mailing-list for TeXmacs Users

[Martin Costabel <address@hidden>]

Martin Costabel wrote:
[]
> are done from inside the texmacs.info file. But this does, of course, 
> not mean that new versions will be guaranteed to compile without new 
> patches. Quite often, little tweaks have been necessary.

This time (version 1.0.2.5), the little tweak that was necessary was 
caused by a bug in src/common.makefile.in:

Lines 245/246:

$(plugins_obj): Objects/%.o: %.cpp
        $(call cxx_incl,$(plugins_incl),$(CXXXWINDOW))

This does *not* put the necessary "-I/usr/X11R6/include" into the 
compiler command line for the Plugins. The reason is that cxx_incl only 
takes 1 argument. One cannot use cxx_incl_sys (which takes 2 arguments) 
either, because then the freetype2 headers will no longer be found, 
because they are included before the "-I-" separator.

The bug makes the compilation of the Ghostscript Plugin crash. I suspect 
it doesn't show up on Linux, because there /usr/include/X11 is probably 
a symlink to /usr/X11R6/include/X11, and therefore the X headers will be 
found without CXXXWINDOW.

A Fink package description that compiles under MacOSX 10.2 with gcc-3.3 
should shortly show up on unstable CVS/rsync. For MacOSX 10.3, I need a 
little more time, because I can use it only since this morning and there 
are a couple of other things to compile first before texmacs gets its turn.

--
Martin