Mailing-List: contact cygwin-help AT cygwin DOT com; run by ezmlm List-Subscribe: List-Archive: List-Post: List-Help: , Sender: cygwin-owner AT cygwin DOT com Mail-Followup-To: cygwin AT cygwin DOT com Delivered-To: mailing list cygwin AT cygwin DOT com Message-ID: <20031013220827.13993.qmail@web21405.mail.yahoo.com> Date: Tue, 14 Oct 2003 08:08:27 +1000 (EST) From: =?iso-8859-1?q?Danny=20Smith?= Subject: Re: merging mingw and cygwin To: cygwin AT cygwin DOT com MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit Ed said: > Like I said, try: > > mingw > gcc -dM -e -xc /dev/null > cygwin > gcc -mno-cygwin -dM -E -xc /dev/null > > cygwin makes 73 defines, mingw makes 38. If a large project uses any of the > cygwin defines, it will behave differently than if compiled with native mingw. > That's because you used gcc-3.2.3 with mingw (3.3.1 is still a release candidate "over there") and gcc-3.3.1 with cygwin. I agree: A program compiled with gcc-3.2.3 will be different than one compiled with 3.3.1. So maybe you should harass the mingw list to update the release status of gcc. On second thought, don't do that. I hear they are petty mean over there, too. > As I said, this is just the tip of the iceberg - who knows what patches that > mingw has made to gcc, ld, make, etc. which could affect the building and > running of large win32 packages. I do. So does Chris, So does anyone who cares to look. The diff for gcc and binutils is not an iceberg. > > If the large packages built in mingw are tested via mingw, then mingw is the > only real way to a 'proper' win32 executable. And the only way to truly > emulate mingw32 would be by merging it. Wrong. I can build binutils for mingw with cygwin gcc -mno-cygwin. It is the same as the binutils that I build with mingw. I have built a mingw gcc with cygwin gcc -mno-cygwin. AFAICT, it is the same as the gcc I build with native mingw. I don't do it any more because I like to say that I can do a native bootstrap of gcc for mingw (with the help of cygwin tools) > > > Maybe the MingW package in Cygwin needs to be updated, however, I fail > > to see the need for a MINGW or NO_CYGWIN flavor aside from what > > currently exists (i.e. -mno-cygwin). > > Because gcc is not the only place that has MINGW-isms in it; msys departs from > the cygwin standard and handles certain things differently. msys != mingw. mingw doesn't need msys. Cygwin provides a more complete building and testing environment than does msys. > In order > to build MinGW packages right, the underlying tools have to work the same > as well. > > MINGW and/or NO_CYGWIN simply wrap all of this up in a nice user friendly > package. cygwin is a nice user friendly package. I won't speak for mingw because I have a personal bias. Danny > > Ed > http://search.yahoo.com.au - Yahoo! Search - Looking for more? Try the new Yahoo! Search -- Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple Problem reports: http://cygwin.com/problems.html Documentation: http://cygwin.com/docs.html FAQ: http://cygwin.com/faq/