w32-enable-(caps|num)-lock vs typematic keyboard

w32-enable-(caps|num)-lock vs typematic keyboard

[Juanma Barranquero]

With `w32-enable-caps-lock' or `w32-enable-num-lock' set to nil, if
you keep the key down on a typematic keyboard the caps-lock or
num-lock state can still change (depending, I suppose, on the number
of keyboard events received).

Perhaps not easy to fix, but should be documented in etc/PROBLEMS IMHO.

                    /L/e/k/t/u

Re: w32-enable-(caps|num)-lock vs typematic keyboard

[Eli Zaretskii]

> Date: Wed, 29 Nov 2006 18:17:18 +0100
> From: "Juanma Barranquero" <lekktu@gmail.com>
> 
> With `w32-enable-caps-lock' or `w32-enable-num-lock' set to nil, if
> you keep the key down on a typematic keyboard the caps-lock or
> num-lock state can still change (depending, I suppose, on the number
> of keyboard events received).

Please provide a recipe to reproduce this.  In a short and naive test,
I couldn't.

Re: w32-enable-(caps|num)-lock vs typematic keyboard

[Juanma Barranquero]

On 11/29/06, Eli Zaretskii <eliz@gnu.org> wrote:

> Please provide a recipe to reproduce this.  In a short and naive test,
> I couldn't.

Just "M-: (setq w32-enable-caps-lock) ENTER nil ENTER" should work.

But... I cannot reproduce it on my laptop. However, it happens
reliably at work, with a clone PC and a Logitech keyboard.

                    /L/e/k/t/u

Re: w32-enable-(caps|num)-lock vs typematic keyboard

[Lennart Borgman]

Juanma Barranquero wrote:
> On 11/29/06, Eli Zaretskii <eliz@gnu.org> wrote:
> 
>> Please provide a recipe to reproduce this.  In a short and naive test,
>> I couldn't.
> 
> Just "M-: (setq w32-enable-caps-lock) ENTER nil ENTER" should work.
> 
> But... I cannot reproduce it on my laptop. However, it happens
> reliably at work, with a clone PC and a Logitech keyboard.


I just tested to add this to the low level keyboard hook that I use in 
the patched version. Seems to work better, no flashing of the caps look 
key.

I suggest that we add this after the release. It would be more in line 
with MS documentation I believe.

Re: w32-enable-(caps|num)-lock vs typematic keyboard

[Juanma Barranquero]

On 11/29/06, Juanma Barranquero <lekktu@gmail.com> wrote:

> Just "M-: (setq w32-enable-caps-lock) ENTER nil ENTER" should work.

Must. Sleep. More.

  M-: (setq w32-enable-caps-lock nil) ENTER

of course.

                    /L/e/k/t/u

Re: w32-enable-(caps|num)-lock vs typematic keyboard

[Eli Zaretskii]

> Date: Wed, 29 Nov 2006 20:30:32 +0100
> From: "Juanma Barranquero" <lekktu@gmail.com>
> Cc: emacs-devel@gnu.org
> 
> Just "M-: (setq w32-enable-caps-lock nil) ENTER" should work.

And then what?  You said:

  if you keep the key down on a typematic keyboard the caps-lock or
  num-lock state can still change

Does that mean that I press and hold the `d' key, for instance, and
after some time the CapsLock key goes on?  If so, why should repeating
a key change CapsLock state?  I never heard about such a ``feature''.

Re: w32-enable-(caps|num)-lock vs typematic keyboard

[Juanma Barranquero]

On 11/30/06, Eli Zaretskii <eliz@gnu.org> wrote:

> And then what?  You said:
>
>   if you keep the key down on a typematic keyboard the caps-lock or
>   num-lock state can still change
>
> Does that mean that I press and hold the `d' key, for instance, and
> after some time the CapsLock key goes on?

No. In my message "the key" tried to mean "they key affected by the
variable previously referred to".

In other words:

  ELISP> (setq w32-enable-caps-lock nil)
 nil

And now I press Caps Lock and keep it down: the caps lock light
flickers like mad. I release the key: the light flicks a few moments
more (while the typematic buffer lasts, I suppose), and then stops.
Sometimes on, sometimes off.

The same thing happens with `w32-enable-num-lock' and the Num Lock key.

                    /L/e/k/t/u

Re: w32-enable-(caps|num)-lock vs typematic keyboard

[Lennart Borgman]

Juanma Barranquero wrote:
> On 11/30/06, Eli Zaretskii <eliz@gnu.org> wrote:
> 
>> And then what?  You said:
>>
>>   if you keep the key down on a typematic keyboard the caps-lock or
>>   num-lock state can still change
>>
>> Does that mean that I press and hold the `d' key, for instance, and
>> after some time the CapsLock key goes on?
> 
> No. In my message "the key" tried to mean "they key affected by the
> variable previously referred to".
> 
> In other words:
> 
>  ELISP> (setq w32-enable-caps-lock nil)
> nil
> 
> And now I press Caps Lock and keep it down: the caps lock light
> flickers like mad. I release the key: the light flicks a few moments
> more (while the typematic buffer lasts, I suppose), and then stops.
> Sometimes on, sometimes off.
> 
> The same thing happens with `w32-enable-num-lock' and the Num Lock key.


And it works like a charm if you instead use a low level keyboard hook...

Re: w32-enable-(caps|num)-lock vs typematic keyboard

[Juanma Barranquero]

On 11/30/06, Lennart Borgman <lennart.borgman.073@student.lu.se> wrote:

> And it works like a charm if you instead use a low level keyboard hook...

Do you mean that without the low level hook you see the same effect as I do?

                    /L/e/k/t/u

Re: w32-enable-(caps|num)-lock vs typematic keyboard

[Lennart Borgman]

Juanma Barranquero wrote:
> On 11/30/06, Lennart Borgman <lennart.borgman.073@student.lu.se> wrote:
> 
>> And it works like a charm if you instead use a low level keyboard hook...
> 
> Do you mean that without the low level hook you see the same effect as I 
> do?


Without the hook I see the CAPS lock indicator blinking when I press or 
hold down the CAPS key.

Re: w32-enable-(caps|num)-lock vs typematic keyboard

[Juanma Barranquero]

On 11/30/06, Lennart Borgman <lennart.borgman.073@student.lu.se> wrote:

> Without the hook I see the CAPS lock indicator blinking when I press or
> hold down the CAPS key.

This is not a bug, but an artifact of the implementation. Not enough
reason to switch to a low-level hook.

Do you see the same bug I'm seeing, or not?

                    /L/e/k/t/u

Re: w32-enable-(caps|num)-lock vs typematic keyboard

[Lennart Borgman]

Juanma Barranquero wrote:
> On 11/30/06, Lennart Borgman <lennart.borgman.073@student.lu.se> wrote:
> 
>> Without the hook I see the CAPS lock indicator blinking when I press or
>> hold down the CAPS key.
> 
> This is not a bug, but an artifact of the implementation. Not enough
> reason to switch to a low-level hook.


Or the implementation is a bug. Is the result really guaranteed to be 
well defined when you try this witout a hook? Could there not be subtle 
differences between different drivers?


> Do you see the same bug I'm seeing, or not?


No, I have tried both with or without Sticky Keys. I have no keybinding 
on <capslock>, maybe that makes a difference?

Re: w32-enable-(caps|num)-lock vs typematic keyboard

[Juanma Barranquero]

On 11/30/06, Lennart Borgman <lennart.borgman.073@student.lu.se> wrote:

> Or the implementation is a bug. Is the result really guaranteed to be
> well defined when you try this witout a hook? Could there not be subtle
> differences between different drivers?

It's quite evident that could be subtle (or not so subtle)
differences, for the simple reason that I'm *already* reporting a bug.
But it is not serious and apparently I'm the only one able to
reproduce it. Not enough reason to switch to a low-level hook, and
even less right at this moment.

> No, I have tried both with or without Sticky Keys. I have no keybinding
> on <capslock>, maybe that makes a difference?

I don't use Sticky Keys. And I have no binding in <capslock>.

                    /L/e/k/t/u

Re: w32-enable-(caps|num)-lock vs typematic keyboard

[Eli Zaretskii]

> Date: Thu, 30 Nov 2006 12:34:20 +0100
> From: "Juanma Barranquero" <lekktu@gmail.com>
> Cc: emacs-devel@gnu.org
> 
>   ELISP> (setq w32-enable-caps-lock nil)
>  nil
> 
> And now I press Caps Lock and keep it down: the caps lock light
> flickers like mad. I release the key: the light flicks a few moments
> more (while the typematic buffer lasts, I suppose), and then stops.
> Sometimes on, sometimes off.

On my machine, I see the flickering, but the light is always off when
I release the key.

How about adding a short comment about this to the variable's doc
string?

Re: w32-enable-(caps|num)-lock vs typematic keyboard

[Juanma Barranquero]

On 11/30/06, Eli Zaretskii <eliz@gnu.org> wrote:

> On my machine, I see the flickering, but the light is always off when
> I release the key.

That's what happens on my laptop.

> How about adding a short comment about this to the variable's doc
> string?

Saying what? "Note: in some keyboards it only works half the time."? :)

                    /L/e/k/t/u

Re: w32-enable-(caps|num)-lock vs typematic keyboard

[Eli Zaretskii]

> Date: Thu, 30 Nov 2006 21:19:54 +0100
> From: "Juanma Barranquero" <lekktu@gmail.com>
> Cc: emacs-devel@gnu.org
> 
> > How about adding a short comment about this to the variable's doc
> > string?
> 
> Saying what? "Note: in some keyboards it only works half the time."? :)

Saying that it flickers when pressed, but returns to off when
released.  I have no idea why it doesn't work for you, but as long as
we don't know the reason, we cannot say anything intelligent about
that.

Re: w32-enable-(caps|num)-lock vs typematic keyboard

[Juanma Barranquero]

On 11/30/06, Eli Zaretskii <eliz@gnu.org> wrote:

> Saying that it flickers when pressed, but returns to off when
> released.

Documenting the flicker is a waste of time IMHO. I don't think is very
common to set either variable to nil, and even less common to keep any
of these keys pressed down (they can not be set as modifier keys).

                    /L/e/k/t/u