Re: Dependency graph of all tuples relied upon in a query answer

From: "Merlin Moncure" <mmoncure(at)gmail(dot)com>
To: "Randall Lucas" <rlucas(at)tercent(dot)com>
Cc: pgsql-general(at)postgresql(dot)org
Subject: Re: Dependency graph of all tuples relied upon in a query answer
Date: 2006-09-05 18:01:58
Message-ID: b42b73150609051101n76d3f6faq7704fd3339452d99@mail.gmail.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-general

On 9/1/06, Randall Lucas <rlucas(at)tercent(dot)com> wrote:
> On Fri, Sep 01, 2006 at 09:30:57AM -0400, Merlin Moncure wrote:
> > A key tenet of relational thinking is to reduce all information to its
> > functional dependencies, and to try and avoid as much as possible
> > keeping information state in the data in a declarative sense.
> > last_query_shown_tuples() is imo a violation in that sense.
>
> I used last_query_shown_tuples() as a quick example. It seems more
> likely that it would be implemented in fact as something like EXPLAIN,
> where it acts upon a given query. Like,
>
> EXPLAIN DEPENDENCIES SELECT * FROM ...
>
> It seems that the query planner *must* know which rows, from which
> tables, actually get used in producing the output for a given query.
> Given that this info is present somewhere in the depths, is there a way
> to get this information out to the app level?

not exactly in the sense you are describing, but you can make the
query such that it gives you the knowledge to get back to the original
data and look it up, yes?

merlin

In response to

Browse pgsql-general by date

  From Date Subject
Next Message Joshua D. Drake 2006-09-05 18:55:22 Re: Porting from ORACLE to PostgSQL
Previous Message Scott Marlowe 2006-09-05 17:33:12 Re: Upgrade Postgres 8.1.2 to 8.1.4