On Fri, 7 Sep 2001 21:20:16 +0000 (UTC), alvherre@atentus.com (Alvaro
Herrera) wrote:
[snip]
>> Au contraire: OID wraparound would cause us to violate the spec.
>> If we want to follow the spec here, then I think we'd need to add
>> a creation-timestamp column to pg_trigger, and sort on that.
>
>If you are going to create a separate column for sorting, why not just
>use a numeric (int8?) value, so one can change that as sees fit?
[snip]
Seconded..
But then again, whats to stop DBA just changing the value in the date
column?