From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 12017 invoked by alias); 7 Aug 2003 18:48:27 -0000 Mailing-List: contact gcc-bugs-help@gcc.gnu.org; run by ezmlm Precedence: bulk List-Archive: List-Post: List-Help: Sender: gcc-bugs-owner@gcc.gnu.org Received: (qmail 12010 invoked by uid 48); 7 Aug 2003 18:48:27 -0000 Date: Thu, 07 Aug 2003 18:48:00 -0000 Message-ID: <20030807184827.12008.qmail@sources.redhat.com> From: "bangerth at dealii dot org" To: gcc-bugs@gcc.gnu.org In-Reply-To: <20030807183306.11847.James.W.Mckelvey@jpl.nasa.gov> References: <20030807183306.11847.James.W.Mckelvey@jpl.nasa.gov> Reply-To: gcc-bugzilla@gcc.gnu.org Subject: [Bug c++/11847] [3.4 regression] reference does not work as template argument any more X-Bugzilla-Reason: CC X-SW-Source: 2003-08/txt/msg01062.txt.bz2 List-Id: PLEASE REPLY TO gcc-bugzilla@gcc.gnu.org ONLY, *NOT* gcc-bugs@gcc.gnu.org. http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11847 ------- Additional Comments From bangerth at dealii dot org 2003-08-07 18:48 ------- Just a further clue that this might have something to do with two-stage name lookup: if I make the variable template dependent, then the code compiles again. Here's an example: --------------------------- template struct A { static const T a; }; template class X {}; template struct Y { X::a> x; }; template struct Y; ----------------------------------- W.