to be
>> able to have the failed query provide hints to the planner telling it
>> where it went wrong. Now, it may be possible to provide (via
>> post-mortem of an execution) a list of actual selectivites like:
>
> Just being able to provide hints to go with EXPLAIN ANALYZE results
> would be useful.
We'd probably get some milage out of collecting statistics equivalent
to EXPLAIN ANALYZE (actual versus estimated), and, over some set of
such statistics, try to improve parameter usage in the optimizer.
That of course requires collecting those stats...
--
output = reverse("moc.liamg" "@" "enworbbc")
http://linuxdatabases.info/info/emacs.html
"...Yet terrible as Unix addiction is, there are worse fates. If Unix
is the heroin of operating systems, then VMS is barbiturate addiction,
the Mac is MDMA, and MS-DOS is sniffing glue. (Windows is filling your
sinuses with lucite and letting it set.) You owe the Oracle a
twelve-step program." --The Usenet Oracle