Re: Segmentation fault in pg_dumpall from master down to 9.1 and other bug introduced by RLS

* Robert Haas (robertmhaas(at)gmail(dot)com) wrote:
> On Thu, Nov 13, 2014 at 6:32 PM, Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> wrote:
> > What's bothering me is that I see this in pg_dumpall output from a 9.4
> > or earlier database:
> >
> > ALTER ROLE postgres WITH SUPERUSER INHERIT CREATEROLE CREATEDB LOGIN REPLICATION NOBYPASSRLS;
>
> What about leaving out NOBYPASSRLS and letting it go to whatever the default is?

I'd be fine with that- but would we want to do it for the other role
attributes also? Specifically rolcreaterole and rolreplication for
older server versions. I'm still of the opinion that we should just
drop the explicit "true" for all the role attributes for the bootstrap
superuser and then go with this suggestion to let it go to the default
for upgrades from older versions.

Thanks,

Stephen