Received: from malur.postgresql.org ([217.196.149.56]) by arkaria.postgresql.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_CBC_SHA1:256) (Exim 4.89) (envelope-from ) id 1j1y0t-0005gw-4r for pgsql-bugs@arkaria.postgresql.org; Wed, 12 Feb 2020 19:49:35 +0000 Received: from localhost ([127.0.0.1] helo=malur.postgresql.org) by malur.postgresql.org with esmtp (Exim 4.89) (envelope-from ) id 1j1y0r-0008FP-D6 for pgsql-bugs@arkaria.postgresql.org; Wed, 12 Feb 2020 19:49:33 +0000 Received: from magus.postgresql.org ([2a02:c0:301:0:ffff::29]) by malur.postgresql.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_CBC_SHA1:256) (Exim 4.89) (envelope-from ) id 1j1y0r-0008FI-4z for pgsql-bugs@lists.postgresql.org; Wed, 12 Feb 2020 19:49:33 +0000 Received: from sss.pgh.pa.us ([66.207.139.130]) by magus.postgresql.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.92) (envelope-from ) id 1j1y0o-0004br-Ui for pgsql-bugs@lists.postgresql.org; Wed, 12 Feb 2020 19:49:32 +0000 Received: from sss1.sss.pgh.pa.us (localhost [127.0.0.1]) by sss.pgh.pa.us (8.14.4/8.14.4) with ESMTP id 01CJnRQv010871; Wed, 12 Feb 2020 14:49:28 -0500 From: Tom Lane To: Vera Gangeskar Johne cc: pgsql-bugs@lists.postgresql.org Subject: Re: pg_dump is not backwards compatible in PG12 In-reply-to: References: Comments: In-reply-to Vera Gangeskar Johne message dated "Wed, 12 Feb 2020 19:59:56 +0100" MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-ID: <10869.1581536967.1@sss.pgh.pa.us> Date: Wed, 12 Feb 2020 14:49:27 -0500 Message-ID: <10870.1581536967@sss.pgh.pa.us> List-Id: List-Help: List-Subscribe: List-Post: List-Owner: List-Archive: Precedence: bulk Vera Gangeskar Johne writes: > PG 12 bumped the pg_backup_archiver version to 1.14, however this version > is not supported in past versions of pg_backup_archiver, resulting in > restores to earlier versions not working. Yup. Use a newer version of pg_restore. > 1. Is it expected that pg_dump will produce backwards incompatible dumps? It's happened before, and it will happen again, no doubt. > 2. The release notes don't mention this at all as far as I could see. > Should they? Possibly. In practice, backwards-incompatibility of the SQL contained in the dump is a bigger issue most of the time than the dump file format proper, and there are issues of that sort in almost every major release. > 3. Should this be a bug? No. It's impractical to freeze that file format, because we keep adding new things that pg_dump has to deal with (as the version history you found shows). Maybe it could've been designed to be more extension-friendly, but such a redesign would itself be backwards-incompatible. regards, tom lane