Patch for emacs/basic.texi

Patch for emacs/basic.texi

[Richard Stallman]

[[[ To any NSA and FBI agents reading my email: please consider    ]]]
[[[ whether defending the US Constitution against all enemies,     ]]]
[[[ foreign or domestic, requires you to follow Snowden's example. ]]]

Would someone please install this and ack?
The log entry can be

Document the easiest way to type the undo key.


diff --git a/doc/emacs/basic.texi b/doc/emacs/basic.texi
index abb385f53d..2372ce972d 100644
--- a/doc/emacs/basic.texi
+++ b/doc/emacs/basic.texi
@@ -461,6 +461,11 @@ Normally, this command undoes the last change, moving point back to
 where it was before the change.  The undo command applies only to
 changes in the buffer; you can't use it to undo cursor motion.
 
+  On a graphics terminal, the easiest way to invoke @code{undo} is
+with @kbd{C-/}; that doesn't need the shift key.  On a text terminal,
+@kbd{C-/} does not exist, but usually you can type @kbd{C-_} without
+the shift key and it will work anyway.
+
   Although each editing command usually makes a separate entry in the
 undo records, very simple commands may be grouped together.
 Sometimes, an entry may cover just part of a complex command.

-- 
Dr Richard Stallman
Chief GNUisance of the GNU Project (https://gnu.org)
Founder, Free Software Foundation (https://fsf.org)
Internet Hall-of-Famer (https://internethalloffame.org)

Re: Patch for emacs/basic.texi

[Alfred M. Szmidt]

   diff --git a/doc/emacs/basic.texi b/doc/emacs/basic.texi
   index abb385f53d..2372ce972d 100644
   --- a/doc/emacs/basic.texi
   +++ b/doc/emacs/basic.texi
   @@ -461,6 +461,11 @@ Normally, this command undoes the last change, moving point back to
    where it was before the change.  The undo command applies only to
    changes in the buffer; you can't use it to undo cursor motion.

   +  On a graphics terminal, the easiest way to invoke @code{undo} is
   +with @kbd{C-/}; that doesn't need the shift key.  On a text terminal,
   +@kbd{C-/} does not exist, but usually you can type @kbd{C-_} without
   +the shift key and it will work anyway.

That doesn't work on my terminal (wsvt25 or xterm with TERM=vt100);
where one has to press C-_ (shifted).  C-- ignores the Control key,
and does a self-insert-command on ?-.

Re: Patch for emacs/basic.texi

[Thibaut Verron]

> diff --git a/doc/emacs/basic.texi b/doc/emacs/basic.texi
> index abb385f53d..2372ce972d 100644
> --- a/doc/emacs/basic.texi
> +++ b/doc/emacs/basic.texi
> @@ -461,6 +461,11 @@ Normally, this command undoes the last change, moving point back to
>  where it was before the change.  The undo command applies only to
>  changes in the buffer; you can't use it to undo cursor motion.
>
> +  On a graphics terminal, the easiest way to invoke @code{undo} is
> +with @kbd{C-/}; that doesn't need the shift key.  On a text terminal,
> +@kbd{C-/} does not exist, but usually you can type @kbd{C-_} without
> +the shift key and it will work anyway.
> +

Does this only make sense with a qwerty keyboard?

Re: Patch for emacs/basic.texi

[Kévin Le Gouguec]

Thibaut Verron <thibaut.verron@gmail.com> writes:

>> diff --git a/doc/emacs/basic.texi b/doc/emacs/basic.texi
>> index abb385f53d..2372ce972d 100644
>> --- a/doc/emacs/basic.texi
>> +++ b/doc/emacs/basic.texi
>> @@ -461,6 +461,11 @@ Normally, this command undoes the last change, moving point back to
>>  where it was before the change.  The undo command applies only to
>>  changes in the buffer; you can't use it to undo cursor motion.
>>
>> +  On a graphics terminal, the easiest way to invoke @code{undo} is
>> +with @kbd{C-/}; that doesn't need the shift key.  On a text terminal,
>> +@kbd{C-/} does not exist, but usually you can type @kbd{C-_} without
>> +the shift key and it will work anyway.
>> +
>
> Does this only make sense with a qwerty keyboard?

This advice probably isn't universal, indeed.  E.g. with AZERTY
keyboards, one needs to hit shift to get a slash, while no shift key is
needed for underscores.  Therefore I always undo with C-_.

Re: Patch for emacs/basic.texi

[Andreas Schwab]

On Sep 16 2020, Kévin Le Gouguec wrote:

> Thibaut Verron <thibaut.verron@gmail.com> writes:
>
>>> diff --git a/doc/emacs/basic.texi b/doc/emacs/basic.texi
>>> index abb385f53d..2372ce972d 100644
>>> --- a/doc/emacs/basic.texi
>>> +++ b/doc/emacs/basic.texi
>>> @@ -461,6 +461,11 @@ Normally, this command undoes the last change, moving point back to
>>>  where it was before the change.  The undo command applies only to
>>>  changes in the buffer; you can't use it to undo cursor motion.
>>>
>>> +  On a graphics terminal, the easiest way to invoke @code{undo} is
>>> +with @kbd{C-/}; that doesn't need the shift key.  On a text terminal,
>>> +@kbd{C-/} does not exist, but usually you can type @kbd{C-_} without
>>> +the shift key and it will work anyway.
>>> +
>>
>> Does this only make sense with a qwerty keyboard?
>
> This advice probably isn't universal, indeed.  E.g. with AZERTY
> keyboards, one needs to hit shift to get a slash, while no shift key is
> needed for underscores.  Therefore I always undo with C-_.

And on a qwertz keyboard neither C-/ nor C-_ work without shift.

Andreas.

-- 
Andreas Schwab, schwab@linux-m68k.org
GPG Key fingerprint = 7578 EB47 D4E5 4D69 2510  2552 DF73 E780 A9DA AEC1
"And now for something completely different."

Re: Patch for emacs/basic.texi

[Antoine Kalmbach]

Andreas Schwab <schwab@linux-m68k.org> writes:

>>>> +  On a graphics terminal, the easiest way to invoke @code{undo} is
>>>> +with @kbd{C-/}; that doesn't need the shift key.  On a text terminal,
>>>> +@kbd{C-/} does not exist, but usually you can type @kbd{C-_} without
>>>> +the shift key and it will work anyway.
>>>> +
>>>
>>> Does this only make sense with a qwerty keyboard?
>>
>> This advice probably isn't universal, indeed.  E.g. with AZERTY
>> keyboards, one needs to hit shift to get a slash, while no shift key is
>> needed for underscores.  Therefore I always undo with C-_.
>
> And on a qwertz keyboard neither C-/ nor C-_ work without shift.
>
> Andreas.

Isn't `C-x u` also a default keybinding? Although it requires two keypresses,
that should work on all keyboard layouts.

-- 
Antoine Kalmbach

Re: Patch for emacs/basic.texi

[Stefan Kangas]

Andreas Schwab <schwab@linux-m68k.org> writes:

> And on a qwertz keyboard neither C-/ nor C-_ work without shift.

Same on standard Swedish (qwerty) keyboards, FWIW.

Re: Patch for emacs/basic.texi

[Stefan Monnier]

> +  On a graphics terminal, the easiest way to invoke @code{undo} is
> +with @kbd{C-/}; that doesn't need the shift key.

This is true for the US keyboard layout but is not true in general, so
it should be appropriately qualified.

> On a text terminal, +@kbd{C-/} does not exist, but usually you can
> type @kbd{C-_} without +the shift key and it will work anyway.

Similarly, I can imagine some terminals where C-/ works, as well as
keyboards where "_" doesn't require a shift key, or terminals where "C-_
without the shift key" won't send a byte-sequence that ends up bound to
`undo`.

E.g. I just tried `C-h k C--` in a "plain old xterm" here with a US keyboard
and Emacs told me:

    C-- (translated from M-[ 2 7 ; 5 ; 4 5 ~) runs the command
    negative-argument (found in global-map), which is an interactive
    compiled Lisp function in ‘simple.el’.
    [...]

OTOH, `C-h k C-/` in this same terminal gave me:

    \^_ runs the command undo (found in global-map), which is an
    interactive compiled Lisp function in ‘simple.el’.
    [...]

Could it be that you misremembered what was the "trick" behind having
both `C-/` and `C-_` keybindings (it's not `C--` but `C-/` that gets
mapped to C-_)?


        Stefan

Re: Patch for emacs/basic.texi

[Eli Zaretskii]

> From: ams@gnu.org (Alfred M. Szmidt)
> Date: Wed, 16 Sep 2020 01:14:53 -0400
> Cc: emacs-devel@gnu.org
> 
>    +  On a graphics terminal, the easiest way to invoke @code{undo} is
>    +with @kbd{C-/}; that doesn't need the shift key.  On a text terminal,
>    +@kbd{C-/} does not exist, but usually you can type @kbd{C-_} without
>    +the shift key and it will work anyway.
> 
> That doesn't work on my terminal (wsvt25 or xterm with TERM=vt100);
> where one has to press C-_ (shifted).  C-- ignores the Control key,
> and does a self-insert-command on ?-.

Richard's text says "usually", so I guess it doesn't need to be
correct always?

Re: Patch for emacs/basic.texi

[Alfred M. Szmidt]

   >    +  On a graphics terminal, the easiest way to invoke @code{undo} is
   >    +with @kbd{C-/}; that doesn't need the shift key.  On a text terminal,
   >    +@kbd{C-/} does not exist, but usually you can type @kbd{C-_} without
   >    +the shift key and it will work anyway.
   > 
   > That doesn't work on my terminal (wsvt25 or xterm with TERM=vt100);
   > where one has to press C-_ (shifted).  C-- ignores the Control key,
   > and does a self-insert-command on ?-.

   Richard's text says "usually", so I guess it doesn't need to be
   correct always?

I tried several more terminals, none of them worked with C-- --
usually usually means that it usually works -- at least sometimes.

Re: Patch for emacs/basic.texi

[Kévin Le Gouguec]

Antoine Kalmbach <ane@iki.fi> writes:

> Isn't `C-x u` also a default keybinding? Although it requires two keypresses,
> that should work on all keyboard layouts.

It is a default binding, but it's more cumbersome for "undo bursts"
(repeating "undo" a bunch of times).

(It's one of those bindings, along with with "C-x o", that I wish came
with a "C-x z"-like transient map.)

RE: Patch for emacs/basic.texi

[Drew Adams]

> > Isn't `C-x u` also a default keybinding? Although it requires two
> > keypresses, that should work on all keyboard layouts.
> 
> It is a default binding, but it's more cumbersome for "undo bursts"
> (repeating "undo" a bunch of times).
> 
> (It's one of those bindings, along with with "C-x o", that I wish came
> with a "C-x z"-like transient map.)

It's trivial to create a repeating command, so you
can use `C-x u u u u ...'.

Either of the definitions below for this (`undo-repeat')
works.  The second one is not specific to any key binding
- use it with any prefix key, not just `C-x u'.

Then use this binding:
(global-set-key [remap undo] 'undo-repeat)

1.

(defun undo-repeat (arg)
  "Same as `undo', but repeatable even on a prefix key.
E.g., if bound to `C-x u' then you can use `C-x u u u...' to repeat."
  (interactive "*P")
  (undo arg)
  (set-transient-map (let ((map  (make-sparse-keymap)))
                       (define-key map "u" 'undo-repeat)
                       map)))

2.

(defun repeat-command (command)
  "Repeat COMMAND."
  (require 'repeat) ; Define its vars before we let-bind them.
  (let ((repeat-previous-repeated-command  command)
        (repeat-message-function           #'ignore)
        (last-repeatable-command           'repeat))
    (repeat nil)))

(defun undo-repeat (arg)
  "Same as `undo', but repeatable even on a prefix key.
E.g., if bound to `C-x u' then you can use `C-x u u u...' to repeat."
  (interactive "*P")
  (repeat-command 'undo))


Similarly, for `C-x o' (`other-window'):

(defun other-window-repeat ()
  "Same as `other-window', but repeatable even on a prefix key.
E.g., if bound to `C-x o' then you can use `C-x o o o...' to repeat."
  (interactive)
  (repeat-command 'other-window))

(global-set-key [remap other-window] 'other-window-repeat)

I do this all over the place, where it makes sense.

Repeating commands (was: Patch for emacs/basic.texi)

[Kévin Le Gouguec]

Drew Adams <drew.adams@oracle.com> writes:

>> (It's one of those bindings, along with with "C-x o", that I wish came
>> with a "C-x z"-like transient map.)
>
> It's trivial to create a repeating command, so you
> can use `C-x u u u u ...'.

Thanks for these tips!  I actually already had setup method 1
(set-transient-map) for other-window, but never bothered with undo.

Method 2 (repeat) is neat!  I wonder whether these three variables you
bind should be mentioned in repeat's docstring, I would never have
guessed this function could be programmed this way.


(Though I did get a kick just now out of reading repeat.el's commentary,
squinting at the existentialist conclusion, checking the "Author:"
attribute just in case, and finding out that yep, this *is* Will
Mengarini's fine work.)

Re: Patch for emacs/basic.texi

[Richard Stallman]

[[[ To any NSA and FBI agents reading my email: please consider    ]]]
[[[ whether defending the US Constitution against all enemies,     ]]]
[[[ foreign or domestic, requires you to follow Snowden's example. ]]]

  > > And on a qwertz keyboard neither C-/ nor C-_ work without shift.

  > Same on standard Swedish (qwerty) keyboards, FWIW.

You can't please everybody, or every keyboard.

-- 
Dr Richard Stallman
Chief GNUisance of the GNU Project (https://gnu.org)
Founder, Free Software Foundation (https://fsf.org)
Internet Hall-of-Famer (https://internethalloffame.org)

Re: Patch for emacs/basic.texi

[Richard Stallman]

[[[ To any NSA and FBI agents reading my email: please consider    ]]]
[[[ whether defending the US Constitution against all enemies,     ]]]
[[[ foreign or domestic, requires you to follow Snowden's example. ]]]

  > Could it be that you misremembered what was the "trick" behind having
  > both `C-/` and `C-_` keybindings (it's not `C--` but `C-/` that gets
  > mapped to C-_)?

Nothing like that.

When I came up with this binding scheme, I had to find a solution that
worked with the terminals that existed.  The text terminals supported
only the control characters of ASCII.  C-- is the same as C-_, which
is ^_, 037.  There is no C-? in ASCII, so C-/ and C-?  send DEL (0177)
-- not available for Undo.

Later I found that C-/ was a convenient binding for terminals that
could support the CTRL key with any printing character.

This history should be in the ChangeLogs.

xterm displays only text, but it supports the CTRL key with any
printing character, like X itself.  For output, it is a text terminal,
but in input it is like a graphics terminal.


-- 
Dr Richard Stallman
Chief GNUisance of the GNU Project (https://gnu.org)
Founder, Free Software Foundation (https://fsf.org)
Internet Hall-of-Famer (https://internethalloffame.org)

Re: Patch for emacs/basic.texi

[Richard Stallman]

[[[ To any NSA and FBI agents reading my email: please consider    ]]]
[[[ whether defending the US Constitution against all enemies,     ]]]
[[[ foreign or domestic, requires you to follow Snowden's example. ]]]

  > > +  On a graphics terminal, the easiest way to invoke @code{undo} is
  > > +with @kbd{C-/}; that doesn't need the shift key.  On a text terminal,
  > > +@kbd{C-/} does not exist, but usually you can type @kbd{C-_} without
  > > +the shift key and it will work anyway.
  > > +

  > Does this only make sense with a qwerty keyboard?

I don't think the keyboard _layout_ makes a difference.  An ASCII-only Dvorak
keyboard would presumably support the same control characters.

What matters is whether the keyboard uses ASCII or some wider character set.


-- 
Dr Richard Stallman
Chief GNUisance of the GNU Project (https://gnu.org)
Founder, Free Software Foundation (https://fsf.org)
Internet Hall-of-Famer (https://internethalloffame.org)

Re: Patch for emacs/basic.texi

[Thibaut Verron]

[-- Attachment #1: Type: text/plain, Size: 1147 bytes --]

Le ven. 18 sept. 2020 à 06:04, Richard Stallman <rms@gnu.org> a écrit :

> [[[ To any NSA and FBI agents reading my email: please consider    ]]]
> [[[ whether defending the US Constitution against all enemies,     ]]]
> [[[ foreign or domestic, requires you to follow Snowden's example. ]]]
>
>   > > +  On a graphics terminal, the easiest way to invoke @code{undo} is
>   > > +with @kbd{C-/}; that doesn't need the shift key.  On a text
> terminal,
>   > > +@kbd{C-/} does not exist, but usually you can type @kbd{C-_} without
>   > > +the shift key and it will work anyway.
>   > > +
>
>   > Does this only make sense with a qwerty keyboard?
>
> I don't think the keyboard _layout_ makes a difference.  An ASCII-only
> Dvorak
> keyboard would presumably support the same control characters.
>
> What matters is whether the keyboard uses ASCII or some wider character
> set.
>

If that's the case, that restriction should be made clear. A lot of people
are using non-ascii keyboards. I agree that we can't expect the bindings to
be convenient for all keyboards, but the documentation at least can be
universal.

>

[-- Attachment #2: Type: text/html, Size: 1718 bytes --]

Re: Patch for emacs/basic.texi

[Mathias Dahl]

[-- Attachment #1: Type: text/plain, Size: 1260 bytes --]

>
>   > > +  On a graphics terminal, the easiest way to invoke @code{undo} is
>>   > > +with @kbd{C-/}; that doesn't need the shift key.  On a text
>> terminal,
>>   > > +@kbd{C-/} does not exist, but usually you can type @kbd{C-_}
>> without
>>   > > +the shift key and it will work anyway.
>>   > > +
>>
>>   > Does this only make sense with a qwerty keyboard?
>>
>> I don't think the keyboard _layout_ makes a difference.  An ASCII-only
>> Dvorak
>> keyboard would presumably support the same control characters.
>>
>
On a Swedish keyboard layout, none of the default bindings for undo is as
convenient as they are when using an English keyboard QWERTY layout.
Regardless of which of the bindings I want to use, I will need to press the
Shift key as well as the Control key, or I have to use the "longest" of the
bindings (C-x u). Same goes for some other bindings that are very
convenient on an English keyboard layout, for example M-/ (dabbrev-expand,
requires the Shift key on a Swedish layout, since they placed / on the same
key as 7 on the number row).

Is this a big problem? Not very. I could rebind these commands if I wanted
to, to something that is convenient to me. And I *can* type the default
bindings as well, both in GUI and terminal Emacs.

[-- Attachment #2: Type: text/html, Size: 1790 bytes --]

Re: Patch for emacs/basic.texi

[Eli Zaretskii]

> From: Richard Stallman <rms@gnu.org>
> Date: Wed, 16 Sep 2020 01:05:54 -0400
> 
> Would someone please install this and ack?

Thanks, I installed this with minor changes (based on subsequent
discussions and comments).