Re: pgsql: In AlterRole, make bypassrls an int

On Friday, August 21, 2015, Alvaro Herrera <alvherre(at)2ndquadrant(dot)com> wrote:

> Stephen Frost wrote:
>
> > Add regression tests to test CREATE/ALTER role for the various role
> > attributes. Arrange to leave roles behind for testing pg_dumpall, but
> > none which have the LOGIN attribute.
>
> There is a problem in this test, which is that it throws errors if the
> roles already exist. Maybe CREATE IF NOT EXISTS is needed, or perhaps
> a DROP beforehand?
>

Right, not sure it's that simple since we don't want the output to differ
based on if they exist already or not. Might be able to adjust the log
level to address that, or a DO block which handles it if nothing else
works.

Thanks!

Stephen