Horton

Re: pg_dump: sortDumpableObjectsByTypeName() doesn't always do that

Jacob Champion <pchampion@pivotal.io>

View thread View on postgr.es
From: Jacob Champion <pchampion@pivotal.io>
To: Tom Lane <tgl@sss.pgh.pa.us>
Cc: PostgreSQL mailing lists <pgsql-hackers@postgresql.org>
Date: 2018-08-06T19:34:47Z
Lists: pgsql-hackers 
On Mon, Aug 6, 2018 at 12:23 PM Jacob Champion <pchampion@pivotal.io> wrote:
> since the
> root cause is that we're not defining a valid ordering, quicksort may
> or may not behave consistently for test purposes.

To expand on this, consider three objects, of the same type, compared
with the current comparator function:

namespace_a.object_3 < namespace_b.object_1 (because a < b)
NULL.object_2 < namespace_a.object_3 (because 2 < 3)
namespace_b.object_1 < NULL.object_2 (because 1 < 2)

This is rock-paper-scissors instead of a partial ordering.

--Jacob

Commits

  1. Ensure pg_dump_sort.c sorts null vs non-null namespace consistently.

  2. Last-minute updates for release notes.