bug#3350: 23.0.93; Deactivation of region by the insert function

bug#3350: 23.0.93; Deactivation of region by the insert function

[Lennart Borgman]

Start from

  emacs -Q

and enter this in the *scratch* buffer and eval it:

(defvar msgtrc-buffer
  "*Messages*"
  "Buffer or name of buffer for trace messages.
See `msgtrc'.")


(defun msgtrc (format-string &rest args)
  "Print message to `msgtrc-buffer'.
Arguments FORMAT-STRING and ARGS are like for `message'."
  (let ((trc-buffer (get-buffer-create msgtrc-buffer)))
    (with-current-buffer trc-buffer
      (goto-char (point-max))
      ;;(insert "MU:" (apply 'format format-string args) "\n")
      (insert "constant string\n")
      (when buffer-file-name (write-region nil nil buffer-file-name)))))

(defun msgtrc-post-command ()
  (msgtrc "msgtrc post-command here"))

(add-hook 'post-command-hook 'msgtrc-post-command nil t)


Then try activating the region with S-<arrow key>. This does not work.
However if the (insert ...) statement in msgtrc is commented out then
it works.

I have not tested if this is a regression from Emacs 22.


In GNU Emacs 23.0.93.1 (i386-mingw-nt5.1.2600)
 of 2009-05-03
Windowing system distributor `Microsoft Corp.', version 5.1.2600
configured using `configure --with-gcc (3.4) --no-opt --cflags
-Ic:/g/include -fno-crossjumping'

bug#3350: 23.0.93; Deactivation of region by the insert function

[Stefan Monnier]

> Thanks, I see.  But it is pretty surprising ...
> What is the reason that deactivate-mark is not buffer local?

There isn't any.


        Stefan

bug#3350: 23.0.93; Deactivation of region by the insert function

[Lennart Borgman]

On Mon, May 25, 2009 at 4:06 AM, Stefan Monnier
<monnier@iro.umontreal.ca> wrote:
>> Thanks, I see.  But it is pretty surprising ...
>> What is the reason that deactivate-mark is not buffer local?
>
> There isn't any.

Then please make it buffer local by default.

Also I think that visible active regions still should be visible when
they are not in the selected window though with another color
(probably grey). That would be more consisten with what other apps do.

bug#3350: 23.0.93; Deactivation of region by the insert function

[Stefan Monnier]

>>> Thanks, I see.  But it is pretty surprising ...
>>> What is the reason that deactivate-mark is not buffer local?
>> There isn't any.
> Then please make it buffer local by default.

Patch welcome.

> Also I think that visible active regions still should be visible when
> they are not in the selected window though with another color
> (probably grey). That would be more consisten with what other apps do.

That's a separate concern.  Please make a separate bug-report for that
feature request.


        Stefan