[prev in list] [next in list] [prev in thread] [next in thread] 

List:       gtkmm
Subject:    Re: [gtkmm] gtkmm on Windows
From:       Cédric Gustin <cgustin () ibelgique ! com>
Date:       2001-12-19 7:31:35
[Download RAW message or body]


My two cents :

1. Compile only with Mingw 1.1 (mingw support in cygwin through the 
-mno-cygwin flag is somewhat broken for c++).
2. Do NOT try to compile gtkmm with cygwin gcc using the gtk+ port for 
win32. The cygwin posix layer does not interact correctly with the MSVCRT.dll.
3. Get libsigc++-1.0.x from CVS. It includes (I submitted a patch a few 
months ago) a modified sigc++config.h.in that handles mingw correctly and 
even generates a DLL on the fly.
4. Do NOT try to generate a c++ DLL with libtool (cygwin or mingw). C++ DLL 
support in libtool, even in the latest release is really broken. What I do 
(see sigc++/Makefile.am in CVS for example) is to create a static library 
then convert it to a DLL and import library using dlltool/dllwrap.

If you don't want to spend 3 days (or more) compiling gtkmm for win32, I 
recommend to use my binaries. Or, if you prefer, I can send you my modified 
sources.

Once again, it's just my 2 cents...

Cedric

At 04:42 PM 12/18/2001 -0800, you wrote:
> > Ah, and something else:
> > >> If I compile with MinGW 1.1 instead, I get an error early during
> > >> compilation:
> > >> > In file included from ../sigc++/node.h:21,
> > >> >                  from node.cc:19:
> > >> > ../sigc++/config/sigcconfig.h:52: #error "Unknown architecture 
> (send me
> > gcc
> > >>  --dumpspecs)"
> > >>
> > >> 'gcc --dumpspecs' says 'gcc.exe: No input files', which probably
> > >> doesn't help much either. :-)
> > >
> > >Hmm.  Okay.  Guess I will need to find out how to get the symbols from 
> that.
> >
> > BTW, using 'gcc -mno-cygwin' gives the same output. Previously, I used
> > the wrong switch (--no-cygwin instead of -mno-cygwin).
> >
> > Encouraged by this, I also tried 'gcc -dumpspecs'. Only one minus! You
> > really like that double-minus, heh? :-)  The result is appended to
> > this message.
>
>Hmm.  Unix specifications for standards say multichar arguments
>should be proceeded by --.  Guess no one told that to gcc group. :-)
>
>--Karl
>
>_______________________________________________
>to unsubscribe or change your subscription parameters :
>https://lists.sourceforge.net/lists/listinfo/gtkmm-main

 
______________________________________________________________________________
ifrance.com, l'email gratuit le plus complet de l'Internet !
vos emails depuis un navigateur, en POP3, sur Minitel, sur le WAP...
http://www.ifrance.com/_reloc/email.emailif



_______________________________________________
to unsubscribe or change your subscription parameters :
https://lists.sourceforge.net/lists/listinfo/gtkmm-main
[prev in list] [next in list] [prev in thread] [next in thread]