Re: How to get around LIKE inefficiencies?

On Sun, 5 Nov 2000, Bruce Momjian wrote:

> > Bruce Momjian <pgman@candle.pha.pa.us> writes:
> > > I am adding a new TODO item:
> > >     * Add SET PERFORMANCE_TIPS option to suggest INDEX, VACUUM, VACUUM
> > >       ANALYZE, and CLUSTER
> > > Seems we should be able to emit NOTICE messages suggesting performance
> > > improvements.
> > 
> > This would be targeted to help those who refuse to read documentation?
> > 
> > I'm not following the point here...
> 
> Well, I think there is some confusion about when CLUSTER is good, and I
> can see people turning it on and running their application to look for
> things they forgot, like indexes or vacuum analyze.

so, we are gonna have an AI built into the database now too?  my
experience to date is that each scenario is different for what can be done
to fix something ... as my last problem shows.  I could remove the index,
since it isn't used anywhere else that I'm aware of, or, as philip pointed
out, I could change my query ...

now, this 'PERFORMANCE_TIPS', would it have to be smart enough to think
about Philips solution, or only Tom's?  How is such a knowledge base
maintained?  Who is turned off of PgSQL when they enable that, and it
doesn't help their performance even though they follow the
recommendations?