Bug in Emacs' batch mode

Bug in Emacs' batch mode

[Davin Pearson]

When I run a file in Emacs' batch mode like so:

emacs --batch --load foo.el

It prints out  the following information:

-*- mode: compilation; default-directory: "~/dlisp/" -*-
Compilation started at Tue May 12 11:06:41

make foo
emacs --batch --load foo.el
Loading cl...
Loading cl-macs...

Compilation finished at Tue May 12 11:06:43
Compilation took 2 seconds

This is a bug because I have redefined the function message to do
nothing.  Therefore it shouldn't print out "Loading cl..." and
"Loading cl-macs".  What follows is the contents of the file foo.el:

(defun message (format-string &rest args) (interactive))

(load-library "cl")
(load-library "cl-macs")
(require 'cl-19 "cl")
(require 'cl)

;; batch code goes here

bug#3283: Bug in Emacs' batch mode

[Kevin Rodgers]

Davin Pearson wrote:
> When I run a file in Emacs' batch mode like so:
> 
> emacs --batch --load foo.el
> 
> It prints out  the following information:
> 
> -*- mode: compilation; default-directory: "~/dlisp/" -*-
> Compilation started at Tue May 12 11:06:41
> 
> make foo
> emacs --batch --load foo.el
> Loading cl...
> Loading cl-macs...
> 
> Compilation finished at Tue May 12 11:06:43
> Compilation took 2 seconds
> 
> This is a bug because I have redefined the function message to do
> nothing.  Therefore it shouldn't print out "Loading cl..." and
> "Loading cl-macs".  What follows is the contents of the file foo.el:
> 
> (defun message (format-string &rest args) (interactive))
> 
> (load-library "cl")
> (load-library "cl-macs")
> (require 'cl-19 "cl")
> (require 'cl)
> 
> ;; batch code goes here

message is a built-in function in `C source code'.
(message format-string &rest args)

There are calls to Fmessage() in Emacs' C source code which do not go
through the `message' Lisp symbol.

-- 
Kevin Rodgers
Denver, Colorado, USA

Re: bug#3283: Bug in Emacs' batch mode

[Davin Pearson]

On May 14, 5:06 pm, Kevin Rodgers <kevin.d.rodg...@gmail.com> wrote:
> There are calls to Fmessage() in Emacs' C source code which do not go
> through the `message' Lisp symbol.

Isn't that a bug, since it seems reasonable to redefine message to a
do-nothing function?

bug#3283: Bug in Emacs' batch mode

[Kevin Rodgers]

Davin Pearson wrote:
> On May 14, 5:06 pm, Kevin Rodgers <kevin.d.rodg...@gmail.com> wrote:
>> There are calls to Fmessage() in Emacs' C source code which do not go
>> through the `message' Lisp symbol.
> 
> Isn't that a bug, since it seems reasonable to redefine message to a
> do-nothing function?

There are many calls to Ffoo() functions in Emacs' C source code which
do not go through the `foo' Lisp symbols.  Are they all bugs?

-- 
Kevin Rodgers
Denver, Colorado, USA