Re: [Adonthell-devel] Partly merging v0.4 back

[Alexandre Courbot]

> g++ -g -Wall -fno-exceptions -fno-rtti -DDATA_DIR=\"/usr/share/adonthell\" -o 
> adonthell main.o -Wl,-E  libadonthell.a -L/usr/lib /usr/lib/libSDL.so 
> -L/usr/X11R6/lib -lX11 -lXext /usr/lib/libvorbis.so /usr/lib/libvorbisfile.so 
> -lm /usr/lib/libogg.so -L/usr/lib/python2.1/config -lpython2.1 -lpthread -ldl 
> -lutil -lz
> ./adonthell -c
> Traceback (most recent call last):
>   File "./adonthell.py", line 3229, in ?

This one was easy to fix: it was because python::init did much more than
it should. When you pass '-c' to adonthell it must NOT import the
adonthell module, as you can't expect it to be installed. However, you
moved this code into python::init, which is called when '-c' is passed.

This is fixed on the CVS, however I had to make python::module public.
Don't remember how we did before! :)

About the po directory failure, this is because top_builddir doesn't
appear in Makefile, while it should. What shall we do here? Add it to
Makefile.in.in? I don't exactly know how things are generated here, as
it's a bit different from the classic automake/autoconf stuff.

Alex.
-- 
http://www.gnurou.org