Re: AUCTeX conflicting with org-mode

[Nick Dokos <nicholas.dokos@hp.com>]

Yagnesh Raghava Yakkala <yagnesh@live.com> wrote:

> 
> Hi Suvayu,
> 
> suvayu ali <fatkasuvayu+linux@gmail.com> writes:
> 
> > Hi,
> >
> > I am having conflicts AUCTeX when I try to open plain latex files after
> > I exporting an org file to latex. These are the exact steps.
> >
> > 1. Start Emacs as: emacs -q (so that the site files load and I have
> >    auctex available).
> > 2. Open any org file and export to latex, C-c C-e l.
> > 3. Open another plain latex file.
> >
> > At this point I see this backtrace:
> >
> > Debugger entered--Lisp error: (void-variable TeX-master)
> >   #[nil "\301\302 !\203
> >   \303=\203  \304\305\211\306#\210\307\306!\207" [TeX-master
> > file-exists-p buffer-file-name shared TeX-master-file nil t
> > TeX-update-style] 4]()
> >   run-hooks(find-file-hook)
> >   after-find-file(nil t)
> >   find-file-noselect-1(#<buffer velo-links.tex>
> > "~/graphics/velo-links.tex" nil nil "~/graphics/velo-links.tex"
> > (263573 64770))
> >   find-file-noselect("~/graphics/velo-links.tex" nil nil t)
> >   find-file("~/graphics/velo-links.tex" t)
> >   call-interactively(find-file nil nil)
> >
> > Since removing the auctex site files out of the site-lisp directory
> > resolves the conflict I am assuming the problem stems from my auctex
> > installation.
> >
> > Can someone help me trouble shoot this?
> 
> I can reproduce this., I think your auctex installation is fine. Its
> definitely from org.
> 
> In org-latex.el Tex-master is set to t to suppress auctex from asking for
> that variable while exporting with org-latex..(IIRC there was a discussion and
> Carsten made that change)
> 
> Some how org leaving that to void at the end of org-latex export (I coundn't
> figure out how?).
> 

That sounds far-fetched to me: once TeX-master is bound, one has to go
to some lengths to unbind it (see makunbound - but I'm pretty sure
that's not used anywhere in org code - there are a few instances of the
function analog, fmakunbound).

It seems much more likely that in these cases, it never got bound in the
first place, possibly because of dependency problems in the loading of
packages. But I cannot reproduce the problem, so I'm speaking
theoretically only and my imagination may be failing me.

Nick