Date: Thu, 21 Jan 2021 22:46:07 -0600
From: Justin Pryzby <pryzby@telsasoft.com>
To: Tomas Vondra <tomas.vondra@enterprisedb.com>
Cc: Dean Rasheed <dean.a.rasheed@gmail.com>, pgsql-hackers@postgresql.org
Subject: Re: PoC/WIP: Extended statistics on expressions
Message-ID: <20210122044606.GG27167@telsasoft.com>
References: <20210116232208.GB8560@telsasoft.com>
 <2f2a3b3b-93e4-723f-144e-bc21f10d6b61@enterprisedb.com>
 <20210118045331.GF8560@telsasoft.com>
 <CAEZATCVnQccV_Z7AuvbvquaOYxUj+cVFk48pPGASYHFjXWgW5g@mail.gmail.com>
 <3f20da4d-89dc-0912-98d2-b6f0ef06c87f@enterprisedb.com>
 <CAEZATCUuj5FPjtOU2=7Z2pTC8LEvrf79+Go7R8yQ+eE0ezfgfQ@mail.gmail.com>
 <f865af90-c19c-769e-8c65-19692e245479@enterprisedb.com>
 <20210122022958.GC27167@telsasoft.com>
 <acf27932-8159-8872-8fca-902c7a35b351@enterprisedb.com>
 <20210122040101.GF27167@telsasoft.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20210122040101.GF27167@telsasoft.com>
User-Agent: Mutt/1.9.4 (2018-02-28)
Precedence: bulk

On Thu, Jan 21, 2021 at 10:01:01PM -0600, Justin Pryzby wrote:
> On Fri, Jan 22, 2021 at 04:49:51AM +0100, Tomas Vondra wrote:
> > > > | Statistics objects:
> > > > |     "public"."s2" (ndistinct, dependencies, mcv) ON  FROM t
> > 
> > Umm, for me that prints:
> 
> >     "public"."s2" ON ((i + 1)), (((i + 1) + 0)) FROM t
> > 
> > which I think is OK. But maybe there's something else to trigger the
> > problem?
> 
> Oh.  It's because I was using /usr/bin/psql and not ./src/bin/psql.
> I think it's considered ok if old client's \d commands don't work on new
> server, but it's not clear to me if it's ok if they misbehave.  It's almost
> better it made an ERROR.

I think you'll maybe have to do something better - this seems a bit too weird:

| postgres=# CREATE STATISTICS s2 ON (i+1) ,i FROM t;
| postgres=# \d t
| ...
|     "public"."s2" (ndistinct, dependencies, mcv) ON i FROM t

It suggests including additional columns in stxkeys for each expression.
Maybe that also helps give direction to response to Dean's concern?

That doesn't make old psql do anything more desirable, though.
Unless you also added attributes, all you can do is make it say things like
"columns: ctid".

> In any case, why are there so many parentheses ?

-- 
Justin