John Hendy <jw.he...@gmail.com> wrote: > On Mon, Sep 10, 2012 at 2:24 PM, Achim Gratz <strom...@nexgo.de> wrote: > > John Hendy writes: > >>> I don't really know where your other problems come from, but it seems > >>> that whatever you are invoking as "emacs" fails to properly set up the > >>> load path. You can try and replace the "-L" option with "--directory". > >> > >> I'll definitely try this. Where should the load path point? My emacs > >> dir, or the org git dir? > > > > What I meant was that if you add something to load-path via this option, > > make sure it actually turns up _first_ in the load-path. Like > > > > emacs -L blafasel > > > > should then have "blafasel" as the first component in load-path. If > > emacs is invoked via a wrapper scriptit may inject other options in > > places that subvert this. Another possibility is that an environment > > variable EMACS_LOADPATH is set (or gets set by some wrapper) when emacs > > is started, this would also explain why the load-path doesn't end up > > what it's supposed to be. > > Thanks, Achim. I just now sent a fresh note to the list about this > specific issue, as I pretty much derailed this into a discussion on > that error. It seemed fitting to make a new thread and I included a > lot more details. I think the issue comes down to the function > =org-find-library-dir=, not the load path. I could be wrong. I don't > know where =org-find-library-dir= is defined. >
C-h f org-find-library-dir RET shows that it's a macro in org-compat.el. Macros are a bit tricky -- see (info "(elisp) Compiling macros") for the gory details. Might this be the problem here? Nick
