Re: using index or check in ALTER TABLE SET NOT NULL

On Mon, 11 Mar 2019 at 00:13, Sergei Kornilov <sk@zsrv.org> wrote:
>
> > Providing I'm imagining it correctly, I do think the patch is slightly
> > cleaner with that change.
>
> Ok, sounds reasonable. I changed patch this way.

Looks good to me.  Good idea to keep the controversial setting of
client_min_messages to debug1 in the tests in a separate patch.

Apart from a few lines that need to be wrapped at 80 chars and some
comment lines that seem to have been wrapped early, I'm happy for it
to be marked as ready for committer, but I'll defer to Ildar in case
he wants to have another look.

-- 
 David Rowley                   http://www.2ndQuadrant.com/
 PostgreSQL Development, 24x7 Support, Training & Services