> If the user exists but is disabled by default, I'm not sure that
> really provides any advantage over not having it at all. And it
> certainly can't be enabled by default.
I was presuming that NOLOGIN wouldn't prevent a replication connections
via the replication user. So the user would be "enabled" as far as
replication was concerned.
And currently, there is no replication connection in the pg_hba.conf.
So that's not a change; in fact, having a sample one would be an
improvement.
--
-- Josh Berkus
PostgreSQL Experts Inc.
http://www.pgexperts.com