From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 24503 invoked by alias); 17 Nov 2007 17:57:53 -0000 Received: (qmail 24495 invoked by uid 22791); 17 Nov 2007 17:57:53 -0000 X-Spam-Check-By: sourceware.org Received: from smtp.cs.tamu.edu (HELO postal.cs.tamu.edu) (128.194.138.100) by sourceware.org (qpsmtpd/0.31) with ESMTP; Sat, 17 Nov 2007 17:57:46 +0000 Received: from soliton.cs.tamu.edu (vpn-18.cs.tamu.edu [128.194.150.18]) by postal.cs.tamu.edu (Postfix) with ESMTP id DA3C746DE11 for ; Sat, 17 Nov 2007 11:57:44 -0600 (CST) Received: by soliton.cs.tamu.edu (Postfix, from userid 1000) id 807A81AB10; Sat, 17 Nov 2007 12:09:11 -0600 (CST) To: "GCC Patches" Subject: Re: [C++ PATCH] Extend SFINAE to all expressions References: <24b520d20711170911g72392e51g36fc2504e8520dc1@mail.gmail.com> From: Gabriel Dos Reis In-Reply-To: <24b520d20711170911g72392e51g36fc2504e8520dc1@mail.gmail.com> Date: Sun, 18 Nov 2007 05:35:00 -0000 Message-ID: <87wssgeoc8.fsf@soliton.cs.tamu.edu> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Mailing-List: contact gcc-patches-help@gcc.gnu.org; run by ezmlm Precedence: bulk List-Id: List-Archive: List-Post: List-Help: Sender: gcc-patches-owner@gcc.gnu.org X-SW-Source: 2007-11/txt/msg00962.txt.bz2 "Doug Gregor" writes: [...] | So, this patch is big but mechanical. Its implementation was trivial | in the sense that only once did the addition of SFINAE support into a | function actually cause any regressions in the GCC testsuite. Weigh | that against the 256k of text in this patch, and you'll see just how | mechanical this patch really is. So, I know this patch is big for | stage 3, but it is relatively simple and will certainly be a boon for | C++ programmers trying out C++0x or digging into some of the more | arcane parts of C++ template argument deduction. Plus, committing it | will save me from my personal torment of keeping such a large patch in | sync with the trunk :) | | Tested i686-pc-linux-gnu and i386-apple-darwin-10.4.10; okay for mainline? * the general agreement on getting things in C++0x mode is that they must at the very least be in the C++ Workng Draft. This resolution to this very issue (close to my heart) is not. We don't even have a standardese for it yet. There still is a debate on where the border line is between hard error and sfinae, even if we generally we agreed (CWG+EWG) on the general phrase 'all expressions except ...' * We're in stage 3; we should come back to this when we are in stage 1 and we know what is in thw WD. -- Gaby