On Sat, Nov 30 2013, David Bremner <david@tethera.net> wrote: > Tomi Ollila <tomi.ollila@iki.fi> writes: > >> >> ifeq ($(WITH_EMACS),1) >> ifeq ($(HAVE_EMACS),1) >> all: $(emacs_bytecode) >> +install-emacs: $(emacs_bytecode) >> endif >> > > It's a bit confusing that we have both HAVE_EMACS and WITH_EMACS. Should > this be documented somewhere? Otherwise the patch seems sensible. > > See also the discussion at id:528B455A.70200@iki.fi. Eventually it's probably good idea to merge HAVE_EMACS to WITH_EMACS but doing in a good way requires quite a few ./configure changes (and while doing that one have to check that some consistency is kept). Eventually it would probably be nice that in case *_EMACS is not set (is 0) the clean rules stay (provided that those do not require emacs) but build rules are ifdeffed out... In the meantime my patch here maintains status quo where things can be done in case HAVE_EMACS is set (1) -- emacs is available in the system, while keeping the cleanup rules in case *_EMACS is not set... and this patch also fixes the compilation whenever there is no emacs in system. I could add short notes of what HAVE_EMACS & WITH_EMACS mean but although it is bit confusing that we currently have both, those are nevertheless quite descriptive themselves. > > d Tomi > _______________________________________________ > notmuch mailing list > notmuch@notmuchmail.org > http://notmuchmail.org/mailman/listinfo/notmuch