On Wed, 1 Mar 2000, John Fortin wrote:

> Marcus Sundberg wrote:
> > 
> > Andreas Beck <[EMAIL PROTECTED]> writes:
> > 
> > > > Is there a way to have it determine the correct extension, or add a
> > > > command-line to configure to change this?
> > >
> > > Sure. It is already prepared in the configure.in script. The DLLEXT variable
> > > has to be set according to the OS detected.
> > >
> > > This can be done like this:
> > >
> > > if test "$ac_cv_header_windows_h" != "yes"; then
> > >   DLLEXT=dll
> > > fi
> > 
> > Not really nice, I'd like to be able to have whatever headers I want
> > on my system without things breaking. Better to do:
> > 
> > case "${target}" in
> >   *-*-mingw* | *-*-cygwin*)
> >         DLLEXT="dll"
> > 
> > I've written a GGI_DLLEXT macro which does just this, please verify
> > that it works correctly when compiling with mingw32/cygwin.
> > 
> 
> Well, it mostly works ;)  DLLEXT get set to DLL.  That is, it gets
> printed during configure.  However, it does not get inserted into
> libggi.conf :( @DLLEXT@ is not replaced.  Here is part of the
> libggi.conf which results.
> 
> John
> 
> # Mapping file for suggest-strings to libraries
> .root: e:/usr/local/lib/ggi
> #
> # suggest_name                        library_name                    versions
> # ------------                        --------------------------      --------
> # Displays
> display-kgi                   display/kgi.@DLLEXT@
> display-suidkgi                       display/suidkgi.@DLLEXT@
> display-x                     display/X.@DLLEXT@
> display-xlib                  display/Xlib.@DLLEXT@
> display-dga                   display/xf86dga.@DLLEXT@
> display-svga                  display/svgalib.@DLLEXT@
> display-svgalib                       display/svgalib.@DLLEXT@
> helper-vgagl                  display/vgagl.@DLLEXT@
> display-aa                    display/aa.@DLLEXT@
> display-directx                       display/directx.@DLLEXT@
> display-directxnt             display/directxnt.@DLLEXT@
> 
> 
> <snip>
> 

Yes, that's true! I have exactly the same problem.
But the problem is not _only_ on libggi, the problem is on _all_ libs
which are in the current cvs-tree...


Christoph Egger
E-Mail: [EMAIL PROTECTED]

Reply via email to