Re: default to WITHOUT OIDS?

From: Christopher Kings-Lynne <chriskl(at)familyhealth(dot)com(dot)au>
To: Neil Conway <neilc(at)samurai(dot)com>
Cc: PostgreSQL Hackers <pgsql-hackers(at)postgresql(dot)org>
Subject: Re: default to WITHOUT OIDS?
Date: 2003-01-11 02:22:20
Message-ID: 20030111102124.U45491-100000@houston.familyhealth.com.au
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

The real question is how tables are dumped from 7.3 and below. Does it
always explicitly specify 'WITH OIDS' for tables with OIDs?

If not, this would have little benefit for me I guess...

I still vote for the ability to drop OIDs from a table :)

Chris

On 10 Jan 2003, Neil Conway wrote:

> Folks,
>
> Is it a good idea for CREATE TABLE to default to WITHOUT OIDS, rather
> than WITH OIDS? Naturally, this would (a) be some time in the future
> (7.5, perhaps) and (b) only apply to user tables.
>
> The two advantages I can see are:
>
> (1) Makes the storage of most tables more efficient; while you *can*
> specify WITHOUT OIDS, a lot of people won't know to, or will forget.
>
> (2) Reduces the usage of the OID counter, making OID wraparound less
> common.
>
> The main disadvantage I can see is just backward compatibility. In order
> to improve that, we could add a GUC var "use_oids_default" (or some
> other name), which would control whether a CREATE TABLE defaults to WITH
> or WITHOUT OIDS. In 7.4, the GUC var would default to false (so there
> would be no change in behavior), but in 7.5 we could switch it to true.
>
> Comments?
>
> Cheers,
>
> Neil
> --
> Neil Conway <neilc(at)samurai(dot)com> || PGP Key ID: DB3C29FC
>
>
>
>
> ---------------------------(end of broadcast)---------------------------
> TIP 4: Don't 'kill -9' the postmaster
>

In response to

Browse pgsql-hackers by date

  From Date Subject
Next Message Christopher Kings-Lynne 2003-01-11 02:27:43 Re: default to WITHOUT OIDS?
Previous Message Tom Lane 2003-01-11 02:13:26 Re: help with PL/PgSQL bug