"Dave Page" <dpage@vale-housing.co.uk> writes:
>> Another point is that Dave added code to pg_dumpall to not dump the
>> postgres database.
> My reading of that code was that I merely stopped it dumping the CREATE
> DATABASE statement (and the ACL) for the database, /not/ the actual
> contents - in the same way as is done for template1. The theory being
> that if you are reloading from into a freshing initdb'ed cluster,
> postgres will already exist so doesn't need to be recreated.
D'oh ... you're right of course. Will fix (and add some comments).
I wonder though if this code isn't a little broken. It should skip the
CREATE DATABASE certainly, but what about the ACL and dumpDatabaseConfig
parts? I suspect those got added in at a handy place without enough
thought taken as to whether they should be excluded for template1.
regards, tom lane