AltGr finger twisters documented?

AltGr finger twisters documented?

[Juanma Barranquero]

After an exchange of comments on the Emacs wiki, that left me wondering...

Is there any place in the Emacs documentation that explains that you
*can* type C-M-@, C-@, C-M-\, C-\ and C-] on an AltGr-challenged
keyboard, *provided* that you use the right order, i.e., AltGr, right
Ctrl, and then the modified key?

For example, on my Spanish keyboard:

   C-M-@ => AltGr RCtrl Alt 2
   C-@   => AltGr RCtrl 2
   C-M-\ => AltGr RCtrl Alt º
   C-\   => AltGr RCtrl º
   C-]   => AltGr RCtrl +

It is non-obvious: it doesn't work if you press other key before
AltGr, or if you use left control.

-- 
                    /L/e/k/t/u

Re: AltGr finger twisters documented?

[Kim F. Storm]

Juanma Barranquero <lekktu@gmail.com> writes:

> After an exchange of comments on the Emacs wiki, that left me wondering...
>
> Is there any place in the Emacs documentation that explains that you
> *can* type C-M-@, C-@, C-M-\, C-\ and C-] on an AltGr-challenged
> keyboard, *provided* that you use the right order, i.e., AltGr, right
> Ctrl, and then the modified key?
>
> For example, on my Spanish keyboard:
>
>    C-M-@ => AltGr RCtrl Alt 2
>    C-@   => AltGr RCtrl 2
>    C-M-\ => AltGr RCtrl Alt º
>    C-\   => AltGr RCtrl º
>    C-]   => AltGr RCtrl +
>
> It is non-obvious: it doesn't work if you press other key before
> AltGr, or if you use left control.

Actually, on my Danish PC keyboard, C-\ can be entered just fine
using either of

       LCtrl AltGr <
       AltGr LCtrl <


My setup:

In GNU Emacs 22.0.50.127 (i686-pc-linux-gnu, X toolkit, Xaw3d scroll bars)
 of 2005-07-04 on kfs-l.imdomain.dk
X server distributor `The XFree86 Project, Inc', version 11.0.40300000
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: en_US.UTF-8
  locale-coding-system: utf-8
  default-enable-multibyte-characters: t


--
Kim F. Storm <storm@cua.dk> http://www.cua.dk

Re: AltGr finger twisters documented?

[Jan D.]

> After an exchange of comments on the Emacs wiki, that left me  
> wondering...
>
> Is there any place in the Emacs documentation that explains that you
> *can* type C-M-@, C-@, C-M-\, C-\ and C-] on an AltGr-challenged
> keyboard, *provided* that you use the right order, i.e., AltGr, right
> Ctrl, and then the modified key?
>
> For example, on my Spanish keyboard:
>
>    C-M-@ => AltGr RCtrl Alt 2
>    C-@   => AltGr RCtrl 2
>    C-M-\ => AltGr RCtrl Alt º
>    C-\   => AltGr RCtrl º
>    C-]   => AltGr RCtrl +
>
> It is non-obvious: it doesn't work if you press other key before
> AltGr, or if you use left control.


I suspect this is keyboard dependent.  I do C-\ all the time with  
both right and left control, and I press Ctrl-AltGr-+ (hold down all  
three keys at the same time, which I press first or last does not  
matter) on my swedish keyboard (and add Alt to that, i.e. Alt-Ctrl- 
AltGr-+, gets me M-C-\, but it is not very ergonomic).  The trick  
above does not work for me at all.

     Jan D.

Re: AltGr finger twisters documented?

[LENNART BORGMAN]

----- Original Message -----
From: Juanma Barranquero <lekktu@gmail.com>
Date: Tuesday, July 5, 2005 11:34 am
Subject: AltGr finger twisters documented?

> After an exchange of comments on the Emacs wiki, that left me 
> wondering...
> Is there any place in the Emacs documentation that explains that you
> *can* type C-M-@, C-@, C-M-\, C-\ and C-] on an AltGr-challenged
> keyboard, *provided* that you use the right order, i.e., AltGr, right
> Ctrl, and then the modified key?
> 
> For example, on my Spanish keyboard:
> 
>   C-M-@ => AltGr RCtrl Alt 2
>   C-@   => AltGr RCtrl 2
>   C-M-\ => AltGr RCtrl Alt º
>   C-\   => AltGr RCtrl º
>   C-]   => AltGr RCtrl +
> 
> It is non-obvious: it doesn't work if you press other key before
> AltGr, or if you use left control.

Wow, I had not the slightest idea! And I have complained and asked about issues related to this ... - maybe I was to unclear? I looked (again) in (emacs) Info but could not find much about AltGr.

Re: AltGr finger twisters documented?

[LENNART BORGMAN]

From: storm@cua.dk (Kim F. Storm)

> Juanma Barranquero <lekktu@gmail.com> writes:
> 
> > After an exchange of comments on the Emacs wiki, that left me 
> wondering...>
> > Is there any place in the Emacs documentation that explains that you
> > *can* type C-M-@, C-@, C-M-\, C-\ and C-] on an AltGr-challenged
> > keyboard, *provided* that you use the right order, i.e., AltGr, 
> right> Ctrl, and then the modified key?
> >
> > For example, on my Spanish keyboard:
> >
> >    C-M-@ => AltGr RCtrl Alt 2
> >    C-@   => AltGr RCtrl 2
> >    C-M-\ => AltGr RCtrl Alt º
> >    C-\   => AltGr RCtrl º
> >    C-]   => AltGr RCtrl +
> >
> > It is non-obvious: it doesn't work if you press other key before
> > AltGr, or if you use left control.
> 
> Actually, on my Danish PC keyboard, C-\ can be entered just fine
> using either of
> 
>       LCtrl AltGr <
>       AltGr LCtrl <
> 
> 
> My setup:
> 
> In GNU Emacs 22.0.50.127 (i686-pc-linux-gnu, X toolkit, Xaw3d 
> scroll bars)
> of 2005-07-04 on kfs-l.imdomain.dk
> X server distributor `The XFree86 Project, Inc', version 11.0.40300000
> 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: en_US.UTF-8
>  locale-coding-system: utf-8
>  default-enable-multibyte-characters: t

Interesting, with this setup I must follow Juanmas advice:

In GNU Emacs 22.0.50.1 (i386-mingw-nt5.0.2195)
 of 2005-07-04 on W2ONE
X server distributor `Microsoft Corp.', version 5.0.2195
configured using `configure --with-gcc (3.2) --cflags -Id:/g/include'

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: SVE
  locale-coding-system: cp1252
  default-enable-multibyte-characters: t

What do I do to test your setup?

Re: AltGr finger twisters documented?

[Juanma Barranquero]

On 7/5/05, LENNART BORGMAN <lennart.borgman.073@student.lu.se> wrote:

> Wow, I had not the slightest idea!

Does it work for you? Are you using Windows ?(I think the answer to
that last is YES :)

> And I have complained and asked about issues related to this ... - maybe > I was to unclear?

Maybe I didn't see your messages, because I've known for a long time
that I could only get C-] and friends using the right control key.

-- 
                    /L/e/k/t/u

Re: AltGr finger twisters documented?

[Mathias Dahl]

"Jan D." <jan.h.d@swipnet.se> writes:

> I suspect this is keyboard dependent.  I do C-\ all the time with
> both right and left control, and I press Ctrl-AltGr-+ (hold down all
> three keys at the same time, which I press first or last does not
> matter) on my swedish keyboard (and add Alt to that, i.e. Alt-Ctrl-
> AltGr-+, gets me M-C-\, but it is not very ergonomic).  The trick
> above does not work for me at all.

Indeed it seems to be. LCtrl-AltGr-+, in that order, produces \ on my
Key Tronic keyboard, running Emacs 22.0.50.2 on Windows
XP. LAlt-LCtrl-AltGr-x does not work at all for me... :)

Re: AltGr finger twisters documented?

[Jason Rumney]

Juanma Barranquero wrote:

>Maybe I didn't see your messages, because I've known for a long time
>that I could only get C-] and friends using the right control key.
>  
>
Does the variable w32-recognize-altgr affect this?

I am not sure whether setting that variable to nil will prevent real 
AltGr keys working, or just the simulated version for keyboards without 
an AltGr key. If the latter, then I wonder if t is the appropriate 
default for this variable.

Re: AltGr finger twisters documented?

[Lennart Borgman]

Jan D. wrote:

>> After an exchange of comments on the Emacs wiki, that left me  
>> wondering...
>>
>> Is there any place in the Emacs documentation that explains that you
>> *can* type C-M-@, C-@, C-M-\, C-\ and C-] on an AltGr-challenged
>> keyboard, *provided* that you use the right order, i.e., AltGr, right
>> Ctrl, and then the modified key?
>>
>> For example, on my Spanish keyboard:
>>
>>    C-M-@ => AltGr RCtrl Alt 2
>>    C-@   => AltGr RCtrl 2
>>    C-M-\ => AltGr RCtrl Alt º
>>    C-\   => AltGr RCtrl º
>>    C-]   => AltGr RCtrl +
>>
>> It is non-obvious: it doesn't work if you press other key before
>> AltGr, or if you use left control.
>
>
>
> I suspect this is keyboard dependent.  I do C-\ all the time with  
> both right and left control, and I press Ctrl-AltGr-+ (hold down all  
> three keys at the same time, which I press first or last does not  
> matter) on my swedish keyboard (and add Alt to that, i.e. Alt-Ctrl- 
> AltGr-+, gets me M-C-\, but it is not very ergonomic).  The trick  
> above does not work for me at all. 

What OS I you using? Which version of Emacs?

Re: AltGr finger twisters documented?

[Lennart Borgman]

Juanma Barranquero wrote:

>On 7/5/05, LENNART BORGMAN <lennart.borgman.073@student.lu.se> wrote:
>
>  
>
>>Wow, I had not the slightest idea!
>>    
>>
>
>Does it work for you? Are you using Windows ?(I think the answer to
>that last is YES :)
>  
>
Yes, it works. And, yes w2k. I have tested three keyboards now.

Re: AltGr finger twisters documented?

[Lennart Borgman]

Jan D. wrote:

>
> I suspect this is keyboard dependent.  I do C-\ all the time with  
> both right and left control, and I press Ctrl-AltGr-+ (hold down all  
> three keys at the same time, which I press first or last does not  
> matter) on my swedish keyboard (and add Alt to that, i.e. Alt-Ctrl- 
> AltGr-+, gets me M-C-\, but it is not very ergonomic).  The trick  
> above does not work for me at all.

I forgot to ask what keyboard layout you are using. Swedish? But in that 
case is not "\" on "AlGr-+"?

Re: AltGr finger twisters documented?

[Lennart Borgman]

Jason Rumney wrote:

> Juanma Barranquero wrote:
>
>> Maybe I didn't see your messages, because I've known for a long time
>> that I could only get C-] and friends using the right control key.
>>  
>>
> Does the variable w32-recognize-altgr affect this?
>
> I am not sure whether setting that variable to nil will prevent real 
> AltGr keys working, or just the simulated version for keyboards 
> without an AltGr key. If the latter, then I wonder if t is the 
> appropriate default for this variable. 

Yes, it surely changes things. I can for example not type \ at all if I 
set this to nil. So setting this to nil gives me a completely new 
keyboard layout which I do not think is useful at all.

Re: AltGr finger twisters documented?

[Juanma Barranquero]

> Yes, it surely changes things. I can for example not type \ at all if I
> set this to nil. So setting this to nil gives me a completely new
> keyboard layout which I do not think is useful at all.

Exactly. With `w32-recognize-altgr' set to nil there's no way I can
type C-\, because there's no way I can type \ directly.

-- 
                    /L/e/k/t/u

Re: AltGr finger twisters documented?

[Juanma Barranquero]

[-- Attachment #1: Type: text/plain, Size: 1076 bytes --]

On 7/5/05, Mathias Dahl <brakjoller@gmail.com> wrote:
> "Jan D." <jan.h.d@swipnet.se> writes:

> Indeed it seems to be. LCtrl-AltGr-+, in that order, produces \ on my
> Key Tronic keyboard, running Emacs 22.0.50.2 on Windows
> XP. LAlt-LCtrl-AltGr-x does not work at all for me... :)

Hmmm. My previous message was wrong. With `w32-recognize-altgr' I
*can* type \, ], etc., for example LCtr-AltGr-+ produces ].

So it seems like, at least on Windows,

  - with w32-recognize-altgr = nil
   LCtrl + AltGr + key produce uncommon chars: []\|@#{}€ (in Spanish kbd)
   There's no way to produce Ctrl/Meta versions of them

  - with w32-recognize-altgr = t
   AltGr + key produce uncommon chars
   AltGr + RCtrl + key produce Ctrl/Meta versions of them
   
I think there's no doubt that w32-recognize-altgr = t is needed, at
least on some language environments. Emacs would be unusable to me if
I had to do LCtrl-AltGr-` every time I wanted to type a [.

So perhaps we should simply document all this in some W32-specific
node of the info files.

-- 
                    /L/e/k/t/u

[-- Attachment #2: Type: text/plain, Size: 142 bytes --]

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

Re: AltGr finger twisters documented?

[Jan D.]

>>> After an exchange of comments on the Emacs wiki, that left me   
>>> wondering...
>>>
>>> Is there any place in the Emacs documentation that explains that you
>>> *can* type C-M-@, C-@, C-M-\, C-\ and C-] on an AltGr-challenged
>>> keyboard, *provided* that you use the right order, i.e., AltGr,  
>>> right
>>> Ctrl, and then the modified key?
>>>
>>> For example, on my Spanish keyboard:
>>>
>>>    C-M-@ => AltGr RCtrl Alt 2
>>>    C-@   => AltGr RCtrl 2
>>>    C-M-\ => AltGr RCtrl Alt º
>>>    C-\   => AltGr RCtrl º
>>>    C-]   => AltGr RCtrl +
>>>
>>> It is non-obvious: it doesn't work if you press other key before
>>> AltGr, or if you use left control.
>>>
>>
>>
>>
>> I suspect this is keyboard dependent.  I do C-\ all the time with   
>> both right and left control, and I press Ctrl-AltGr-+ (hold down  
>> all  three keys at the same time, which I press first or last does  
>> not  matter) on my swedish keyboard (and add Alt to that, i.e. Alt- 
>> Ctrl- AltGr-+, gets me M-C-\, but it is not very ergonomic).  The  
>> trick  above does not work for me at all.
>>
>
> What OS I you using? Which version of Emacs?

Well, GNU/Linux in many flavors, FreeBSD, Solaris, MacOSX (but there  
I have no idea how to get C-\).  A very recent (a few days old) CVS  
Emacs.

     Jan D.

Re: AltGr finger twisters documented?

[Jan D.]

>> I suspect this is keyboard dependent.  I do C-\ all the time with   
>> both right and left control, and I press Ctrl-AltGr-+ (hold down  
>> all  three keys at the same time, which I press first or last does  
>> not  matter) on my swedish keyboard (and add Alt to that, i.e. Alt- 
>> Ctrl- AltGr-+, gets me M-C-\, but it is not very ergonomic).  The  
>> trick  above does not work for me at all.
>>
>
> I forgot to ask what keyboard layout you are using. Swedish? But in  
> that case is not "\" on "AlGr-+"?

Isn't that what I wrote?  Ctrl-AltGr-+ in the text above.  I may have  
misunderstood, but I took it to mean he pressed and released AltGr,  
then pressed and released Ctrl, and then pressed and released another  
key.  That does not work.

     Jan D

Re: AltGr finger twisters documented?

[Juanma Barranquero]

On 7/5/05, Jan D. <jan.h.d@swipnet.se> wrote:

> Well, GNU/Linux in many flavors, FreeBSD, Solaris, MacOSX (but there
> I have no idea how to get C-\).  A very recent (a few days old) CVS
> Emacs.

It seems clearer and clearer than it is an issue related to AltGr
management by Windows (and Emacs).


-- 
                    /L/e/k/t/u

Re: AltGr finger twisters documented?

[Juanma Barranquero]

On 7/5/05, Jan D. <jan.h.d@swipnet.se> wrote:

> I may have
> misunderstood, but I took it to mean he pressed and released AltGr,
> then pressed and released Ctrl, and then pressed and released another
> key.  That does not work.

No, no.

Pressing and releasing AltGr does not work. You must have all keys
pressed, but the order in which you do press them is important. So, to
me:

 Press AltGr
 Press RCtrl
 Press +

produces C-], while

 Press RCtrl
 Press AltGr
 Press +

produces C-M-+

-- 
                    /L/e/k/t/u

Re: AltGr finger twisters documented?

[Jan D.]

>
> No, no.
>
> Pressing and releasing AltGr does not work. You must have all keys
> pressed, but the order in which you do press them is important. So, to
> me:
>
>  Press AltGr
>  Press RCtrl
>  Press +
>
> produces C-], while
>
>  Press RCtrl
>  Press AltGr
>  Press +
>
> produces C-M-+

On non-Windows OS they both produce C-], as does LCtrl.

     Jan D.

Re: AltGr finger twisters documented?

[Lennart Borgman]

Jan D. wrote:

>>
>> No, no.
>>
>> Pressing and releasing AltGr does not work. You must have all keys
>> pressed, but the order in which you do press them is important. So, to
>> me:
>>
>>  Press AltGr
>>  Press RCtrl
>>  Press +
>>
>> produces C-], while
>>
>>  Press RCtrl
>>  Press AltGr
>>  Press +
>>
>> produces C-M-+
>
>
> On non-Windows OS they both produce C-], as does LCtrl.
>
>     Jan D. 

Thanks. I then consider the behaviour on w32 to be a bug (if there is no 
intention with this behaviour of course). What would other think of this?

Re: AltGr finger twisters documented?

[Juanma Barranquero]

On 7/5/05, Lennart Borgman <lennart.borgman.073@student.lu.se> wrote:

> Thanks. I then consider the behaviour on w32 to be a bug (if there is no
> intention with this behaviour of course). What would other think of this?

I do not consider it a bug. Every necessary key combination can be
typed. It is weird, but do you know for sure it is not an artifact of
Windows keyboard handling? Perhaps changing the behavior would require
major efforts.

-- 
                    /L/e/k/t/u

Re: AltGr finger twisters documented?

[Jason Rumney]

Lennart Borgman wrote:

> Thanks. I then consider the behaviour on w32 to be a bug (if there is 
> no intention with this behaviour of course). What would other think of 
> this?

Sure it's a bug, but a bug of the OS. Setting w32-recognize-altgr to nil 
appears to avoid this bug for people who don't need AltGr at all.

The bug is that Windows does not have a separate modifier flag for 
AltGr, it uses Left-Ctrl + Right-Alt flags to indicate that AltGr is 
being pressed. So we can't tell that Left-Ctrl is pressed when AltGr is 
down.

If you can find a proper fix for this bug, we can probably get rid of 
w32-recognize-altgr.

Re: AltGr finger twisters documented?

[Juanma Barranquero]

On 7/5/05, Jason Rumney <jasonr@gnu.org> wrote:

> Sure it's a bug, but a bug of the OS. Setting w32-recognize-altgr to nil
> appears to avoid this bug for people who don't need AltGr at all.

I know you're somewhat tongue-in-cheek, but I agree with you. There's
nothing to change here (IMO), but it would be good to document the
issue, don't you agree?

-- 
                    /L/e/k/t/u

Re: AltGr finger twisters documented?

[Lennart Borgman]

Jason Rumney wrote:

> Lennart Borgman wrote:
>
>> Thanks. I then consider the behaviour on w32 to be a bug (if there is 
>> no intention with this behaviour of course). What would other think 
>> of this?
>
>
> Sure it's a bug, but a bug of the OS. Setting w32-recognize-altgr to 
> nil appears to avoid this bug for people who don't need AltGr at all.
>
> The bug is that Windows does not have a separate modifier flag for 
> AltGr, it uses Left-Ctrl + Right-Alt flags to indicate that AltGr is 
> being pressed. So we can't tell that Left-Ctrl is pressed when AltGr 
> is down.
>
> If you can find a proper fix for this bug, we can probably get rid of 
> w32-recognize-altgr.

Have you thought about the possibility to use a low level keyboard hook?

I thought of this as a possibility for leaving the Alt key to windows 
and still have a comfortable Meta key. <lwindow> and <rwindow> could be 
used for that I believe but you must then use a low level keyboard hook. 
The current approach with w32-pass-lwindow-to-system does not work as 
far as I can see. If do

   (setq w32-pass-alt-to-system nil), and then C-h c, <lwindow>, e

I get a Windows Explorer w32 window. Using a low level keyboard hook 
this can be trapped.

However I do not know much about them. Can they be added if the user 
have no privilege? As far as I understand it ought to be decided when 
Emacs has keyboard focus and then it does not have it so that the 
keyboard hook can be added and removed at the right moments. Can 
WM_APPACTIVATE be used for this (or was there some other message)?

Otherwise there does not seem to be needed much code for the keyboard 
hook itself and adding and removing it. And I guess the rest of the code 
already is in Emacs?

Re: AltGr finger twisters documented?

[Lennart Borgman]

Lennart Borgman wrote:

>
> However I do not know much about them. Can they be added if the user 
> have no privilege? As far as I understand it ought to be decided when 
> Emacs has keyboard focus and then it does not have it so that the 
> keyboard hook can be added and removed at the right moments. Can 
> WM_APPACTIVATE be used for this (or was there some other message)?

Seems like a low level keyboard hook can be added to a specific thread 
so I guess there are no problems with privs.

Re: AltGr finger twisters documented?

[Jason Rumney]

Juanma Barranquero <lekktu@gmail.com> writes:

> On 7/5/05, Jason Rumney <jasonr@gnu.org> wrote:
>
>> Sure it's a bug, but a bug of the OS. Setting w32-recognize-altgr to nil
>> appears to avoid this bug for people who don't need AltGr at all.
>
> I know you're somewhat tongue-in-cheek, but I agree with you. There's
> nothing to change here (IMO), but it would be good to document the
> issue, don't you agree?

w32-recognize-altgr could do with better documentation. Its affect on
AltGr was not apparent to me from reading the doc string. There is
already an entry in etc/PROBLEMS relating to the use of left Ctrl and
right Alt together. That could perhaps be expanded to include an
explanation of how AltGr is affected.

Re: AltGr finger twisters documented?

[Juanma Barranquero]

On 7/5/05, Jason Rumney <jasonr@gnu.org> wrote:
> Juanma Barranquero <lekktu@gmail.com> writes:

> w32-recognize-altgr could do with better documentation. Its affect on
> AltGr was not apparent to me from reading the doc string. There is
> already an entry in etc/PROBLEMS relating to the use of left Ctrl and
> right Alt together. That could perhaps be expanded to include an
> explanation of how AltGr is affected.

I can try to expand it (although I imagine Richard or someone will
then indulge in a rewording frenzy to turn it into Emacs English, as
opposed to Juanma's Spanglish), but, it is etc/PROBLEMS the right
place if we agree that it is not a bug, but the way it works? (Which
is my opinion since the start, BTW).

-- 
                    /L/e/k/t/u

Re: AltGr finger twisters documented?

[Jason Rumney]

Juanma Barranquero <lekktu@gmail.com> writes:

> I can try to expand it (although I imagine Richard or someone will
> then indulge in a rewording frenzy to turn it into Emacs English, as
> opposed to Juanma's Spanglish), but, it is etc/PROBLEMS the right
> place if we agree that it is not a bug, but the way it works? (Which
> is my opinion since the start, BTW).

Yes. It is a known problem, and PROBLEMS is the first place to look
when you have problems.

Re: AltGr finger twisters documented?

[Eli Zaretskii]

> Date: Tue, 5 Jul 2005 11:34:13 +0200
> From: Juanma Barranquero <lekktu@gmail.com>
> 
> Is there any place in the Emacs documentation that explains that you
> *can* type C-M-@, C-@, C-M-\, C-\ and C-] on an AltGr-challenged
> keyboard, *provided* that you use the right order, i.e., AltGr, right
> Ctrl, and then the modified key?

As you already found out, this is specific to the MS-Windows port.

Let me reiterate what I already said a week or two ago: the Windows
specific aspects of Emacs are almost entirely undocumented.  This is
one of them.

It would be nice if someone started to write the MS-Windows appendix
to the User's Manual, IMHO.

Re: AltGr finger twisters documented?

[YAMAMOTO Mitsuharu]

>>>>> On Tue, 5 Jul 2005 16:00:05 +0200, "Jan D." <jan.h.d@swipnet.se> said:

> Well, GNU/Linux in many flavors, FreeBSD, Solaris, MacOSX (but there
> I have no idea how to get C-\).

As for Mac OS X/Carbon, could you test if you can type C-\ with "Ctrl
+ [the key combination for '\']" with the following patch?

				     YAMAMOTO Mitsuharu
				mituharu@math.s.chiba-u.ac.jp


Index: src/macterm.c
===================================================================
RCS file: /cvsroot/emacs/emacs/src/macterm.c,v
retrieving revision 1.120
diff -c -r1.120 macterm.c
*** src/macterm.c	4 Jul 2005 16:06:33 -0000	1.120
--- src/macterm.c	6 Jul 2005 00:13:57 -0000
***************
*** 93,99 ****
  #define macMetaKey     (NILP (Vmac_reverse_ctrl_meta) ?			\
  			(NILP (Vmac_command_key_is_meta) ? optionKey : cmdKey) \
  			: controlKey)
! #define macAltKey      (NILP (Vmac_command_key_is_meta) ? cmdKey : optionKey)
  
  #define mac_window_to_frame(wp) (((mac_output *) GetWRefCon (wp))->mFP)
  \f
--- 93,101 ----
  #define macMetaKey     (NILP (Vmac_reverse_ctrl_meta) ?			\
  			(NILP (Vmac_command_key_is_meta) ? optionKey : cmdKey) \
  			: controlKey)
! #define macAltKey      (NILP (Vmac_command_key_is_meta) ? \
! 			(NILP (Vmac_command_key_is_meta) ? cmdKey : optionKey) \
! 			: 0)
  
  #define mac_window_to_frame(wp) (((mac_output *) GetWRefCon (wp))->mFP)
  \f
***************
*** 7537,7543 ****
      result |= ctrl_modifier;
    if (mods & macMetaKey)
      result |= meta_modifier;
!   if (NILP (Vmac_command_key_is_meta) && (mods & macAltKey))
      result |= alt_modifier;
    if (!NILP (Vmac_option_modifier) && (mods & optionKey)) {
        Lisp_Object val = Fget(Vmac_option_modifier, Qmodifier_value);
--- 7539,7545 ----
      result |= ctrl_modifier;
    if (mods & macMetaKey)
      result |= meta_modifier;
!   if (mods & macAltKey)
      result |= alt_modifier;
    if (!NILP (Vmac_option_modifier) && (mods & optionKey)) {
        Lisp_Object val = Fget(Vmac_option_modifier, Qmodifier_value);
***************
*** 9479,9487 ****
  	      }
  	    else
  	      {
! 		if (er.modifiers & (controlKey |
! 				    (NILP (Vmac_command_key_is_meta) ? optionKey
! 				     : cmdKey)))
  		  {
  		    /* This code comes from Keyboard Resource,
  		       Appendix C of IM - Text.  This is necessary
--- 9481,9492 ----
  	      }
  	    else
  	      {
! 		EventModifiers mask = macCtrlKey | macMetaKey | macAltKey;
! 
! 		if (!NILP (Vmac_option_modifier))
! 		  mask |= optionKey;
! 
! 		if (er.modifiers & mask)
  		  {
  		    /* This code comes from Keyboard Resource,
  		       Appendix C of IM - Text.  This is necessary
***************
*** 9490,9513 ****
  		       It also does not translate correctly
  		       control-shift chars like C-% so mask off shift
  		       here also */
! 		    int new_modifiers = er.modifiers & 0xe600;
! 		    /* mask off option and command */
! 		    int new_keycode = keycode | new_modifiers;
! 		    Ptr kchr_ptr = (Ptr) GetScriptManagerVariable (smKCHRCache);
! 		    unsigned long some_state = 0;
! 		    inev.code = KeyTranslate (kchr_ptr, new_keycode,
! 					      &some_state) & 0xff;
! 		  }
! 		else if (!NILP (Vmac_option_modifier)
! 			 && (er.modifiers & optionKey))
! 		  {
! 		    /* When using the option key as an emacs modifier,
! 		       convert the pressed key code back to one
! 		       without the Mac option modifier applied. */
! 		    int new_modifiers = er.modifiers & ~optionKey;
! 		    int new_keycode = keycode | new_modifiers;
  		    Ptr kchr_ptr = (Ptr) GetScriptManagerVariable (smKCHRCache);
! 		    unsigned long some_state = 0;
  		    inev.code = KeyTranslate (kchr_ptr, new_keycode,
  					      &some_state) & 0xff;
  		  }
--- 9495,9506 ----
  		       It also does not translate correctly
  		       control-shift chars like C-% so mask off shift
  		       here also */
! 		    EventModifiers new_modifiers = er.modifiers & ~mask;
! 		    /* mask off modifiers */
! 		    UInt16 new_keycode = keycode | new_modifiers & 0xff00;
  		    Ptr kchr_ptr = (Ptr) GetScriptManagerVariable (smKCHRCache);
! 		    UInt32 some_state = 0;
! 
  		    inev.code = KeyTranslate (kchr_ptr, new_keycode,
  					      &some_state) & 0xff;
  		  }

Re: AltGr finger twisters documented?

[Jan D.]

>> Well, GNU/Linux in many flavors, FreeBSD, Solaris, MacOSX (but there
>> I have no idea how to get C-\).
>>
>
> As for Mac OS X/Carbon, could you test if you can type C-\ with "Ctrl
> + [the key combination for '\']" with the following patch?
>
>

It seems to work fine on initial testing.  I'll run Emacs with this  
patch for a while.  Actually \ is on option-shift-7 so C-\ becomes  
option-shift-ctrl-7, and M-C-\ is command-option-shift-ctrl-7.  Not  
so bad as it looks, because option, ctrl, shift and command are all  
close to each other.

     Jan D.


> Index: src/macterm.c
> ===================================================================
> RCS file: /cvsroot/emacs/emacs/src/macterm.c,v
> retrieving revision 1.120
> diff -c -r1.120 macterm.c
> *** src/macterm.c    4 Jul 2005 16:06:33 -0000    1.120
> --- src/macterm.c    6 Jul 2005 00:13:57 -0000
> ***************
> *** 93,99 ****
>   #define macMetaKey     (NILP (Vmac_reverse_ctrl_meta) ?            \
>               (NILP (Vmac_command_key_is_meta) ? optionKey : cmdKey) \
>               : controlKey)
> ! #define macAltKey      (NILP (Vmac_command_key_is_meta) ?  
> cmdKey : optionKey)
>
>   #define mac_window_to_frame(wp) (((mac_output *) GetWRefCon (wp))- 
> >mFP)
>   \f
> --- 93,101 ----
>   #define macMetaKey     (NILP (Vmac_reverse_ctrl_meta) ?            \
>               (NILP (Vmac_command_key_is_meta) ? optionKey : cmdKey) \
>               : controlKey)
> ! #define macAltKey      (NILP (Vmac_command_key_is_meta) ? \
> !             (NILP (Vmac_command_key_is_meta) ? cmdKey : optionKey) \
> !             : 0)
>
>   #define mac_window_to_frame(wp) (((mac_output *) GetWRefCon (wp))- 
> >mFP)
>   \f
> ***************
> *** 7537,7543 ****
>       result |= ctrl_modifier;
>     if (mods & macMetaKey)
>       result |= meta_modifier;
> !   if (NILP (Vmac_command_key_is_meta) && (mods & macAltKey))
>       result |= alt_modifier;
>     if (!NILP (Vmac_option_modifier) && (mods & optionKey)) {
>         Lisp_Object val = Fget(Vmac_option_modifier, Qmodifier_value);
> --- 7539,7545 ----
>       result |= ctrl_modifier;
>     if (mods & macMetaKey)
>       result |= meta_modifier;
> !   if (mods & macAltKey)
>       result |= alt_modifier;
>     if (!NILP (Vmac_option_modifier) && (mods & optionKey)) {
>         Lisp_Object val = Fget(Vmac_option_modifier, Qmodifier_value);
> ***************
> *** 9479,9487 ****
>             }
>           else
>             {
> !         if (er.modifiers & (controlKey |
> !                     (NILP (Vmac_command_key_is_meta) ? optionKey
> !                      : cmdKey)))
>             {
>               /* This code comes from Keyboard Resource,
>                  Appendix C of IM - Text.  This is necessary
> --- 9481,9492 ----
>             }
>           else
>             {
> !         EventModifiers mask = macCtrlKey | macMetaKey | macAltKey;
> !
> !         if (!NILP (Vmac_option_modifier))
> !           mask |= optionKey;
> !
> !         if (er.modifiers & mask)
>             {
>               /* This code comes from Keyboard Resource,
>                  Appendix C of IM - Text.  This is necessary
> ***************
> *** 9490,9513 ****
>                  It also does not translate correctly
>                  control-shift chars like C-% so mask off shift
>                  here also */
> !             int new_modifiers = er.modifiers & 0xe600;
> !             /* mask off option and command */
> !             int new_keycode = keycode | new_modifiers;
> !             Ptr kchr_ptr = (Ptr) GetScriptManagerVariable  
> (smKCHRCache);
> !             unsigned long some_state = 0;
> !             inev.code = KeyTranslate (kchr_ptr, new_keycode,
> !                           &some_state) & 0xff;
> !           }
> !         else if (!NILP (Vmac_option_modifier)
> !              && (er.modifiers & optionKey))
> !           {
> !             /* When using the option key as an emacs modifier,
> !                convert the pressed key code back to one
> !                without the Mac option modifier applied. */
> !             int new_modifiers = er.modifiers & ~optionKey;
> !             int new_keycode = keycode | new_modifiers;
>               Ptr kchr_ptr = (Ptr) GetScriptManagerVariable  
> (smKCHRCache);
> !             unsigned long some_state = 0;
>               inev.code = KeyTranslate (kchr_ptr, new_keycode,
>                             &some_state) & 0xff;
>             }
> --- 9495,9506 ----
>                  It also does not translate correctly
>                  control-shift chars like C-% so mask off shift
>                  here also */
> !             EventModifiers new_modifiers = er.modifiers & ~mask;
> !             /* mask off modifiers */
> !             UInt16 new_keycode = keycode | new_modifiers & 0xff00;
>               Ptr kchr_ptr = (Ptr) GetScriptManagerVariable  
> (smKCHRCache);
> !             UInt32 some_state = 0;
> !
>               inev.code = KeyTranslate (kchr_ptr, new_keycode,
>                             &some_state) & 0xff;
>             }
>
>
> _______________________________________________
> Emacs-devel mailing list
> Emacs-devel@gnu.org
> http://lists.gnu.org/mailman/listinfo/emacs-devel
>