Re: Will default key bindings spell the death of Emacs?

Re: Will default key bindings spell the death of Emacs?

[Lars Hansen]

[-- Attachment #1.1: Type: text/plain, Size: 2219 bytes --]

I think this is a very important discussion.

>>I believe that Emacs key bindings are effectively preventing 99% or more
>>of Windows programmers and other younger programmers from ever using
>>it.
>>    
>>
>
>I think you greatly exaggerate the difficulty of adapting to Emacs
>key bindings.
>  
>
Users should not adopt to Emacs. Emacs should adopt to users!

Emacs is said to be customizable, and key bindings is the first thing users
want to customize. But key bindings are much to hard to change! When I 
started
using Emacs some years ago, I spend several weeks trying to figure out 
how to
bind the most basic operations to the keys I wanted. In particular I wanted
ESC to do keyboard-escape-quit and C-z, C-x, C-c, C-v to do what they do on
MS Windows. That ought not to be difficult at all, it should be a piece of
cake, especially to new users! So we do indeed have a problem!

>>I propose a toggle that sets emacs and all its IDEs into "Windows
>>Mode".
>>    
>>
>
>While this might be a good idea, I think we need to change only a
>small set of bindings frequently used by Windows users.  Otherwise,
>almost all of Emacs bindings will need to be changed, and most of the
>user manual will instantly become invalid where it describes key
>bindings.
>  
>
Yes, it is a large task to implement a possibility to completely change
keymappings of Emacs, but never the less, IMHO it is the right thing to do.
And it could be done in small steps if we lay out a strategy.

>>Do any of you share these views?  Why, or why not.  Who else cares about
>>this issue and what can we do to bring about this important change?
>>    
>>
>
>AFAIK, many (most?) Emacs users (certainly its developers) want Emacs
>to work the same on all supported platforms.  
>
I work on MS Windows as well as on two Unix-like systems, and I am very
pleased with the uniformity in user interface I get from protable 
applications
like Emacs. But to me it is not enough that the interface is the same. 
It should
also be one that suits me. And the default Emacs key bindings do not. They
seem to be selected from the idea that all operations should have their 
own key
binding simultainously! The result are long and complicated keystrokes.


[-- Attachment #1.2: Type: text/html, Size: 3065 bytes --]

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

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

Re: Will default key bindings spell the death of Emacs?

[Miles Bader]

On Thu, May 29, 2003 at 10:57:06PM +0200, Lars Hansen wrote:
> Yes, it is a large task to implement a possibility to completely change
> keymappings of Emacs, but never the less, IMHO it is the right thing to do.
> And it could be done in small steps if we lay out a strategy.

No it's a completely silly thing to do (or even waste time arguing about).

Emacs keybindings are not arbitrary, and they're a highly ingrained part of
the editor and its culture.  To change them in any large way, there would
have to be a _very_ good reason, and you certainly haven't shown one.

MSWindows keybindings are not particularly great, and in my experience, most
MSWindow users don't even _use_ keybindings (or know them), beyond the
C-c/C-x/C-v triumvirate and the `shift-movement' selection -- both of which
cua-mode handles.

-Miles
-- 
Occam's razor split hairs so well, I bought the whole argument!

Re: Will default key bindings spell the death of Emacs?

[Lars Hansen]

>
>
>Emacs keybindings are not arbitrary, and they're a highly ingrained part of
>the editor and its culture.
>
Yes, and in my opinion it is a problem.

>To change them in any large way, there would
>have to be a _very_ good reason, and you certainly haven't shown one.
>  
>
The reason is to make Emacs more user friendly. More customizable. To make
more people use Emacs, and realize that free software is not just for nerds.

>MSWindows keybindings are not particularly great, and in my experience, most
>MSWindow users don't even _use_ keybindings (or know them), beyond the
>C-c/C-x/C-v triumvirate and the `shift-movement' selection -- both of which
>cua-mode handles.
>  
>
I do not argue that we should implement MS Windows-like keybindings. I just
argue that it should be easyer to change key bindings.

Re: Will default key bindings spell the death of Emacs?

[Miles Bader]

On Fri, May 30, 2003 at 12:04:56AM +0200, Lars Hansen wrote:
> >Emacs keybindings are not arbitrary, and they're a highly ingrained part of
> >the editor and its culture.
> >
> Yes, and in my opinion it is a problem.

Why?


Since I'm not entirely 

> > MSWindows keybindings are not particularly great, and in my experience,
> > most MSWindow users don't even _use_ keybindings (or know them), beyond
> > the C-c/C-x/C-v triumvirate and the `shift-movement' selection -- both
> > of which cua-mode handles.
>
> I do not argue that we should implement MS Windows-like keybindings. I just
> argue that it should be easyer to change key bindings.

Ok, maybe I misread your post.  What exactly are you arguing for?
Do you mean there should be a nicer user interface for users to edit their
personal keybindings?

> > To change them in any large way, there would have to be a _very_ good
> > reason, and you certainly haven't shown one.
>
> The reason is to make Emacs more user friendly. More customizable. To make
> more people use Emacs, and realize that free software is not just for nerds.

I don't think this is as much an issue as you think -- in observing `naive'
(used to MSWindows apps) users using emacs, I've noticed that they usually
just ignore the keybindings altogether (even with cua-mode active) and use
the mouse/menus for _everything_.  To me this seems nuts, but I guess many
people operate this way in MSWindows apps anyway.

[and there's a certain point at which we _want_ to force people to come to
grips with emacs' different style, because it will help them in the future
(the whole give-a-fish/teach-to-fish thing)]

-Miles
-- 
Occam's razor split hairs so well, I bought the whole argument!

Re: Will default key bindings spell the death of Emacs?

[Lars Hansen]

Miles Bader wrote:

>[and there's a certain point at which we _want_ to force people to come to
>grips with emacs' different style, because it will help them in the future
>(the whole give-a-fish/teach-to-fish thing)]
>  
>
That is a good point. But I believe people have so different ways to work
that there is no single style that is the best for everybody.

Re: Will default key bindings spell the death of Emacs?

[Thien-Thi Nguyen]

Lars Hansen <larsh@math.ku.dk> writes:

   people have so different ways to work that there is no single style
   that is the best for everybody.

so the challenge for you is to upscope and see how to separate the
specific from the general, and put in place (directly or otherwise)
mechanisms that allow the general to support the specific.  otherwise,
this valuable realization goes to waste; substituting one specific w/
another does not improve the general condition.  ("meet the new boss,
same as the old boss." -- the who)

for things where there is no possible separation, the earlier that can
be determined the more time/effort you save (which is good if you are
lazy like me :-).

thi

Re: Will default key bindings spell the death of Emacs?

[Alex Schroeder]

Lars Hansen <larsh@math.ku.dk> writes:

> That is a good point. But I believe people have so different ways to
> work that there is no single style that is the best for everybody.

But then again, usability expert Jeff Raskin claims that whenever
there is something to customize, it means that the developpers didn't
know what design decision  was the best and left it up to the users
(who aren't usability experts, either).

Obviously Emacs is the mother-of-customizable-things, so we are
failing badly according to Raskin's standards.  But we at least try to
make the *defaults* usable according to our own standards.  We might
not be the best usability experts out there, but we all use Emacs day
in and day out.

Therefore, if you really want to change the keybindings -- go right
ahead.  But Emacs is complex software with tons of files, and some of
them maintained by people not part of emacs-devel, and so resolving
new inconsistencies, documenting new keybindings, etc. is going to be
a very very difficult and boring job.  If you look at CUA mode, you
will see how much hard work went into it.

There are good news, too, however:  For certain specific tasks,
people have already documented how to do it, or written the code to
do it -- eg. pc-selection-mode.

The thing to be aware of, however, is that most people on this list
think it is a low-priority problem and they have other more pressing
things on their mind.  So if you really want it, you will have to do
it.

Alex.
-- 
http://www.emacswiki.org/cgi-bin/alex.pl

Re: Will default key bindings spell the death of Emacs?

[Miles Bader]

Alex Schroeder <alex@gnu.org> writes:
> Obviously Emacs is the mother-of-customizable-things, so we are
> failing badly according to Raskin's standards.

Of course, that is as much a comment on Raskin as it is on emacs...

-Miles
-- 
P.S.  All information contained in the above letter is false,
      for reasons of military security.

Re: Will default key bindings spell the death of Emacs?

[Jan D.]

fredagen den 30 maj 2003 kl 12.52 skrev Miles Bader:

> Alex Schroeder <alex@gnu.org> writes:
>> Obviously Emacs is the mother-of-customizable-things, so we are
>> failing badly according to Raskin's standards.
>
> Of course, that is as much a comment on Raskin as it is on emacs...

But on the other hand he praises Emacs for incremental search so he
can't be all wrong ...

	Jan D.

Re: Will default key bindings spell the death of Emacs?

[Lars Hansen]

Alex Schroeder wrote:

>The thing to be aware of, however, is that most people on this list
>think it is a low-priority problem and they have other more pressing
>things on their mind.
>
I appreciate that. This is just my personal opinion.

Re: Will default key bindings spell the death of Emacs?

[Juanma Barranquero]

On Fri, 30 May 2003 12:32:06 +0200
Alex Schroeder <alex@gnu.org> wrote:

> But then again, usability expert Jeff Raskin claims that whenever
> there is something to customize, it means that the developpers didn't
> know what design decision  was the best and left it up to the users
> (who aren't usability experts, either).

Seems like a fairly narrow POV. "Whenever", indeed? What if there's *no*
best decision?

                                                                Juanma

Re: Will default key bindings spell the death of Emacs?

[Alex Schroeder]

Juanma Barranquero <jmbarranquero@laley.wke.es> writes:

> Alex Schroeder <alex@gnu.org> wrote:
>
>> But then again, usability expert Jeff Raskin claims that whenever
>> there is something to customize, it means that the developpers didn't
>> know what design decision  was the best and left it up to the users
>> (who aren't usability experts, either).
>
> Seems like a fairly narrow POV. "Whenever", indeed? What if there's *no*
> best decision?

Heh, simple: Try harder!  :)

Alex.
-- 
http://www.emacswiki.org/cgi-bin/alex.pl

Re: Will default key bindings spell the death of Emacs?

[Luc Teirlinck]

Alex Schroeder wrote:

   Juanma Barranquero <jmbarranquero@laley.wke.es> writes:

   > Alex Schroeder <alex@gnu.org> wrote:
   >
   >> But then again, usability expert Jeff Raskin claims that whenever
   >> there is something to customize, it means that the developpers didn't
   >> know what design decision  was the best and left it up to the users
   >> (who aren't usability experts, either).
   >
   > Seems like a fairly narrow POV. "Whenever", indeed? What if there's *no*
   > best decision?

   Heh, simple: Try harder!  :)

This is completely ridiculous.  Denying the importance of
customizability is denying the importance of individuality.  People
are not just all isomorphic carbon copies of each other.

Sincerely,

Luc.

Re: Will default key bindings spell the death of Emacs?

[Richard Stallman]

    Do you mean there should be a nicer user interface for users to edit their
    personal keybindings?

We definitely want to have that.  We would like to make it possible
to do this thru Custom.  ISTR that Alex was working on this last fall
but it never got installed.

Re: Will default key bindings spell the death of Emacs?

[David Kastrup]

Miles Bader <miles@gnu.org> writes:

> On Fri, May 30, 2003 at 12:04:56AM +0200, Lars Hansen wrote:

> > I do not argue that we should implement MS Windows-like
> > keybindings. I just argue that it should be easyer to change key
> > bindings.
> 
> Ok, maybe I misread your post.  What exactly are you arguing for?
> Do you mean there should be a nicer user interface for users to edit
> their personal keybindings?

`Take some more tea,' the March Hare said to Alice, very earnestly.

`I've had nothing yet,' Alice replied in an offended tone, `so I can't
 take more.'

`You mean you can't take less,' said the Hatter: `it's very easy to
 take more than nothing.'

Uh, we don't _have_ a user interface for users to edit their personal
keybindings, nice or not.  We have a programmer interface.  While you
nay call global-set-key interactively, it is nowhere in the menus, and
of course any binding you make with it does not survive into the next
session.  And yes I know about repeat-complex-command and
cut-and-paste from the minibuffer, but that is not really a user
interface.

A user interface is something like click with right mouse key on any
menu entry and get a small submenu
Bind-to-key locally
Bind-to-key globally
Save Keybinding
Add function to menu

Oh, of course, manipulating the menu entries with mouse clicks would
probably be appreciated, too.

-- 
David Kastrup, Kriemhildstr. 15, 44793 Bochum

Re: Will default key bindings spell the death of Emacs?

[Robert J. Chassell]

David.Kastrup@t-online.de (David Kastrup) wrote:

    A user interface is something like click with right mouse key on any
    menu entry and get a small submenu

No, that is not correct.  You are specifying a particular kind of user
interface for people who are sighted and who do not want to learn a
more efficent technique.

That is an important audience and Emacs should aim for it (as far as I
know, Emacs does not).

But at the same time, Emacs should also provide

  * a user interface for people who are permanently or situationally
    blind (Emacs does), and,

  * a user interface for people who desire to change more than three
    bindings at one time efficiently and who are willing to spend
    time, but less than the time of doing it awkwardly, learning how
    to do this (Emacs does, but GNOME/sawfish, as I found out
    yesterday, does not).

(Incidentally, I personally think the first "important feature" in the
etc/TODO list is more important than extending the user interface, but
the second "important feature" is less important.)

--
    Robert J. Chassell                         Rattlesnake Enterprises
    http://www.rattlesnake.com                  GnuPG Key ID: 004B4AC8
    http://www.teak.cc                             bob@rattlesnake.com

Re: Will default key bindings spell the death of Emacs?

[David Kastrup]

"Robert J. Chassell" <bob@rattlesnake.com> writes:

> David.Kastrup@t-online.de (David Kastrup) wrote:
> 
>     A user interface is something like click with right mouse key on any
>     menu entry and get a small submenu
> 
> No, that is not correct.

So you think the above description should not be called a user
interface?

> You are specifying a particular kind of user interface for people
> who are sighted and who do not want to learn a more efficent
> technique.

I was giving an example for a user interface.  What Emacs has nowadays
does not really deserve that name.

> But at the same time, Emacs should also provide
> 
>   * a user interface for people who are permanently or situationally
>     blind (Emacs does), and,

This is a somewhat orthogonal aim.  I don't see why one should make
the availability of one user interface depend on that of another.

-- 
David Kastrup, Kriemhildstr. 15, 44793 Bochum

Re: Will default key bindings spell the death of Emacs?

[Robert J. Chassell]

David.Kastrup@t-online.de (David Kastrup) wrote:

   I was giving an example for a user interface.  What Emacs has nowadays
   does not really deserve that name.

But that particular kind of user interface is not the only kind that
exists.  Emacs provides two excellent user interfaces.  However,
neither are the kind you are talking about.  (I agree that Emacs is
missing that kind.)

My point is that there are several different kinds of user interface.
It is misleading to say that

    .... an example for a user interface.  What Emacs has nowadays
    does not really deserve that name.

because that suggests that Emacs lacks a decent user interface.  

On the contrary, Emacs provides two really good user interfaces that
should not be devalued:

  * a user interface for people who are permanently or situationally
    blind (Emacs does), and,

  * a user interface for people who desire to change more than three
    bindings at one time efficiently and who are willing to spend
    time, but less than the time of doing it awkwardly, learning how
    to do this.

I agree that Emacs could do better to also provide a user interface
for people who are sighted and who do not want to learn a more
efficent technique, but it is a mistake to write as if that particular
kind of user interface is the only worthy kind.

   > But at the same time, Emacs should also provide ...

   This is a somewhat orthogonal aim.  I don't see why one should make
   the availability of one user interface depend on that of another.

I am not saying we should make one depend on another.  I am saying
that a user interface for the blind is important, as is a user
interface for those who desire to work with computers comfortably,
quickly, and efficiently.  And neither of these two kinds of user
interface should be relegated to the near invisible status of a notion
that does `not really deserve [a] name.'

-- 
    Robert J. Chassell                         Rattlesnake Enterprises
    http://www.rattlesnake.com                  GnuPG Key ID: 004B4AC8
    http://www.teak.cc                             bob@rattlesnake.com

Re: Will default key bindings spell the death of Emacs?

[David Kastrup]

"Robert J. Chassell" <bob@rattlesnake.com> writes:

> David.Kastrup@t-online.de (David Kastrup) wrote:
> 
>    I was giving an example for a user interface.  What Emacs has nowadays
>    does not really deserve that name.
> 
> But that particular kind of user interface is not the only kind that
> exists.  Emacs provides two excellent user interfaces.

No.  It provides an interface to the functionality, but "user
interface" implies something more than just the capability to manually
edit some configuration files.  Even if the format of the
configuration files can be found in "user documentation".

> My point is that there are several different kinds of user interface.
> It is misleading to say that
> 
>     .... an example for a user interface.  What Emacs has nowadays
>     does not really deserve that name.
> 
> because that suggests that Emacs lacks a decent user interface.

It does.  It is an interface, but not a user interface.

> On the contrary, Emacs provides two really good user interfaces that
> should not be devalued:
> 
>   * a user interface for people who are permanently or situationally
>     blind (Emacs does), and,
> 
>   * a user interface for people who desire to change more than three
>     bindings at one time efficiently and who are willing to spend
>     time, but less than the time of doing it awkwardly, learning how
>     to do this.

I don't see how the ability to edit .emacs should count as one user
interface, let alone two.

> I agree that Emacs could do better to also provide a user interface
> for people who are sighted and who do not want to learn a more
> efficent technique, but it is a mistake to write as if that
> particular kind of user interface is the only worthy kind.

Does `customize' not play together with things like Emacspeak?

>    > But at the same time, Emacs should also provide ...
> 
>    This is a somewhat orthogonal aim.  I don't see why one should
>    make the availability of one user interface depend on that of
>    another.
> 
> I am not saying we should make one depend on another.  I am saying
> that a user interface for the blind is important, as is a user
> interface for those who desire to work with computers comfortably,
> quickly, and efficiently.

It is the internal Emacs interface to key bindings and thus will
remain usable, and in fact, underlying any attempt of creating a user
interface to keybindings.

> And neither of these two kinds of user interface should be relegated
> to the near invisible status of a notion that does `not really
> deserve [a] name.'

Oh, it deserves a name, just not that of a user interface.  Being able
to call functionality with Lisp does not make it a user interface.
You could call the _interactive_ binding of global-set-key and its ilk
a user interface, if you really wanted to, but this interface is not
catering for persistence: you still have to edit your .emacs manually
if you want your keybindings to stay.  Not even the somewhat quaint
`disabled' commands require you to do the editing yourself in case you
want to enable commands for permanent.

-- 
David Kastrup, Kriemhildstr. 15, 44793 Bochum

Re: Will default key bindings spell the death of Emacs?

[Jan D.]

> "Robert J. Chassell" <bob@rattlesnake.com> writes:
>
>> David.Kastrup@t-online.de (David Kastrup) wrote:
>>
>>    I was giving an example for a user interface.  What Emacs has 
>> nowadays
>>    does not really deserve that name.
>>
>> But that particular kind of user interface is not the only kind that
>> exists.  Emacs provides two excellent user interfaces.
>
> No.  It provides an interface to the functionality, but "user
> interface" implies something more than just the capability to manually
> edit some configuration files.  Even if the format of the
> configuration files can be found in "user documentation".

Do you mean "Graphical user interface" when you are saying "user 
interface"?
If so, what you say makes some sense.


	Jan D.

Re: Will default key bindings spell the death of Emacs?

[David Kastrup]

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

> > "Robert J. Chassell" <bob@rattlesnake.com> writes:
> >
> >> David.Kastrup@t-online.de (David Kastrup) wrote:
> >>
> >>    I was giving an example for a user interface.  What Emacs has
> >> nowadays
> >>    does not really deserve that name.
> >>
> >> But that particular kind of user interface is not the only kind that
> >> exists.  Emacs provides two excellent user interfaces.
> >
> > No.  It provides an interface to the functionality, but "user
> > interface" implies something more than just the capability to manually
> > edit some configuration files.  Even if the format of the
> > configuration files can be found in "user documentation".
> 
> Do you mean "Graphical user interface" when you are saying "user
> interface"?
> If so, what you say makes some sense.

No, I don't mean that.  If you had bothered to read my mail to the end
before replying, you would have noticed that I called the interactive
binding for global-set-key an (although rather simplistic) user
interface for one-time setting of keybindings, but completely lacking
a way to make the keybinding persist into future sessions.

-- 
David Kastrup, Kriemhildstr. 15, 44793 Bochum

Re: Will default key bindings spell the death of Emacs?

[Jan D.]

>> Do you mean "Graphical user interface" when you are saying "user
>> interface"?
>> If so, what you say makes some sense.
>
> No, I don't mean that.  If you had bothered to read my mail to the end
> before replying, you would have noticed that I called the interactive
> binding for global-set-key an (although rather simplistic) user
> interface for one-time setting of keybindings, but completely lacking
> a way to make the keybinding persist into future sessions.

I did read your mail, you said:

"Uh, we don't _have_ a user interface for users to edit their personal
keybindings, nice or not."

and before that
"What Emacs has nowadays does not really deserve that name."

You did not say global-set-key is a simplistic user interface, you just
said it could be called interactively, which is why I asked.  Sorry for
not understanding all implied meanings.

	Jan D.

Re: Will default key bindings spell the death of Emacs?

[Alan Shutko]

Lars Hansen <larsh@math.ku.dk> writes:

> I do not argue that we should implement MS Windows-like keybindings. I just
> argue that it should be easyer to change key bindings.

In what way?  I don't think it's hard to change keybindings.  I'd say
it's difficult to completely change the keymap to something
else... but that's only because there's so much functionality, it's
hard to come up with a keymap which covers everything I want, and
does so better than Emacs.  I've done this on other editors, and I
found it easier to learn Emacs's key bindings than to come up with a
set I liked better.

There's room to make key binding customization easier (people have
talked about adding it to custom) but nobody's really come up with a
good interface yet.  What do you think?

-- 
Alan Shutko <ats@acm.org> - I am the rocks.
Looking for a developer in St. Louis? http://web.springies.com/~ats/
We are all worms. But I do believe I am a glowworm. -- Winston Churchill

Re: Will default key bindings spell the death of Emacs?

[Stefan Monnier]

> I do not argue that we should implement MS Windows-like keybindings. I just
> argue that it should be easyer to change key bindings.

What exactly are you suggesting to make it easier ?
You'll get better feedback if you show us actual concrete problems
rather than philosophical arguments about how things should be easier.

I'm not sure what you mean by keybindings being hard to change,
but let's suppose it's because C-c, C-x and ESC are so pervasive
that you can't really move them.  I think you can solve this
using mappings such as:

(define-key key-translation-map [?\C-c] [control-c])

so that hitting C-c will now lookup bindings on the `control-c'
key which is completely free for you to use.

Or you can swap to keys:

(define-key key-translation-map [?\C-x] [?\A-x])
(define-key key-translation-map [?\A-x] [?\C-x])

so that all the C-x bindings are now actually on your Alt-x key
(assuming you have both Alt and Meta keys) and vice versa.

Very few applications are as flexible when it comes to configuring
key bindings.  If you don't like long-key-combos and prefer having
different modes, you can try M-x viper RET which offers a completely
different set of bindings which many people like.


	Stefan

Re: Will default key bindings spell the death of Emacs?

[Kai Großjohann]

"Stefan Monnier" <monnier+gnu/emacs@rum.cs.yale.edu> writes:

>> I do not argue that we should implement MS Windows-like keybindings. I just
>> argue that it should be easyer to change key bindings.
>
> What exactly are you suggesting to make it easier ?

Remapping the whole keyboard doesn't make sense, but some interface
that reads a key sequence and a function name, and perhaps a map
name, and then does the right thing would be good.

You can get close by hitting the key sequence and then C-h l and
copying and pasting, but I think that's not enough.

-- 
This line is not blank.

Re: Will default key bindings spell the death of Emacs?

[Kevin Rodgers]

Kai Großjohann wrote:

> "Stefan Monnier" <monnier+gnu/emacs@rum.cs.yale.edu> writes:
> 
>>>I do not argue that we should implement MS Windows-like keybindings. I just
>>>argue that it should be easyer to change key bindings.
>>>
>>What exactly are you suggesting to make it easier ?
>>
> 
> Remapping the whole keyboard doesn't make sense, but some interface
> that reads a key sequence and a function name, and perhaps a map
> name, and then does the right thing would be good.

(defun read-keymap (prompt &optional default-map)
  "Read the name of a keymap and return as a symbol.
Prompts with PROMPT.  By default, return DEFAULT-MAP."
  (intern (completing-read prompt obarray
			   'keymapp t nil nil default-map)))

(defadvice define-key (before interactive activate)
  "Read KEY and DEF interactively, plus KEYMAP if a prefix arg is specified
\(otherwise default to `global-map'\)."
  (interactive (list (if current-prefix-arg
			 (read-keymap "Keymap: " 'global-map)
		       global-map)
		     (read-key-sequence "Key: " nil t)
		     (read-command "Command: " 'ignore))))


-- 
<a href="mailto:&lt;kevin.rodgers&#64;ihs.com&gt;">Kevin Rodgers</a>

Re: Will default key bindings spell the death of Emacs?

[Kai Großjohann]

Kevin Rodgers <ihs_4664@yahoo.com> writes:

> 		     (read-key-sequence "Key: " nil t)

This doesn't read long prefix key sequences.  That is, if C-c a is
not bound yet, I can do (kbd "C-c a b c d") to bind a key, but
read-key-sequence will stop at C-c a.

Hm.

I guess it also won't work to rebind C-x, say, since it's a prefix key.
-- 
This line is not blank.

Re: Will default key bindings spell the death of Emacs?

[Kevin Rodgers]

Kai Großjohann wrote:

> Kevin Rodgers <ihs_4664@yahoo.com> writes:
>>		     (read-key-sequence "Key: " nil t)
> 
> This doesn't read long prefix key sequences.  That is, if C-c a is
> not bound yet, I can do (kbd "C-c a b c d") to bind a key, but
> read-key-sequence will stop at C-c a.
> 
> Hm.
> 
> I guess it also won't work to rebind C-x, say, since it's a prefix key.
 
How about:


	(kbd (read-string "Key (`kbd' format): "))

That seems to require the following patch; is this a known bug in `kbd'?

*** emacs-21.3/lisp/subr.el.orig	Fri Oct 18 19:21:10 2002
--- emacs-21.3/lisp/subr.el	Mon Jun  2 15:25:36 2003
***************
*** 438,444 ****
     "Convert KEYS to the internal Emacs key representation.
   KEYS should be a string constant in the format used for
   saving keyboard macros (see `insert-kbd-macro')."
!   (read-kbd-macro keys))

   (put 'keyboard-translate-table 'char-table-extra-slots 0)

--- 438,444 ----
     "Convert KEYS to the internal Emacs key representation.
   KEYS should be a string constant in the format used for
   saving keyboard macros (see `insert-kbd-macro')."
!   `(read-kbd-macro ,keys))

   (put 'keyboard-translate-table 'char-table-extra-slots 0)


-- 
<a href="mailto:&lt;kevin.rodgers&#64;ihs.com&gt;">Kevin Rodgers</a>

Re: Will default key bindings spell the death of Emacs?

[David Kastrup]

Kevin Rodgers <ihs_4664@yahoo.com> writes:

>  How about:
> 
> 
> 	(kbd (read-string "Key (`kbd' format): "))
> 
> That seems to require the following patch; is this a known bug in `kbd'?
> 
> *** emacs-21.3/lisp/subr.el.orig	Fri Oct 18 19:21:10 2002
> --- emacs-21.3/lisp/subr.el	Mon Jun  2 15:25:36 2003
> ***************
> *** 438,444 ****
>      "Convert KEYS to the internal Emacs key representation.
>    KEYS should be a string constant in the format used for
>    saving keyboard macros (see `insert-kbd-macro')."
> !   (read-kbd-macro keys))
> 
>    (put 'keyboard-translate-table 'char-table-extra-slots 0)
> 
> --- 438,444 ----
>      "Convert KEYS to the internal Emacs key representation.
>    KEYS should be a string constant in the format used for
>    saving keyboard macros (see `insert-kbd-macro')."
> !   `(read-kbd-macro ,keys))
> 
>    (put 'keyboard-translate-table 'char-table-extra-slots 0)

Well, what about "KEYS should be a string constant" don't you
understand?  It is documented right there in the piece you have quoted
in the patch, and it is probably this way for efficiency reasons.

-- 
David Kastrup, Kriemhildstr. 15, 44793 Bochum

Re: Will default key bindings spell the death of Emacs?

[Kevin Rodgers]

David Kastrup wrote:

> Kevin Rodgers <ihs_4664@yahoo.com> writes:
>>That seems to require the following patch; is this a known bug in `kbd'?
...
> Well, what about "KEYS should be a string constant" don't you
> understand?


All of it, apparently.  :-)

> It is documented right there in the piece you have quoted
> in the patch, and it is probably this way for efficiency reasons.

I can't see any significant efficiency savings in restricting such a simple
macro's sole argument to a constant.


-- 
<a href="mailto:&lt;kevin.rodgers&#64;ihs.com&gt;">Kevin Rodgers</a>

Re: Will default key bindings spell the death of Emacs?

[David Kastrup]

Kevin Rodgers <ihs_4664@yahoo.com> writes:

> David Kastrup wrote:
> 
> > Kevin Rodgers <ihs_4664@yahoo.com> writes:
> >>That seems to require the following patch; is this a known bug in `kbd'?
> ...
> > Well, what about "KEYS should be a string constant" don't you
> > understand?
> 
> 
> All of it, apparently.  :-)
> 
> > It is documented right there in the piece you have quoted
> > in the patch, and it is probably this way for efficiency reasons.
> 
> I can't see any significant efficiency savings in restricting such a simple
> macro's sole argument to a constant.

You can't evaluate it at compile time if it isn't a constant.  The
macro is there to provide a more readable (and portable) way to encode
fixed key sequences.

-- 
David Kastrup, Kriemhildstr. 15, 44793 Bochum

Re: Will default key bindings spell the death of Emacs?

[Luc Teirlinck]

Kevin Rodgers wrote:

   How about:


	   (kbd (read-string "Key (`kbd' format): "))

   That seems to require the following patch; is this a known bug in `kbd'?

No, as David already pointed out.  However, you could just as well use
read-kbd-macro directly (unless I am confused about something).

Sincerely,

Luc.

Re: Will default key bindings spell the death of Emacs?

[Kevin Rodgers]

Luc Teirlinck wrote:

> Kevin Rodgers wrote:
>    How about:
> 
> 	   (kbd (read-string "Key (`kbd' format): "))
> 
>    That seems to require the following patch; is this a known bug in `kbd'?
> 
> No, as David already pointed out.  However, you could just as well use
> read-kbd-macro directly (unless I am confused about something).

Doh!

Thanks,
-- 
<a href="mailto:&lt;kevin.rodgers&#64;ihs.com&gt;">Kevin Rodgers</a>

Re: Will default key bindings spell the death of Emacs?

[Lars Hansen]

Stefan Monnier wrote:

>I'm not sure what you mean by keybindings being hard to change,
>but let's suppose it's because C-c, C-x and ESC are so pervasive
>that you can't really move them.
>
Yes, that quite much my problem. And I do use key-translation-map
to swap keys as you suggest. But I feel that is a workaround rather
than a real solution. Messages from Emacs then become wrong when
they refer to those key bindings.

Re: Will default key bindings spell the death of Emacs?

[Stefan Monnier]

> >I'm not sure what you mean by keybindings being hard to change,
> >but let's suppose it's because C-c, C-x and ESC are so pervasive
> >that you can't really move them.
> >
> Yes, that quite much my problem. And I do use key-translation-map
> to swap keys as you suggest. But I feel that is a workaround rather
> than a real solution. Messages from Emacs then become wrong when
> they refer to those key bindings.

Good point.  We could change those messages to take key-translation-map
into account, but it would probably be wrong (because of other uses of
key-translation-map where it shouldn't happen).

So we'd need to add a new key-remapping layer, after key-translation-map
and before the actual key-lookup.  And the remapping should not be
done by `read-key-sequence' but should be done just before looking
up the various global/local/minor maps instead.


	Stefan

Re: Will default key bindings spell the death of Emacs?

[Richard Stallman]

    Good point.  We could change those messages to take key-translation-map
    into account, but it would probably be wrong (because of other uses of
    key-translation-map where it shouldn't happen).

Could you describe the two scenarios more precisely?
In which case should the messages be changed?
What is the case you are concerned about not changing them?

    So we'd need to add a new key-remapping layer, after key-translation-map
    and before the actual key-lookup.

Please don't do this.  It is not worth putting a lot of effort into
making it somewhat easier to do radical changes in the Emacs command
set.  The users for that would be a tiny minority of Emacs users,
and there are many areas (see etc/TODO) where we could spend the same
effort and provide far more important features.

Re: Will default key bindings spell the death of Emacs?

[Stefan Monnier]

>     Good point.  We could change those messages to take key-translation-map
>     into account, but it would probably be wrong (because of other uses of
>     key-translation-map where it shouldn't happen).
> 
> Could you describe the two scenarios more precisely?
> In which case should the messages be changed?
> What is the case you are concerned about not changing them?

I don't really know.  Especially because I don't know when key-translation-map
is supposed to be used (it seems that it tends to be used more or less
interchangeably with function-key-map for very low-level things).

Let's say you bind `foo' to [?é].
Should (where-is 'foo) tell you that it's on C-x 8 ' e ?

>     So we'd need to add a new key-remapping layer, after key-translation-map
>     and before the actual key-lookup.
> 
> Please don't do this.

Don't worry: I have no intention to spend any time on this.
I have plenty of other silly things to do,


	Stefan

Re: Will default key bindings spell the death of Emacs?

[Richard Stallman]

The purpose of function-key-map is to convert ESC-sequences
to function keys.

The purpose of key-translation-map is so that the user can remap
one key sequence (perhaps including function keys) into another.

That's why key-translation-map is processed after function-key-map.

Re: Will default key bindings spell the death of Emacs?

[Stefan Monnier]

> The purpose of function-key-map is to convert ESC-sequences
> to function keys.
> 
> The purpose of key-translation-map is so that the user can remap
> one key sequence (perhaps including function keys) into another.
> 
> That's why key-translation-map is processed after function-key-map.

So maybe the C-x 8 stuff should be moved to function-key-map ?


	Stefan

Re: Will default key bindings spell the death of Emacs?

[Richard Stallman]

    So maybe the C-x 8 stuff should be moved to function-key-map ?

It ought to work the same in either keymap, but the logical place for
it is in key-translation-map.  function-key-map is only for
translating ESC-sequences to function keys.

Re: Will default key bindings spell the death of Emacs?

[Kai Großjohann]

Lars Hansen <larsh@math.ku.dk> writes:

> Yes, that quite much my problem. And I do use key-translation-map
> to swap keys as you suggest. But I feel that is a workaround rather
> than a real solution. Messages from Emacs then become wrong when
> they refer to those key bindings.

That seems to be a bug in Emacs.  The messages can be changed to
always display the right bindings.  Please report those places.
-- 
This line is not blank.

Re: Will default key bindings spell the death of Emacs?

[Tak Ota]

Thu, 29 May 2003 17:47:28 -0400: Miles Bader <miles@gnu.org> wrote:

> On Thu, May 29, 2003 at 10:57:06PM +0200, Lars Hansen wrote:
> > Yes, it is a large task to implement a possibility to completely change
> > keymappings of Emacs, but never the less, IMHO it is the right thing to do.
> > And it could be done in small steps if we lay out a strategy.
> 
> No it's a completely silly thing to do (or even waste time arguing about).

I agree with Miles.  It is silly to change the standard emacs binding
because for example C-n is not simply bound to next-line but it is
bound to the concept or notion "next" so that some other package also
binds C-n to a function that intuitively matches to the notion "next",
which may not necessarily perform next-line.  Sometimes ?n is also
bound to "next".

But at the same time I also sympathize with what Lars and others are
pointing out.

Is it worth considering as a long term development item to introduce a
notion layer between key sequence and function?  Each mode package
binds a notion to a function instead of a key sequence to a function
in its local map.  The key sequence to notion binding is defined
elsewhere.  This way if a user binds some other key sequence than C-n
to "next" it applies to entire emacs living environment including but
not limited to next-line.

This idea of notion layer is interesting and easy to say but actual
implementation work is unimaginably enormous because it involves all
lisp libraries.  I am not so positive if it is really worth daring.

-Tak

Re: Will default key bindings spell the death of Emacs?

[Miles Bader]

Tak Ota <Takaaki.Ota@am.sony.com> writes:
> This idea of notion layer is interesting and easy to say but actual
> implementation work is unimaginably enormous because it involves all
> lisp libraries.  I am not so positive if it is really worth daring.

It's been explored a bit in the past, for instance in the Hemlock editor
(CMUCL's emacs-a-like).  I think it's a useful concept, but runs into
problems if you try to apply it widely -- for instance, in emacs a
major part of the command set is `related' keybindings, like n and C-n;
are these one role?  Two?  If two, how are they related?

In emacs, you can kind of get a similar effect using Kim's `command
remapping' code.

[e.g., (local-set-key [remap next-line] 'forward-char) ]

-Miles
-- 
Next to fried food, the South has suffered most from oratory.
  			-- Walter Hines Page

Re: Will default key bindings spell the death of Emacs?

[Kai Großjohann]

Tak Ota <Takaaki.Ota@am.sony.com> writes:

> Is it worth considering as a long term development item to introduce a
> notion layer between key sequence and function?  Each mode package
> binds a notion to a function instead of a key sequence to a function
> in its local map.  The key sequence to notion binding is defined
> elsewhere.  This way if a user binds some other key sequence than C-n
> to "next" it applies to entire emacs living environment including but
> not limited to next-line.

Emacs has this already, in a way.  C-d is perceived as a generic
deletion command, and <delete> also invokes it.

It's kind of weird to consider concrete bindings to be notions, as
well, but then, why not?  There is not *so* much gain from having
better names.
-- 
This line is not blank.

Re: Will default key bindings spell the death of Emacs?

[Miles Bader]

kai.grossjohann@gmx.net (Kai Großjohann) writes:
> It's kind of weird to consider concrete bindings to be notions, as
> well, but then, why not?

Because it's much more confusing?  `concrete bindings' are not only very
cryptic as names, but by mixing concepts like that, you provide lots of
room for confusion about what's going on.

If I were designing a new editor today, I'd certainly try to use a
straight-forward progression of clearly delineated levels where
possible...

-Miles
-- 
`Life is a boundless sea of bitterness'

Re: Will default key bindings spell the death of Emacs?

[Richard Stallman]

    Is it worth considering as a long term development item to introduce a
    notion layer between key sequence and function?

We already have a feature that does more or less that job: the feature
of rebinding a command to another alternate command.  For instance, a
major mode, instead of rebinding C-n, could rebind next-line.

This feature is relatively new, and it probably isn't used in
most the places where using it would make sense.  So if someone
is interested in changing various modes to use this feature where
appropriate, that would be good.

Re: Will default key bindings spell the death of Emacs?

[Kai Großjohann]

Richard Stallman <rms@gnu.org> writes:

> We already have a feature that does more or less that job: the feature
> of rebinding a command to another alternate command.  For instance, a
> major mode, instead of rebinding C-n, could rebind next-line.

I think what's needed is the other way round: modes bind a symbolic
<next-line> function key, and the Church of Emacs uses C-n to invoke
that symbolic function key, whereas the heretic Viperians use j to
invoke the same.

And then dired, say, could bind something to <next-line> and
Viperians can then type j to invoke the special dired function.
-- 
This line is not blank.

Re: Will default key bindings spell the death of Emacs?

[Richard Stallman]

    > We already have a feature that does more or less that job: the feature
    > of rebinding a command to another alternate command.  For instance, a
    > major mode, instead of rebinding C-n, could rebind next-line.

    I think what's needed is the other way round: modes bind a symbolic
    <next-line> function key, and the Church of Emacs uses C-n to invoke
    that symbolic function key, whereas the heretic Viperians use j to
    invoke the same.


    And then dired, say, could bind something to <next-line> and
    Viperians can then type j to invoke the special dired function.

Sorry, I don't see the difference.  As far as I can tell, the existing
feature does exactly this.

Re: Will default key bindings spell the death of Emacs?

[Kai Großjohann]

Richard Stallman <rms@gnu.org> writes:

>     I think what's needed is the other way round: modes bind a symbolic
>     <next-line> function key, and the Church of Emacs uses C-n to invoke
>     that symbolic function key, whereas the heretic Viperians use j to
>     invoke the same.
>
> Sorry, I don't see the difference.  As far as I can tell, the existing
> feature does exactly this.

Me sorry.  Me brainfart.  *blush*
-- 
This line is not blank.

Re: Will default key bindings spell the death of Emacs?

[Kevin Rodgers]

Lars Hansen wrote:

> Users should not adopt to Emacs. Emacs should adopt to users!


Users should adopt Emacs' ways -- it is far more mature than they are.

Emacs _can_ adapt to users.


> Emacs is said to be customizable, and key bindings is the first thing users
> want to customize. But key bindings are much to hard to change! When I 
> started
> using Emacs some years ago, I spend several weeks trying to figure out 
> how to
> bind the most basic operations to the keys I wanted. In particular I wanted
> ESC to do keyboard-escape-quit and C-z, C-x, C-c, C-v to do what they do on
> MS Windows. That ought not to be difficult at all, it should be a piece of
> cake, especially to new users! So we do indeed have a problem!


Then you should submit a proper bug report: exactly what you did, how
Emacs responded, and what you expected to happen.


> Yes, it is a large task to implement a possibility to completely change
> keymappings of Emacs, but never the less, IMHO it is the right thing to do.
> And it could be done in small steps if we lay out a strategy.


"We" will only lay out a strategy if you convince us that it's the right thing
to do.  But you haven't.


> I work on MS Windows as well as on two Unix-like systems, and I am very
> pleased with the uniformity in user interface I get from protable 
> applications
> like Emacs. But to me it is not enough that the interface is the same. 
> It should
> also be one that suits me. And the default Emacs key bindings do not.


Yes, you've said that you prefer the Microsoft Windows conventions.  But
you haven't said how CUA mode fails to meet your preferences.

> They
> seem to be selected from the idea that all operations should have their 
> own key
> binding simultainously! The result are long and complicated keystrokes.

I'm not sure what you mean.  Are you referring to Control-Meta- key
combinations?  You can avoid them by using ESC as a prefix to the
corresponding Control- key combination, or using the menu bar, or
defining your own "shortcuts" (key bindings).  Or are you referring
to the C-x and C-c prefix keys?

In any case, I resent the fact that you think your individual preferences
should become the default for me, even though I've been using Emacs on
Unix exclusively for almost 20 years.

-- 
<a href="mailto:&lt;kevin.rodgers&#64;ihs.com&gt;">Kevin Rodgers</a>

Re: Will default key bindings spell the death of Emacs?

[Kai Großjohann]

Lars Hansen <larsh@math.ku.dk> writes:

> Emacs is said to be customizable, and key bindings is the first
> thing users want to customize. But key bindings are much to hard to
> change! When I started using Emacs some years ago, I spend several
> weeks trying to figure out how to bind the most basic operations to
> the keys I wanted. In particular I wanted ESC to do
> keyboard-escape-quit and C-z, C-x, C-c, C-v to do what they do on MS
> Windows. That ought not to be difficult at all, it should be a piece
> of cake, especially to new users! So we do indeed have a problem!

I'm not so sure it's desirable to make it that easy.  I have fairly
strong feelings about my window manager keybindings, for instance.
So when I started to use KDE, I also started to configure its
keybindings.  KDE has a very intuitive interface for doing this, so I
had no trouble to do the changes themselves.  But it was not so easy
to find a consistent set of keybindings.

I also started to use Gnome a short while ago, and there I also
changed the keybindings to my liking.  I had similar problems, only
fewer of them, because Gnome doesn't allow for configuring so many
bindings.

Given that Emacs has *hundreds* of keybindings, people are going to
have real problems with the rebinding.  For example, if you rebind
ESC and C-x and C-c, then you lose a lot of other bindings that you
somehow need to make accessible.

These days, there is the cua package which tries hard to make C-z,
C-x, C-c and C-v available.  It was hard work.

I like the Gnome approach where you can select some flavors of
keybindings and don't have to change each one of them individually.
This is also offered by Emacs, via cua and crisp-mode and
wordstar-mode and so on.

-- 
This line is not blank.

Re: Will default key bindings spell the death of Emacs?

[Miles Bader]

kai.grossjohann@gmx.net (Kai Großjohann) writes:
> I also started to use Gnome a short while ago, and there I also
> changed the keybindings to my liking.  I had similar problems, only
> fewer of them, because Gnome doesn't allow for configuring so many
> bindings.
> 
> Given that Emacs has *hundreds* of keybindings, people are going to
> have real problems with the rebinding.  For example, if you rebind
> ESC and C-x and C-c, then you lose a lot of other bindings that you
> somehow need to make accessible.

I think in part this simply points out that the tools available for
changing bindings are not very good.  It's not hard to image something
that would give you a broad visual presentation of the current bindings,
highlight which key ranges are `reserved for users,' help you avoid
conflicts, allow you to define your own `binding sets,' etc.

> I like the Gnome approach where you can select some flavors of
> keybindings and don't have to change each one of them individually.
> This is also offered by Emacs, via cua and crisp-mode and
> wordstar-mode and so on.

It's nice to have some preselected flavors available that capture
various popular preferences, but the Gnome viewpoint that users Just
Shouldn't Change Anything is assinine.  Boo, gnome.*

* Yes I know that it makes life much simpler for the maintainers, and
  support/help/etc, but none the less it's user hostile, and I am a user.

-Miles
-- 
Ich bin ein Virus. Mach' mit und kopiere mich in Deine .signature.

Re: Will default key bindings spell the death of Emacs?

[Kai Großjohann]

Miles Bader <miles@lsi.nec.co.jp> writes:

> kai.grossjohann@gmx.net (Kai Großjohann) writes:
>> 
>> Given that Emacs has *hundreds* of keybindings, people are going to
>> have real problems with the rebinding.  For example, if you rebind
>> ESC and C-x and C-c, then you lose a lot of other bindings that you
>> somehow need to make accessible.
>
> I think in part this simply points out that the tools available for
> changing bindings are not very good.  It's not hard to image something
> that would give you a broad visual presentation of the current bindings,
> highlight which key ranges are `reserved for users,' help you avoid
> conflicts, allow you to define your own `binding sets,' etc.

But there is a relationship between keybindings.  How to capture that?

If a user has remapped transpose-chars to somewhere, how should the
system know where transpose-words and transpose-lines should go?

The problem is AI complete, IMHO.
-- 
This line is not blank.

Re: Will default key bindings spell the death of Emacs?

[Miles Bader]

kai.grossjohann@gmx.net (Kai Großjohann) writes:
> > I think in part this simply points out that the tools available for
> > changing bindings are not very good.  It's not hard to image something
> > that would give you a broad visual presentation of the current bindings,
> > highlight which key ranges are `reserved for users,' help you avoid
> > conflicts, allow you to define your own `binding sets,' etc.
> 
> But there is a relationship between keybindings.  How to capture that?
> 
> If a user has remapped transpose-chars to somewhere, how should the
> system know where transpose-words and transpose-lines should go?

Why should the software care?  For the relatively simple task of
allowing users to easily add their own bindings, you don't have to
ensure that the result is somehow a perfect symmetrical set of bindings,
you merely have to help the user avoid some common problems.

If the user cares about other related bindings, he'll try to figure
something out, the software's only job is to assist him by providing
good tools.

-miles
-- 
o The existentialist, not having a pillow, goes everywhere with the book by
  Sullivan, _I am going to spit on your graves_.

Re: Will default key bindings spell the death of Emacs?

[Kai Großjohann]

Miles Bader <miles@lsi.nec.co.jp> writes:

> Why should the software care?  For the relatively simple task of
> allowing users to easily add their own bindings, you don't have to
> ensure that the result is somehow a perfect symmetrical set of
> bindings,

When I was composing the message, I was still under the (mistaken)
impression that Lars wanted easy customization of all the default
bindings.

Sorry.
-- 
This line is not blank.

Re: Will default key bindings spell the death of Emacs?

[Andreas Schwab]

Lars Hansen <larsh@math.ku.dk> writes:

|> I think this is a very important discussion.
|> 
|> >>I believe that Emacs key bindings are effectively preventing 99% or more
|> >>of Windows programmers and other younger programmers from ever using
|> >>it.
|> >>    
|> >
|> >I think you greatly exaggerate the difficulty of adapting to Emacs
|> >key bindings.
|> >  
|> Users should not adopt to Emacs. Emacs should adopt to users!

You might be surprised how many GUIs support Emacs keybindings like C-k.

Andreas.

-- 
Andreas Schwab, SuSE Labs, schwab@suse.de
SuSE Linux AG, Deutschherrnstr. 15-19, D-90429 Nürnberg
Key fingerprint = 58CA 54C7 6D53 942B 1756  01D3 44D5 214B 8276 4ED5
"And now for something completely different."

Re: Will default key bindings spell the death of Emacs?

[Lars Hansen]

Andreas Schwab wrote:

>Lars Hansen <larsh@math.ku.dk> writes:
>
>|> I think this is a very important discussion.
>|> 
>|> >>I believe that Emacs key bindings are effectively preventing 99% or more
>|> >>of Windows programmers and other younger programmers from ever using
>|> >>it.
>|> >>    
>|> >
>|> >I think you greatly exaggerate the difficulty of adapting to Emacs
>|> >key bindings.
>|> >  
>|> Users should not adopt to Emacs. Emacs should adopt to users!
>
>You might be surprised how many GUIs support Emacs keybindings like C-k.
>
>Andreas.
>  
>
I never intended to discuss which keybindings people should use.
On the contrary I think people should have the oppotunity to decide
for themselves.

Re: Will default key bindings spell the death of Emacs?

[David Kastrup]

Lars Hansen <larsh@math.ku.dk> writes:

> Andreas Schwab wrote:
> >
> >You might be surprised how many GUIs support Emacs keybindings like
> >C-k.
>
> I never intended to discuss which keybindings people should use.  On
> the contrary I think people should have the oppotunity to decide for
> themselves.

They don't have the opportunity if you don't have
a) an easy way to configure keybindings manually
b) an easy way to carry your keybindings around on a floppy, web site
or USB stick.

Just copying .emacs will not cut it, really, since that tends to
contain machine-specific stuff as well.

If you will lose your personal keybindings the moment you go to a
different machine, there is not much point in customizing to your
convenience.

Should keybindings be themeable?  Well, quite a few of people complain
about changed bindings and other stuff on upgrades.  But it would
probably be impractical to cater for them too much by providing Emacs
with a set of more historical keybindings.

But maybe more radical keybinding redesigns that are optional might
be possible.  I don't know enough of, say, viper-mode, to know
whether such experiments are bound to work well.

-- 
David Kastrup, Kriemhildstr. 15, 44793 Bochum

Re: Will default key bindings spell the death of Emacs?

[Benjamin Riefenstahl]

Hi David,


I don't want to enter into an argument here, but for some of the
points that you made I had associations.

David.Kastrup@t-online.de (David Kastrup) writes:
> b) an easy way to carry your keybindings around on a floppy, web
> site or USB stick.
>
> Just copying .emacs will not cut it, really, since that tends to
> contain machine-specific stuff as well.

Just a thought:

If you want carry your .emacs around on some media, than you will have
to install it somehow, unless I am misunderstanding something about
the setup that you mean.  In that case you can just as easily call it
dak-keybindings.el and either just M-x load-library it or put a (load)
form into the .emacs that is there.  IOW, it doesn't have to be a
complete .emacs.

> If you will lose your personal keybindings the moment you go to a
> different machine, there is not much point in customizing to your
> convenience.

Right.  As I do it (and I'd guess some other people, too), I customize
my setup, but I also try to learn some of the default bindings, so I
can get around without my customizations.  I end up using a mixture of
my own shortcuts, default keys, and keys that I know from other
environments, like CUA.  Which is fine for me.

> Should keybindings be themeable?

CUA-mode and pc-selection-mode actually already do that, and also
brief-mode and wordstar-mode, I think.  All those package only address
a limited set of keys, I think.  Most single Emacs application modes
define more keys.


so long, benny

Re: Will default key bindings spell the death of Emacs?

[Richard Stallman]

    They don't have the opportunity if you don't have
    a) an easy way to configure keybindings manually
    b) an easy way to carry your keybindings around on a floppy, web site
    or USB stick.

Please don't argue about this.  We certainly want to have a better
user interface for customizing and saving key bindings.  Nobody is
against it.  We just need someone to write it.

Would you like to write it?

Re: Will default key bindings spell the death of Emacs?

[David Kastrup]

Richard Stallman <rms@gnu.org> writes:

>     They don't have the opportunity if you don't have
>     a) an easy way to configure keybindings manually
>     b) an easy way to carry your keybindings around on a floppy, web
>     site or USB stick.
> 
> Please don't argue about this.  We certainly want to have a better
> user interface for customizing and saving key bindings.  Nobody is
> against it.  We just need someone to write it.
> 
> Would you like to write it?

I'd like to manage having the time for properly working on those Emacs
related projects that I am supposed to be actively maintaining.  So I
am afraid that in my personal queue, it is not likely to reach the
front anytime soon.

-- 
David Kastrup, Kriemhildstr. 15, 44793 Bochum

Re: Will default key bindings spell the death of Emacs?

[Robert J. Chassell]

    Should keybindings be themeable?

They have been for a really long time.

For example, VIP mode is from 1986 and a note for VI mode says its
last revision was in 1987 for GNU Emacs 18.33.  Wordstar mode is from
1991.

CUA minor mode, which "provides a complete emulation of the standard
CUA key bindings (Motif/Windows/Mac GUI)", is more recent.  It is from
1997.

-- 
    Robert J. Chassell                         Rattlesnake Enterprises
    http://www.rattlesnake.com                  GnuPG Key ID: 004B4AC8
    http://www.teak.cc                             bob@rattlesnake.com

Re: Will default key bindings spell the death of Emacs?

[Lars Hansen]

Tak Ota wrote:

>Is it worth considering as a long term development item to introduce a
>notion layer between key sequence and function?  Each mode package
>binds a notion to a function instead of a key sequence to a function
>in its local map.  The key sequence to notion binding is defined
>elsewhere.  This way if a user binds some other key sequence than C-n
>to "next" it applies to entire emacs living environment including but
>not limited to next-line.
>
It sounds as a very good idea to have notions, at least in those places
where it makes sense. It would greatly ease making consistent key
rebindings.

Re: Will default key bindings spell the death of Emacs?

[Lars Hansen]

Kevin Rodgers wrote:

>In any case, I resent the fact that you think your individual preferences
>should become the default for me, even though I've been using Emacs on
>Unix exclusively for almost 20 years.
>
I am sorry that you read my posting in that way. It is certainly not my
intension to impose my personal preferences to anybody. I just think Emacs
keybindings should be easier to change.
So actually the subject line "Will default key bindings..." (which does not
originate from me) is misleading.

Re: Will default key bindings spell the death of Emacs?

[Eli Zaretskii]

[I suggest to resend your message to emacs-devel.]

> From: Glen Peterson <glen@organicdesign.org>
> Reply-To: glen@organicdesign.org
> Date: 28 May 2003 23:16:59 -0400
> 
> I believe that Emacs key bindings are effectively preventing 99% or more
> of Windows programmers and other younger programmers from ever using
> it.

I think you greatly exaggerate the difficulty of adapting to Emacs
key bindings.

> I propose a toggle that sets emacs and all its IDEs into "Windows
> Mode".

While this might be a good idea, I think we need to change only a
small set of bindings frequently used by Windows users.  Otherwise,
almost all of Emacs bindings will need to be changed, and most of the
user manual will instantly become invalid where it describes key
bindings.

I suggest to try CUA mode (in the CVS sources) and see if it solves
your problem well enough for it to become much less grave.

> * The kill ring should not be used.  Cut and copy and paste should
> operate on the Windows clipboard

Emacs does both, so I don't see why is the default operation a
problem.  A user who is unaware of the kill ring should be able to
simply ignore it.

> Delete - delete 1 char to the right.
> Backspace - move and delete 1 char to the left

These two already work like that.  If you have any problems, please
describe them.

> cua-mode bindings is a good idea, but I don't think it goes far enough. 

The question is not whether CUA mode makes Emacs be a Windows program,
the question is does it go far enough as to allow Windows users to
become efficient Emacs users with minimal adaptation pain.

> Particularly disturbing are the state sensitive M-x and C-x keys:
> ftp://ftp.xemacs.org/pub/xemacs/contrib/cua-mode.el

I sugegst to try the version of CUA mode that is part of the Emacs CVS
sources, not the one in XEmacs (perhaps those are the same, I don't
know).

> Do any of you share these views?  Why, or why not.  Who else cares about
> this issue and what can we do to bring about this important change?

AFAIK, many (most?) Emacs users (certainly its developers) want Emacs
to work the same on all supported platforms.

Re: Will default key bindings spell the death of Emacs?

[Jan D.]

>> I believe that Emacs key bindings are effectively preventing 99% or 
>> more
>> of Windows programmers and other younger programmers from ever using
>> it.
>
> I think you greatly exaggerate the difficulty of adapting to Emacs
> key bindings.

At my work there are many MS Windows users.  None had any trouble 
adopting
to Emacs.  Well, one, but he uses vim, so being a Windows user wasn't
a factor in that case :-)


>> * The kill ring should not be used.  Cut and copy and paste should
>> operate on the Windows clipboard

This assumes the MS Windows model is superior to Emacs.  I think it
is the opposite.

>> Do any of you share these views?  Why, or why not.  Who else cares 
>> about
>> this issue and what can we do to bring about this important change?
>
> AFAIK, many (most?) Emacs users (certainly its developers) want Emacs
> to work the same on all supported platforms.

Yes, this is a big advantage.

	Jan D.