bug#7014: 23.2; `C-h c' command doesn't show the info when invoked from the minibuffer

bug#7014: 23.2; `C-h c' command doesn't show the info when invoked from the minibuffer

[Dani Moncayo]

Hello,

Try the following:
1.- Start emacs (anyhow).
2.- Type `C-x C-f'.
3.- Now, from the minibuffer, type `C-h c' --> Here, Emacs prompts me
for a key to describe. OK.
4.- Type `C-n' (for example). --> Now Emacs should show me the
description of the `C-n' key (*), but that doesn't happens. Instead,
after a couple of seconds, the minibuffer comes back with the prompt
of the `find-file' command.



(*) Look at the Emacs manual, chapter 8 ("The Minibuffer"):
>   Since the minibuffer appears in the echo area, it can conflict with
> other uses of the echo area.  If an error occurs while the minibuffer
> is active, the error message hides the minibuffer for a few seconds, or
> until you type something; then the minibuffer comes back.  If a command
> such as `C-x =' needs to display a message in the echo area, the
> message hides the minibuffer for a few seconds, or until you type
> something; then the minibuffer comes back.  While the minibuffer is in
> use, keystrokes do not echo.


Thanks in advance. Dani.


In GNU Emacs 23.2.1 (i686-pc-linux-gnu, GTK+ Version 2.20.1)
 of 2010-07-31 on jesus-desktop
Windowing system distributor `The X.Org Foundation', version 11.0.10706000
Important settings:
  value of $LC_ALL: nil
  value of $LC_COLLATE: nil
  value of $LC_CTYPE: nil
  value of $LC_MESSAGES: nil
  value of $LC_MONETARY: nil
  value of $LC_NUMERIC: nil
  value of $LC_TIME: nil
  value of $LANG: es_ES.utf8
  value of $XMODIFIERS: nil
  locale-coding-system: utf-8-unix
  default enable-multibyte-characters: t

bug#7014: 23.2; `C-h c' command doesn't show the info when invoked from the minibuffer

[Glenn Morris]

Dani Moncayo wrote:

> 2.- Type `C-x C-f'.
> 3.- Now, from the minibuffer, type `C-h c' --> Here, Emacs prompts me
> for a key to describe. OK.
> 4.- Type `C-n' (for example). --> Now Emacs should show me the
> description of the `C-n' key (*), but that doesn't happens. Instead,
> after a couple of seconds, the minibuffer comes back with the prompt
> of the `find-file' command.

For me, the message "C-n runs the command next-line" appears for a
second or two, then the "Find file" prompt reappears. This seems correct
to me.

bug#7014: 23.2; `C-h c' command doesn't show the info when invoked from the minibuffer

[Dani Moncayo]

On Wed, Jul 13, 2011 at 03:50, Glenn Morris <rgm@gnu.org> wrote:
> Dani Moncayo wrote:
>
>> 2.- Type `C-x C-f'.
>> 3.- Now, from the minibuffer, type `C-h c' --> Here, Emacs prompts me
>> for a key to describe. OK.
>> 4.- Type `C-n' (for example). --> Now Emacs should show me the
>> description of the `C-n' key (*), but that doesn't happens. Instead,
>> after a couple of seconds, the minibuffer comes back with the prompt
>> of the `find-file' command.
>
> For me, the message "C-n runs the command next-line" appears for a
> second or two, then the "Find file" prompt reappears. This seems correct
> to me.
>

I've just checked this in [1], and I can confirm that:
1. It works right from plain "emacs -Q".
2. It is still reproducible if you disable blink-cursor-mode.


Footnotes
[1]
In GNU Emacs 24.0.50.1 (i386-mingw-nt5.1.2600)
 of 2011-06-27 on 3249CTO
Windowing system distributor `Microsoft Corp.', version 5.1.2600
configured using `configure --with-gcc (4.5) --no-opt --cflags
-Ic:/build/include'


-- 
Dani Moncayo

bug#7014: 23.2; `C-h c' command doesn't show the info when invoked from the minibuffer

[Eli Zaretskii]

> Date: Wed, 13 Jul 2011 08:30:28 +0200
> From: Dani Moncayo <dmoncayo@gmail.com>
> Cc: 7014@debbugs.gnu.org
> 
> I've just checked this in [1], and I can confirm that:
> 1. It works right from plain "emacs -Q".
> 2. It is still reproducible if you disable blink-cursor-mode.

Look in the *Messages* buffer, and you will see that the message is,
in fact, displayed, it just gets immediately overwritten.

I guess some incoming event causes some sit-for to exit early, or
something.

bug#7014: 23.2; `C-h c' command doesn't show the info when invoked from the minibuffer

[Chong Yidong]

Dani Moncayo <dmoncayo@gmail.com> writes:

> Try the following:
> 1.- Start emacs (anyhow).
> 2.- Type `C-x C-f'.
> 3.- Now, from the minibuffer, type `C-h c' --> Here, Emacs prompts me
> for a key to describe. OK.
> 4.- Type `C-n' (for example). --> Now Emacs should show me the
> description of the `C-n' key (*), but that doesn't happens. Instead,
> after a couple of seconds, the minibuffer comes back with the prompt
> of the `find-file' command.

Fixed in trunk.

bug#7014: 23.2; `C-h c' command doesn't show the info when invoked from the minibuffer

[Drew Adams]

> > Try the following:
> > 1.- Start emacs (anyhow).
> > 2.- Type `C-x C-f'.
> > 3.- Now, from the minibuffer, type `C-h c' --> Here, Emacs 
> prompts me
> > for a key to describe. OK.
> > 4.- Type `C-n' (for example). --> Now Emacs should show me the
> > description of the `C-n' key (*), but that doesn't happens. Instead,
> > after a couple of seconds, the minibuffer comes back with the prompt
> > of the `find-file' command.
> 
> Fixed in trunk.

I have not seen the fix yet, but I certainly hope that the currently active
keymaps are used for this.

In particular, with the example given, the doc for `C-n' should list the command
it is bound to as being `exit-minibuffer', since the question about its binding
is being asked with a minibuffer keymap current, and that is the binding of
`C-n' in the minibuffer.

This is important.  Emacs should always faithfully answer questions about itself
and its current state.  When in the minibuffer, users should be able to ask and
get answers to questions about the current, minibuffer, state.

If for some reason some particular context/code needs to instead interrogate the
keymaps that were current before the initial command (`C-x C-f in this case), it
can do that (explicitly).

But a simple `C-h c' when the minibuffer is active should (must) DTRT and give
the currently active binding.  Anything else would be counter the Emacs
philosophy of dynamic introspection.  Emacs is not only self-documenting editor,
it documents itself in real-time.  And nothing about Emacs is more important
than that.

bug#7014: 23.2; `C-h c' command doesn't show the info when invoked from the minibuffer

[Dani Moncayo]

> I have not seen the fix yet, but I certainly hope that the currently active
> keymaps are used for this.

It's apparently so.

> In particular, with the example given, the doc for `C-n' should list the command
> it is bound to as being `exit-minibuffer', since the question about its binding
> is being asked with a minibuffer keymap current, and that is the binding of
> `C-n' in the minibuffer.

?? `C-n' from the minifuffer doesn't exit the minibuffer.  It seems to
be bound to `next-line', as reported if you do the experiment in the
original message.

> This is important.  Emacs should always faithfully answer questions about itself
> and its current state.  When in the minibuffer, users should be able to ask and
> get answers to questions about the current, minibuffer, state.

Agreed, but that seems to be the current behavior.  Try for example
[C-x C-f C-h c M-p].  Emacs will tell you that "M-p runs the command
previous-history-element".  It looks ok to me.

-- 
Dani Moncayo

bug#7014: 23.2; `C-h c' command doesn't show the info when invoked from the minibuffer

[Drew Adams]

> > I have not seen the fix yet, but I certainly hope that the 
> > currently active keymaps are used for this.
> 
> It's apparently so.
> 
> > In particular, with the example given, the doc for `C-n' 
> > should list the command it is bound to as being
> > `exit-minibuffer', since the question about its binding
> > is being asked with a minibuffer keymap current, and that 
> > is the binding of `C-n' in the minibuffer.
> 
> ?? `C-n' from the minifuffer doesn't exit the minibuffer.  It seems to
> be bound to `next-line', as reported if you do the experiment in the
> original message.

Sorry, I was thinking that the request example concerned `C-m' (`RET'), not
`C-n'.  (And of course for `C-m' the binding depends on which minibuffer map is
current.  It is `exit-minibuffer' in some cases but not all.)

> > This is important.  Emacs should always faithfully answer 
> > questions about itself and its current state.  When in the
> > minibuffer, users should be able to ask and get answers to
> > questions about the current, minibuffer, state.
> 
> Agreed, but that seems to be the current behavior.  Try for example
> [C-x C-f C-h c M-p].  Emacs will tell you that "M-p runs the command
> previous-history-element".  It looks ok to me.

Good.