Date: Wed, 31 Jan 2007 10:39:14 -0500
From: Christopher Faylor <cgf-use-the-mailinglist-please AT cygwin DOT com>
To: cygwin AT cygwin DOT com
Subject: Re: Eliminating -mno-cygwin from gcc?
Message-ID: <20070131153914.GB19137@trixie.casa.cgf.cx>
Reply-To: cygwin AT cygwin DOT com
Mail-Followup-To: cygwin AT cygwin DOT com
References: <20070131131337 DOT GA17256 AT trixie DOT casa DOT cgf DOT cx> <45C0971E DOT 4080305 AT byu DOT net> <20070131132700 DOT GA3478 AT implementation DOT labri DOT fr> <20070131133102 DOT GA17405 AT trixie DOT casa DOT cgf DOT cx> <20070131134842 DOT GU27843 AT calimero DOT vinschen DOT de>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20070131134842.GU27843@calimero.vinschen.de>
User-Agent: Mutt/1.5.11
Mailing-List: contact cygwin-help AT cygwin DOT com; run by ezmlm
Precedence: bulk
Sender: cygwin-owner AT cygwin DOT com
Mail-Followup-To: cygwin AT cygwin DOT com

On Wed, Jan 31, 2007 at 02:48:42PM +0100, Corinna Vinschen wrote:
>On Jan 31 08:31, Christopher Faylor wrote:
>> On Wed, Jan 31, 2007 at 02:27:00PM +0100, Samuel Thibault wrote:
>> >Eric Blake, le Wed 31 Jan 2007 06:18:22 -0700, a ?crit :
>> >> I would much rather call the cross-compiler i686-mingw-gcc than the
>> >> current name of 'gcc -mno-cygwin'.
>> >
>> >Same for me.
>> 
>> Thinking about this some more, it seems like we'd need a real
>> cygwin-based mingw cross compiler rather than a wrapped mingw compiler
>> since otherwise there would be path and signal issues.
>
>While I agree with the general idea, I have to add the obligatory hint
>that there are many projects out there which build environment requires
>`gcc -mno-cygwin' to work.  All of them will break with at least 50% of
>the lost user base asking on the Cygwin list for help.

Yes.  I think I made this point, although in a facetious manner.

The problem with "-mno-cygwin" is that its presence opens us up to all
sorts of confusion, as witnessed in the recent thread.  If the option is
there then surely we must have to support it right?  Why can't I pull in
cygwin headers?  That's what I want!

Anyway, we've decided to drop windows 98 support in Cygwin 1.7.0.  I
don't see this as much different.  We'll surely hear wails about no
support for Windows 9x/Me, too, but we've made the decision that support
for Windows 98 is not worth the effort.

Removing -mno-cygwin means that we'd be trading "How come I can't use
sys/socket.h when I type -mno-cygwin?" questions for "Why does gcc say
cc1: invalid option `no-cygwin'?"  If we did provide some sort of optional
shell script wrapper to be used with the mingw cross-compiler that should
satisfy people who really have to upgrade their C compilers.

What this may imply is that we need some sort of early warning system in
setup.exe.  If we had some way of saying "Warning: the following changes
have been made which could affect you.  Are you sure?" then maybe we
could eliminate a lot of people who are upset about make, bash, gcc, cygwin...

>> And, as long as we're talking about cross-compilers, a cygwin -> linux
>> cross compiler would probably be nice, too.
>
>Sorry, but we don't support Linux on this mailing list. ;)

Yes we do.  I just changed the cygwin web page.  :-)

cgf

--
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/