From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 17309 invoked by alias); 5 Nov 2004 00:39:26 -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 17295 invoked by uid 48); 5 Nov 2004 00:39:26 -0000 Date: Fri, 05 Nov 2004 00:39:00 -0000 Message-ID: <20041105003926.17294.qmail@sourceware.org> From: "giovannibajo at libero dot it" To: gcc-bugs@gcc.gnu.org In-Reply-To: <20041104224005.18304.boris@kolpackov.net> References: <20041104224005.18304.boris@kolpackov.net> Reply-To: gcc-bugzilla@gcc.gnu.org Subject: [Bug c++/18304] unnecessary typename expected X-Bugzilla-Reason: CC X-SW-Source: 2004-11/txt/msg00500.txt.bz2 List-Id: ------- Additional Comments From giovannibajo at libero dot it 2004-11-05 00:39 ------- Anyway, it is invalid because a nested class (like "t") is always dependent, as you can specialize it. think of what happens if you define this later: template <> struct s::t { }; (now, in this very case, it would be invalid because you cannot specialize this nested class after having already used the primary template s::t, but the point is to show that s::t is always a dependent context, even within s). -- What |Removed |Added ---------------------------------------------------------------------------- Status|UNCONFIRMED |RESOLVED Resolution| |INVALID http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18304