Received: from sss2.sss.pgh.pa.us (sss.pgh.pa.us [209.114.166.2]) by hub.org (8.9.3/8.9.3) with ESMTP id RAA09848; Fri, 14 Apr 2000 17:31:09 -0400 (EDT) (envelope-from tgl@sss.pgh.pa.us) Received: from sss2.sss.pgh.pa.us (tgl@localhost [127.0.0.1]) by sss2.sss.pgh.pa.us (8.9.3/8.9.3) with ESMTP id RAA04565; Fri, 14 Apr 2000 17:30:15 -0400 (EDT) To: Bruce Momjian cc: PostgreSQL-development , PostgreSQL-documentation Subject: Re: [HACKERS] SESSION_USER In-reply-to: <200004142035.QAA04417@candle.pha.pa.us> References: <200004142035.QAA04417@candle.pha.pa.us> Comments: In-reply-to Bruce Momjian message dated "Fri, 14 Apr 2000 16:35:28 -0400" Date: Fri, 14 Apr 2000 17:30:14 -0400 Message-ID: <4562.955747814@sss.pgh.pa.us> From: Tom Lane X-Archive-Number: 200004/83 Bruce Momjian writes: > create_table.sgml mentions SESSION_USER is not supported. I don't think > that is true anymore, is it? regression=# select session_user; getpgusername --------------- postgres (1 row) Looking at the spec, we do seem to be missing SYSTEM_USER out of the four variants the spec mentions. However, considering that we map them all to the same thing anyway, I can't get too excited about it. If something is done about making functions have setuid-like behavior, we'd want to make sure that the appropriate ones of these functions change value inside a function. regards, tom lane