Received: from makus.postgresql.org (makus.postgresql.org [98.129.198.125]) by mail.postgresql.org (Postfix) with ESMTP id 0BCC91C7037D for ; Tue, 13 Dec 2011 20:12:25 -0400 (AST) Received: from mail-ey0-f174.google.com ([209.85.215.174]) by makus.postgresql.org with esmtp (Exim 4.72) (envelope-from ) id 1RacSK-00064x-7e for pgsql-docs@postgresql.org; Wed, 14 Dec 2011 00:12:24 +0000 Received: by eaaj10 with SMTP id j10so234053eaa.19 for ; Tue, 13 Dec 2011 16:12:07 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=gamma; h=mime-version:in-reply-to:references:date:message-id:subject:from:to :cc:content-type:content-transfer-encoding; bh=cJRMIpi8L589Bb2ORmlFkE6/qSy543sDVlxZPNfbkY8=; b=VAFt1wLum/SpPvoy42Knf36gpUqc4hZ1jK0Ay9OXID51hrVe37WLbzP4Msc3guXKlK bnw6cI+32fp0sf9sIQgvd0Nj+cjQcZquq5wZY9HHXxPyV8p5z/wXFpuwnhnUIdMHPvIE nXguPhKFLVfP7jakeAT8Jd8KKZel1Y+XIODUA= MIME-Version: 1.0 Received: by 10.204.199.132 with SMTP id es4mr64950bkb.4.1323821527240; Tue, 13 Dec 2011 16:12:07 -0800 (PST) Received: by 10.204.66.136 with HTTP; Tue, 13 Dec 2011 16:12:07 -0800 (PST) In-Reply-To: <201112132336.pBDNaRB21797@momjian.us> References: <9796.1323487277@sss.pgh.pa.us> <201112132336.pBDNaRB21797@momjian.us> Date: Tue, 13 Dec 2011 17:12:07 -0700 Message-ID: Subject: Re: CREATE USER From: Scott Marlowe To: Bruce Momjian Cc: Tom Lane , Jaime Casanova , pgsql-docs Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: quoted-printable X-Pg-Spam-Score: -2.6 (--) X-Archive-Number: 201112/36 X-Sequence-Number: 7143 On Tue, Dec 13, 2011 at 4:36 PM, Bruce Momjian wrote: > Tom Lane wrote: >> Jaime Casanova writes: >> > I just notice that when you create a user using CREATE USER and give >> > the user the ability to create other users it gets created as >> > SUPERUSER. >> >> Are you talking about the CREATEUSER option? =A0That is documented, quit= e >> clearly I think: >> >> =A0 =A0 =A0 CREATEUSER >> =A0 =A0 =A0 NOCREATEUSER >> >> =A0 =A0 =A0 These clauses are an obsolete, but still accepted, spelling = of >> =A0 =A0 =A0 SUPERUSER and NOSUPERUSER. Note that they are not equivalent= to >> =A0 =A0 =A0 CREATEROLE as one might naively expect! >> >> The only way we could really make this any better is to remove these >> keywords, which might be something to consider. =A0The preferred SUPERUS= ER >> spelling has been accepted since 8.1 ... is that long enough to ensure >> everyone's converted their client-side tools? =A0Probably not :-( > > How does this relate to people restoring pre-8.1 pg_dumpall dumps? > Seems removing this option would cause the restore to fail. True, but 1: that's a really old release, and you should always dump from the version you're going to not, going from, and 2: If you dumped with 8.1 from 8.1 you're likely to have plenty of other issues loading into later versions as well. How about it's marked as deprecated to be removed in the next major release (9.1) and removed the one after that (9.2)?