To: "Viorel Dragomir" <bc@vio.ro>
Cc: "Greg Stark" <gsstark@mit.edu>, <pgsql-sql@postgresql.org>
Subject: Re: summing tables
References: <20030715140921.7c5a6a6c.erik@thiele-hydraulik.de>
	<001501c34aca$e70ca280$0600a8c0@fix.ro>
	<20030715143925.54716ed4.erik@thiele-hydraulik.de>
	<007101c34ad6$3e0249e0$0600a8c0@fix.ro>
	<878yqzn8k5.fsf@stark.dyndns.tv>
	<010201c34ae2$b35b3830$0600a8c0@fix.ro>
In-Reply-To: <010201c34ae2$b35b3830$0600a8c0@fix.ro>
From: Greg Stark <gsstark@mit.edu>
Organization: The Emacs Conspiracy; member since 1992
Date: 15 Jul 2003 11:38:21 -0400
Message-ID: <87u19nls0i.fsf@stark.dyndns.tv>
Lines: 20
User-Agent: Gnus/5.09 (Gnus v5.9.0) Emacs/21.3
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii


"Viorel Dragomir" <bc@vio.ro> writes:

> Anyway, in real life this update modifies only one row with a value wich is
> diff of null. It was really handy if it was specified the option ORDER for
> the update command.

Are you hoping to produce a running total? That's very difficult in standard
SQL. That would be very different from the query you asked for. 

Running totals, ranking, lead/lag, are all things that are very difficult to
do in standard SQL. They don't fit in the unordered set model that SQL follows
so doing them without special non-standard functions is very hard and
inefficient. 

The functions to do them don't fit well within the SQL universe either, which
might be why they don't exist yet in postgres.

-- 
greg