On Fri, Jul 18, 2025 at 3:17 PM Noah Misch wrote:
> > This comment is useful, but if I were to be critical, it does a better
> > job saying what this field isn't than what it is.
>
> True. I've changed it to this:
That looks great.
> - /* To have a stable sort order, break ties for some o
On Thu, Jul 17, 2025 at 09:24:02AM -0400, Robert Haas wrote:
> On Mon, Jul 7, 2025 at 3:27 PM Noah Misch wrote:
> > Let's get rid of pg_dump's need to sort by OID, apart from catalog
> > corruption
> > scenarios.
>
> +1. I had at one point believed that sorting by OID was a good way to
> make du
On Mon, Jul 7, 2025 at 3:27 PM Noah Misch wrote:
> Let's get rid of pg_dump's need to sort by OID, apart from catalog corruption
> scenarios.
+1. I had at one point believed that sorting by OID was a good way to
make dumps stable, but this disproves that theory. Sorting by logical
properties of t
