Re: Redundant (harmful) duplication of run-hooks in define-globalized-minor-mode [patch-2]

[Alan Mackenzie]

Hi, Stefan.

On Thu, Jan 31, 2013 at 09:38:31AM -0500, Stefan Monnier wrote:
> > +    (MODE-cancel-disable
> > +     (intern (concat global-mode-name "-cancel-disable")))
> > +    (MODE-disable-in-buffer
> > +     (intern (concat global-mode-name "-disable-in-buffer")))

> If you defvar-local the MODE-disable-in-buffer, .....

That would be disable-MODE I think you meant.  Yes, I've now made
disable-MODE buffer local.  It has to be buffer local, considering the
way that MODE-enable-in-buffers checks all buffers each time it runs.

> .... then you shouldn't need MODE-cancel-disable since
> kill-all-local-variables will have reset MODE-disable-in-buffer to nil
> already.

I'm not so sure about this.  These complicated structures of macros and
hooks and generated functions are making my head hurt.  ;-(  Are you
sure there are no ways of invoking this which won't bypass
kill-all-local-variables?  (That's a real question, not a rhetorical
one.)

Another thought.  As I've coded it up, the disable-MODE flag, once it
becomes t, stays t (upto the next major mode change).  However, running
global-MODE still enables MODE on this buffer.  Should we worry about
this?  I don't think we need to.

[Amended patch not included.]

>         Stefan

-- 
Alan Mackenzie (Nuremberg, Germany).