Re: quick question abt pg_dump and restore

On Wed, 09 Jan 2008 10:54:21 -0500
Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> wrote:

> Alvaro Herrera <alvherre(at)commandprompt(dot)com> writes:
> > Josh Harrison escribió:
> >> Fine. I can use order by when I want to order it in terms of
> >> some columns. But What if I want to maintain the same order as
> >> in the database1? ie., I want my rows of TableABC in Database2
> >> to be the same order as the rows in TableABC in Database 1 ???
>
> > You can't.
>
> According to the SQL standard, a table is an *unordered* collection
> of rows, and the results of any query are produced in an
> unspecified order (unless you use ORDER BY). The ambiguity about
> row ordering is intentional and is exploited by most DBMSes
> including Postgres to improve implementation efficiency. If you
> assume there is such a thing as a specific ordering within a table,
> you'll live to regret it eventually.

Does it make any sense *knowing* how the implementation works to load
records in a table in a specific order to improve performances?

And yeah I know that once you start deleting/updating row you may
lose the advantage you gained betting on some peculiarity of the
implementation... but in case you're dealing with a mostly static
table?

eg. if I'm importing a table does it make any sense to pre-sort it
before importing it in postgres?

--
Ivan Sergio Borgonovo
http://www.webthatworks.it