Re: [HACKERS] Slow count(*) again...

Robert Haas wrote:
> On Thu, Feb 3, 2011 at 6:33 PM, Mladen Gogala <mladen.gogala@vmsinfo.com> wrote:
>
>> Kevin Grittner wrote:
>>
>>> Mladen Gogala <mladen.gogala@vmsinfo.com> wrote:
>>>
>>>> Maybe we can agree to remove that ridiculous "we don't want hints"
>>>> note from Postgresql wiki?
>>>>
>>>>
>>>  I'd be against that.  This is rehashed less frequently since that
>>> went in.  Less wasted time and bandwidth with it there.
>>>
>> Well, the problem will not go away.  As I've said before, all other
>> databases have that feature and none of the reasons listed here convinced me
>> that everybody else has a crappy optimizer.  The problem may go away
>> altogether if people stop using PostgreSQL.
>>
>
> You seem to be asserting that without hints, problem queries can't be
> fixed.  But you haven't offered any evidence for that proposition, and
> it doesn't match my experience, or the experience of other people on
> this list who have been using PostgreSQL for a very long time.  If you
> want to seriously advance this conversation, you should (1) learn how
> people who use PostgreSQL solve these problems and then (2) if you
> think there are cases where those methods are inadequate, present
> them, and let's have a discussion about it.  People in this community
> DO change their mind about things - but they do so in response to
> *evidence*.  You haven't presented one tangible example of where the
> sort of hints you seem to want would actually help anything, and yet
> you're accusing the people who don't agree with you of being engaged
> in a religious war.  It seems to me that the shoe is on the other
> foot.  Religion is when you believe something first and then look for
> evidence to support it.  Science goes the other direction.
>
>
Actually, it is not unlike a religious dogma, only stating that "hints
are bad". It even says so in the wiki. The arguments are
1) Refusal to implement hints is motivated by distrust toward users,
citing that some people may mess things up.
    Yes, they can, with and without hints.
2) All other databases have them. This is a major feature and if I were
in the MySQL camp, I would use it as an
   argument. Asking me for some "proof" is missing the point. All other
databases have hints precisely because
   they are useful. Assertion that only Postgres is so smart that can
operate without hints doesn't match the
   reality. As a matter of fact, Oracle RDBMS on the same machine will
regularly beat PgSQL in performance.
   That has been my experience so far.   I even posted counting query
results.
3) Hints are "make it or break it" feature. They're absolutely needed in
the fire extinguishing situations.

I see no arguments to say otherwise and until that ridiculous "we don't
want hints" dogma is on wiki, this is precisely what it is:  a dogma.
Dogmas do not change and I am sorry that you don't see it that way.
However, this discussion
did convince me that I need to take another look at MySQL and tone down
my engagement with PostgreSQL community. This is my last post on the
subject because posts are becoming increasingly personal. This level of
irritation is also
characteristic of a religious community chastising a sinner. Let me
remind you again: all other major databases have that possibility:
Oracle, MySQL, DB2, SQL Server and Informix. Requiring burden of proof
about hints is equivalent to saying that all these databases are
developed by idiots and have a crappy optimizer.
I am not going to back down, but I may stop using Postgres altogether.
If that was your goal, you almost achieved it. Oh yes, and good luck
with the world domination. If there is not enough common sense even to
take down  that stupid dogma on the wiki, there isn't much hope left.
With this post, my participation in this group is finished, for the
foreseeable future.


--

Mladen Gogala
Sr. Oracle DBA
1500 Broadway
New York, NY 10036
(212) 329-5251
http://www.vmsinfo.com
The Leader in Integrated Media Intelligence Solutions