On Jul 24, 2006, at 8:06pm, Kalle Olavi Niemitalo writes: >"M. Levinson" <[EMAIL PROTECTED]> writes: > >> 4. ./configure --with-python didn't work because the Python tests were >> trying an incorrect library name and didn't produce all the necessary >> compiler and linker flags. > >Witek already fixed that as bug 768. Consequently, your patch >does not apply cleanly to the current configure.in (now at commit >098fb87065e2c69f4f9ef65d16901e336a72b789).
Ah, I must be too slow. :-) I'll update my tree.... >> (Also, they were setting CFLAGS where CPPFLAGS is probably >> more appropriate.) > >Agreed! But, doesn't this require changing the Makefiles too? Hmm, it looks like none of this actually matters for the scripting backends: At least in my now-outdated tree (from earlier this month), the makefiles under src/scripting/* use $(LUA_CFLAGS), $(PERL_CFLAGS), $(PYTHON_CFLAGS), etc., so it made no difference that $(CPPFLAGS) was never used. Consequently there appears to be no reason for the configure script to add any of these to CPPFLAGS in the first place. _______________________________________________ elinks-dev mailing list elinks-dev@linuxfromscratch.org http://linuxfromscratch.org/mailman/listinfo/elinks-dev
