This is the mail archive of the cygwin@cygwin.com mailing list for the Cygwin project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]

Re: Up-to-date info on '-mno-cygwin' vs. Mingw32


On Fri, Aug 31, 2001 at 07:30:16AM -0400, David Carter wrote:
>Jesper Eskilson wrote:
>> According to the FAQ, support for the -mno-cygwin flag "has been weak
>>and flaky, [...] and maintenance of the option has *not* been a
>>priority in development", 
>
>I would tend to agree with this.

I'm not sure what FAQ is referring to this but it is rather
disappointing to hear that all of the work I've put into getting this
right is considered "weak" and "flaky".

For instance, I recently added code to binutils to allow isolation of
the various required libraries so that -mno-cygwin can only use mingw
and w32api libraries.

The only lack that I'm currently aware of is C++.  I have asked for
someone to send me a drop in libstdc++.a library for the current cygwin
version of gcc.  I don't think that anyone has taken me up on that
offer.

>>and the FAQ recommends that one uses a
>>separate MingW compiler set. Is this accurate?
>
>Here's what I recommend:
>- use the cygwin tools, passing -mno-cygwin to the compiler (& linker? I
>do it to be sure, but not sure if it's required)
>- use the mingw headers & libraries -- download them from mingw, and
>copy or symlink them into your cygwin /usr/include/mingw and
>/usr/lib/mingw directory trees. This is absolutely required if you're
>doing anything with c++ that you want to run on win32 without the
>cygwin1.dll

Hopefully most of the mingw headers and libraries are the same as
the ones in the current cygwin release.  If they aren't, they should
be updated.

Basically, I have been saying that -mno-cygwin is not something that I
enjoy working on since it is rather diametrically opposed to my main
cygwin activities.  However, I *have* spent a lot of time trying to
get it right, mainly because I don't like to release code that only
half works.

One of the reasons that I haven't released gcc 3.0* is that I would have
to patch the preprocessor to handle some of my "hacks" and the
preprocessor has changed drastically since 2.95.*.

cgf

--
Unsubscribe info:      http://cygwin.com/ml/#unsubscribe-simple
Bug reporting:         http://cygwin.com/bugs.html
Documentation:         http://cygwin.com/docs.html
FAQ:                   http://cygwin.com/faq/


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]