eval-after-load weirdness

eval-after-load weirdness

[Trent Buck]

[-- Attachment #1.1: Type: text/plain, Size: 2569 bytes --]

[Please CC me; I'm not subscribed to the list.]

I'm noticing the most bizarre symptom with my .emacs[0]:
eval-after-load works right when passed a symbol, but not when passed
a string.  Some examples:

    $ emacs-unicode-2 -f toggle-debug-on-error `mktemp`
    Debugger entered--Lisp error: (void-variable vc-directory-exclusion-list)
      add-to-list(vc-directory-exclusion-list ".bzr" t)
      eval((add-to-list (quote vc-directory-exclusion-list) ".bzr" t))
      eval-after-load("vc" (add-to-list (quote vc-directory-exclusion-list) ".bzr" t))

    $ emacs-snapshot -f toggle-debug-on-error -f gnus
    Debugger entered--Lisp error: (void-variable mail-mode-map)
      (define-key mail-mode-map "\x03\x01" (quote mail-abbrev-insert-alias))
      (progn (define-key mail-mode-map "\x03\x01" (quote mail-abbrev-insert-alias))
             (define-key mail-mode-map "^[	" (quote mail-abbrev-complete-alias)))
      eval((progn (define-key mail-mode-map "\x03\x01" (quote mail-abbrev-insert-alias))
                  (define-key mail-mode-map "^[	" (quote mail-abbrev-complete-alias))))
      eval-after-load("sendmail" (progn (define-key mail-mode-map "\x03\x01" (quote mail-abbrev-insert-alias))
                                        (define-key mail-mode-map "^[	" (quote mail-abbrev-complete-alias))))

Adding -q (i.e. disabling my .emacs) stops these errors.  I'm at my
wits end... can anyone suggest what could be causing this?

[0] http://twb.ath.cx/~twb/scratch/profile/.emacs
    http://twb.ath.cx/~twb/scratch/profile/.emacs-prefs/
    http://twb.ath.cx/~twb/scratch/profile/.emacs-apps/

In GNU Emacs 23.0.0.1 (x86_64-pc-linux-gnu, X toolkit, Xaw3d scroll bars)
 of 2007-05-13 on baal, modified by Debian
 (Debian emacs-unicode-2 package, version 1:20070513-1)
Windowing system distributor `The X.Org Foundation', version 11.0.10300000
configured using `configure  '--build' 'x86_64-linux-gnu' '--host' 'x86_64-linux-gnu' '--prefix=/usr' '--sharedstatedir=/var/lib' '--libexecdir=/usr/lib' '--localstatedir=/var' '--infodir=/usr/share/info' '--mandir=/usr/share/man' '--with-pop=yes' '--enable-locallisppath=/etc/emacs-unicode-2:/etc/emacs:/usr/local/share/emacs/23.0.0/site-lisp:/usr/local/share/emacs/site-lisp:/usr/share/emacs/23.0.0/site-lisp:/usr/share/emacs/site-lisp:/usr/share/emacs/23.0.0/leim' '--enable-font-backend' '--with-xft' '--with-x=yes' '--with-x-toolkit=athena' '--with-toolkit-scroll-bars' 'build_alias=x86_64-linux-gnu' 'host_alias=x86_64-linux-gnu' 'CFLAGS=-DDEBIAN -DSITELOAD_PURESIZE_EXTRA=5000 -g -O2''
-- 
Trent Buck

[-- Attachment #1.2: Digital signature --]
[-- Type: application/pgp-signature, Size: 189 bytes --]

[-- Attachment #2: Type: text/plain, Size: 149 bytes --]

_______________________________________________
bug-gnu-emacs mailing list
bug-gnu-emacs@gnu.org
http://lists.gnu.org/mailman/listinfo/bug-gnu-emacs

Re: eval-after-load weirdness

[Trent Buck]

"Trent Buck" <trentbuck@gmail.com> writes:
> I'm noticing the most bizarre symptom with my .emacs[0]:
> eval-after-load works right when passed a symbol, but not when
> passed a string.  Some examples:
>
>     $ emacs-unicode-2 -f toggle-debug-on-error `mktemp`
>     Debugger entered--Lisp error: (void-variable vc-directory-exclusion-list)
>       add-to-list(vc-directory-exclusion-list ".bzr" t)
>       eval((add-to-list (quote vc-directory-exclusion-list) ".bzr" t))
>       eval-after-load("vc" (add-to-list (quote vc-directory-exclusion-list) ".bzr" t))

I worked out what the trouble is: because my modular .emacs setup has
each configuration module named after the library it configures
(e.g. ~/.emacs-prefs/vc.el for the vc library), eval-after-load thinks
"vc" has already been loaded.  Obviously (featurep 'vc) is still nil
so giving symbols to e-a-l works.

Some solutions suggested on #emacs:

    - stuff everything back into a single .emacs;

    - prefix filenames with digits (e.g. 20vc.el), like system V init
      scripts; and

    - instead of LOADing each config module, FIND-FILE it and then
      EVAL-BUFFER it.

I produced a minimal example of the problem at
http://paste.lisp.org/display/41451
--
Trent Buck

Re: eval-after-load weirdness

[Miles Bader]

Trent Buck <trentbuck@gmail.com> writes:
> because my modular .emacs setup has each configuration module named
> after the library it configures (e.g. ~/.emacs-prefs/vc.el for the vc
> library), eval-after-load thinks "vc" has already been loaded.
...
> Some solutions suggested on #emacs:
>     - stuff everything back into a single .emacs;
>     - prefix filenames with digits (e.g. 20vc.el), like system V init
>       scripts; and
>     - instead of LOADing each config module, FIND-FILE it and then
>       EVAL-BUFFER it.

You could also just remove the init files from `load-history' after the
fact.

E.g. (at the end of your .emacs):

   (let ((remove-regexp
          (concat "\\`" (regexp-quote (file-truename "~/.emacs-prefs/"))))
         (new-load-hist nil))
     (dolist (entry load-history)
       (unless (string-match remove-regexp (car entry))
         (push entry new-load-hist)))
     (setq load-history (nreverse new-load-hist)))

-miles
-- 
(\(\
(^.^)
(")")
*This is the cute bunny virus, please copy this into your sig so it can spread.

Re: eval-after-load weirdness

[Trent Buck]

[-- Attachment #1.1: Type: text/plain, Size: 2519 bytes --]

On Wed, May 23, 2007 at 07:59:22AM +0900, Miles Bader wrote:
> Trent Buck <trentbuck@gmail.com> writes:
> > because my modular .emacs setup has each configuration module named
> > after the library it configures (e.g. ~/.emacs-prefs/vc.el for the vc
> > library), eval-after-load thinks "vc" has already been loaded.
> ...
> > Some solutions suggested on #emacs:
> >     - stuff everything back into a single .emacs;
> >     - prefix filenames with digits (e.g. 20vc.el), like system V init
> >       scripts; and
> >     - instead of LOADing each config module, FIND-FILE it and then
> >       EVAL-BUFFER it.
> 
> You could also just remove the init files from `load-history' after the
> fact.
> 
> E.g. (at the end of your .emacs):
> 
>    (let ((remove-regexp
>           (concat "\\`" (regexp-quote (file-truename "~/.emacs-prefs/"))))
>          (new-load-hist nil))
>      (dolist (entry load-history)
>        (unless (string-match remove-regexp (car entry))
>          (push entry new-load-hist)))
>      (setq load-history (nreverse new-load-hist)))

I think that would solve EVAL-AFTER-LOADs in Emacs itself, but not
those scattered throughout .emacs-prefs/ itself.  Here is the change I
eventually made per method (3), which solves the problem.

    Mon May 21 19:30:33 EST 2007  Trent Buck <trentbuck@gmail.com>
      * .emacs: don't use LOAD in .emacs-prefs.
      Doing so breaks EVAL-AFTER-LOAD when passed a string as argument.
      See http://paste.lisp.org/display/41451 for details.
    diff -rN -u old-profile/.emacs new-profile/.emacs
    --- old-profile/.emacs  2007-05-23 11:39:23.219566437 +1000
    +++ new-profile/.emacs  2007-05-23 11:39:23.267571168 +1000
    @@ -17,9 +17,11 @@
     (let ((default-directory "~/.emacs-prefs/"))
       (when (file-accessible-directory-p ".")
         (load (expand-file-name "Emacs.el") t)
         (load (expand-file-name
                (format "window-system-%s.el" window-system)) t)
    -    (mapc (lambda (x)
    -            (when (locate-library (file-name-sans-extension x))
    -              (load (expand-file-name x))))
    -          (directory-files "." nil "\\`[a-z].*\\.el\\'"))))
    +    (with-temp-buffer
    +      (mapc (lambda (x)
    +              (when (locate-library (file-name-sans-extension x))
    +                (insert-file-contents (expand-file-name x))))
    +            (directory-files "." nil "\\`[a-z].*\\.el\\'"))
    +      (eval-buffer))))
-- 
Trent Buck

[-- Attachment #1.2: Digital signature --]
[-- Type: application/pgp-signature, Size: 189 bytes --]

[-- Attachment #2: Type: text/plain, Size: 149 bytes --]

_______________________________________________
bug-gnu-emacs mailing list
bug-gnu-emacs@gnu.org
http://lists.gnu.org/mailman/listinfo/bug-gnu-emacs