What is keyboard-local?

What is keyboard-local?

[Xue Fuqiao]

In the docstring of the variable `key-translation-map':

  This keymap works like `function-key-map', but comes after that,
  and its non-prefix bindings override ordinary bindings.
  Another difference is that it is global rather than keyboard-local.

What does the "keyboard-local" mean here?  I have searched the Emacs manuals, Emacs FAQ, and the archive of this newsgroup, but I don't find anything useful.  Can anybody help?  Thanks.

-- 
Best regards, Xue Fuqiao.
http://www.emacswiki.org/emacs/XueFuqiao

Re: What is keyboard-local?

[Michael Heerdegen]

Xue Fuqiao <xfq.free@gmail.com> writes:

> In the docstring of the variable `key-translation-map':
>
>   This keymap works like `function-key-map', but comes after that,
>   and its non-prefix bindings override ordinary bindings.
>   Another difference is that it is global rather than keyboard-local.
>
> What does the "keyboard-local" mean here?  I have searched the Emacs
> manuals, Emacs FAQ, and the archive of this newsgroup, but I don't
> find anything useful.  Can anybody help?  Thanks.

I never heard of such a term "keyboard-local".  I guess "terminal-local"
or something like that is meant.  Please file a bug report, Xue.


Thanks,

Michael.

Re: What is keyboard-local?

[Stefan Monnier]

> I never heard of such a term "keyboard-local".  I guess "terminal-local"
> or something like that is meant.  Please file a bug report, Xue.

Right.  Actually, there are no real "terminal-local" variables, since
all those are in reality keyboard-local.

Emacs's "terminals" are visible in Elisp (e.g. `frame-terminal' returns
the "terminal" from where a particular frame comes).

OTOH Emacs's "keyboards" are not directly visible.  A "keyboard" belongs
to a particular terminal, but terminals can have several "keyboards",
although in practice 99.99% of the terminals have only 1 keyboard.

IIRC the only cases where you can have several keyboards for a given
terminal, is when you have an X11 server (e.g. "foo:0") with various
"X11 screens" (i.e. you have foo:0.0 and foo:0.1) in which case each X11
screen gets a corresponding Emacs "keyboard", whereas they all share the
same Emacs "terminal".

I'm pretty sure this subtlety is not handled correctly everywhere in
Emacs (IOW we have bugs in there).  We should probably get rid of
this distinction.


        Stefan

RE: What is keyboard-local?

[Drew Adams]

> Actually, there are no real "terminal-local" variables, since
> all those are in reality keyboard-local.
> 
> Emacs's "terminals" are visible in Elisp (e.g. `frame-terminal'
> returns the "terminal" from where a particular frame comes).
> 
> OTOH Emacs's "keyboards" are not directly visible.  A "keyboard"
> belongs to a particular terminal, but terminals can have several
> "keyboards", although in practice 99.99% of the terminals have
> only 1 keyboard.
> 
> IIRC the only cases where you can have several keyboards for a given
> terminal, is when you have an X11 server (e.g. "foo:0") with various
> "X11 screens" (i.e. you have foo:0.0 and foo:0.1) in which case each
> X11 screen gets a corresponding Emacs "keyboard", whereas they 
> all share the same Emacs "terminal".
> 
> I'm pretty sure this subtlety is not handled correctly everywhere in
> Emacs (IOW we have bugs in there).  We should probably get rid of
> this distinction.

While waiting for such a code fix, at least, please consider updating the doc
with info such as you just provided here, unless you feel it is already there
somewhere.