Adding missing C-x 5 C-j and C-x t C-j commands

Adding missing C-x 5 C-j and C-x t C-j commands

[Sean Whitton]

[-- Attachment #1: Type: text/plain, Size: 264 bytes --]

We have C-x d, C-x 4 d, C-x 5 d and C-x t d, but only C-x C-j
and C-x 4 C-j.  My fingers try to use C-x 5 C-j, and find it isn't
there, at least biweekly.  Here is a draft patch to fill the gap.

C-x t C-j temporarily doesn't work due to #55582.

-- 
Sean Whitton

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #2: 0001-Add-dired-jump-other-frame-and-dired-jump-other-tab.patch --]
[-- Type: text/x-diff, Size: 7690 bytes --]

From 2b5ced057a959ab51387ac5d45d2a7b10a414591 Mon Sep 17 00:00:00 2001
From: Sean Whitton <spwhitton@spwhitton.name>
Date: Sun, 22 May 2022 12:08:54 -0700
Subject: [PATCH] Add dired-jump-other-frame and dired-jump-other-tab

* lisp/bindings.el: Drop binding C-x C-j and C-x 4 C-j.
* lisp/dired.el: Bind C-x C-j, C-x 4 C-j, C-x 5 C-j and C-x t C-j
using autoload cookies.  (This parallels the way in which C-x d,
C-x 4 d, C-x 5 d and C-x t d are bound.)
(dired-up-directory): When optional argument is a function, switch to
the buffer using that function.
(dired-jump): Refactor to call dired-jump--internal.
(dired-jump-other-window): Refactor to call dired-jump--internal and
add "... in other window" to read-file-name prompt.
(dired-jump-other-frame, dired-jump-other-tab): New commands.
(dired-jump--internal): New function factored out of dired-jump.
* etc/NEWS:
* doc/emacs/dired.texi (Entering Dired): Document the change.
---
 doc/emacs/dired.texi |  9 +++++-
 etc/NEWS             |  4 +++
 lisp/bindings.el     |  3 --
 lisp/dired.el        | 66 ++++++++++++++++++++++++++++++++------------
 4 files changed, 60 insertions(+), 22 deletions(-)

diff --git a/doc/emacs/dired.texi b/doc/emacs/dired.texi
index ed4ff5213f..5a95d6e8e1 100644
--- a/doc/emacs/dired.texi
+++ b/doc/emacs/dired.texi
@@ -111,8 +111,12 @@ Dired Enter
 
 @findex dired-jump
 @findex dired-jump-other-window
+@findex dired-jump-other-frame
+@findex dired-jump-other-tab
 @kindex C-x C-j
 @kindex C-x 4 C-j
+@kindex C-x 5 C-j
+@kindex C-x t C-j
   You can ask Emacs to invoke Dired on the default-directory
 (@pxref{File Names, default-directory}) of any buffer, by typing
 @kbd{C-x C-j} (@code{dired-jump}).  If the buffer visits a file, this
@@ -123,7 +127,10 @@ Dired Enter
 and places point on the line that corresponds to the directory where
 you invoked @code{dired-jump}.  Typing @kbd{C-x 4 C-j}
 (@code{dired-jump-other-window}) has the same effect, but displays the
-Dired buffer in a new window.
+Dired buffer in a new window; similarly @kbd{C-x 5 C-j}
+(@code{dired-jump-other-frame}) to display the Dired buffer in a new
+frame and @kbd{C-x t C-j} (@code{dired-jump-other-tab}) to display the
+Dired buffer in a new tab.
 
   The variable @code{dired-listing-switches} specifies the options to
 give to @command{ls} for listing the directory; this string
diff --git a/etc/NEWS b/etc/NEWS
index 190620619f..31105bfbec 100644
--- a/etc/NEWS
+++ b/etc/NEWS
@@ -1493,6 +1493,10 @@ If non-nil (which is the default), hitting 'RET' or 'mouse-1' on
 the directory components at the directory displayed at the start of
 the buffer will take you to that directory.
 
++++
+*** New commands 'dired-jump-other-frame' and 'dired-jump-other-tab'.
+These are bound to 'C-x 5 C-j' and 'C-x t C-j' respectively.
+
 ** Exif
 
 ---
diff --git a/lisp/bindings.el b/lisp/bindings.el
index ed1325e326..4ed197f1d4 100644
--- a/lisp/bindings.el
+++ b/lisp/bindings.el
@@ -1501,9 +1501,6 @@ esc-map
 (define-key ctl-x-map "'" 'expand-abbrev)
 (define-key ctl-x-map "\C-b" 'list-buffers)
 
-(define-key ctl-x-map "\C-j" 'dired-jump)
-(define-key ctl-x-4-map "\C-j" 'dired-jump-other-window)
-
 (define-key ctl-x-map "z" 'repeat)
 
 (defvar ctl-x-x-map
diff --git a/lisp/dired.el b/lisp/dired.el
index 89fbd52aa6..d8fa726c6f 100644
--- a/lisp/dired.el
+++ b/lisp/dired.el
@@ -2583,8 +2583,13 @@ dired-up-directory
   "Run Dired on parent directory of current directory.
 Find the parent directory either in this buffer or another buffer.
 Creates a buffer if necessary.
-If OTHER-WINDOW (the optional prefix arg), display the parent
-directory in another window."
+
+When called interactively with a prefix argument, display the
+parent directory in another window.
+
+When called from Lisp, if OTHER-WINDOW is a function, switch to
+the buffer by calling that function.  When OTHER-WINDOW is
+otherwise non-nil, use `switch-to-buffer-other-window'."
   (interactive "P")
   (let* ((dir (dired-current-directory))
 	 (up (file-name-directory (directory-file-name dir))))
@@ -2592,11 +2597,12 @@ dired-up-directory
 	;; Only try dired-goto-subdir if buffer has more than one dir.
 	(and (cdr dired-subdir-alist)
 	     (dired-goto-subdir up))
-	(progn
-	  (if other-window
-	      (dired-other-window up)
-	    (dired--find-possibly-alternative-file up))
-	  (dired-goto-file dir)))))
+        (progn (pcase other-window
+                 ('nil (dired--find-possibly-alternative-file up))
+                 ((or (pred functionp) (pred symbolp))
+                  (funcall other-window (dired-noselect up)))
+                 (_ (dired-other-window up)))
+               (dired-goto-file dir)))))
 
 (defun dired-get-file-for-visit ()
   "Get the current line's file name, with an error if file does not exist."
@@ -4704,6 +4710,7 @@ archive-superior-buffer
 (defvar tar-superior-buffer)
 (declare-function dired-omit-mode "dired-x" (&optional arg))
 
+;;;###autoload (define-key ctl-x-map "\C-j" #'dired-jump)
 ;;;###autoload
 (defun dired-jump (&optional other-window file-name)
   "Jump to Dired buffer corresponding to current buffer.
@@ -4723,6 +4730,39 @@ dired-jump
   (interactive
    (list nil (and current-prefix-arg
                   (read-file-name "Jump to Dired file: "))))
+  (dired-jump--internal (if other-window
+                            #'switch-to-buffer-other-window
+                          #'pop-to-buffer-same-window)
+                        file-name))
+
+;;;###autoload (define-key ctl-x-4-map "\C-j" #'dired-jump-other-window)
+;;;###autoload
+(defun dired-jump-other-window (&optional file-name)
+  "Like \\[dired-jump] (`dired-jump') but in other window."
+  (interactive
+   (list (and current-prefix-arg
+              (read-file-name "Jump to Dired file in other window: "))))
+  (dired-jump--internal #'switch-to-buffer-other-window file-name))
+
+;;;###autoload (define-key ctl-x-5-map "\C-j" #'dired-jump-other-frame)
+;;;###autoload
+(defun dired-jump-other-frame (&optional file-name)
+  "Like \\[dired-jump] (`dired-jump') but in other frame."
+  (interactive
+   (list (and current-prefix-arg
+              (read-file-name "Jump to Dired file in other frame: "))))
+  (dired-jump--internal #'switch-to-buffer-other-frame file-name))
+
+;;;###autoload (define-key tab-prefix-map "\C-j" #'dired-jump-other-tab)
+;;;###autoload
+(defun dired-jump-other-tab (&optional file-name)
+  "Like \\[dired-jump] (`dired-jump') but make new tab."
+  (interactive
+   (list (and current-prefix-arg
+              (read-file-name "Jump to Dired file in other tab: "))))
+  (dired-jump--internal #'switch-to-buffer-other-tab file-name))
+
+(defun dired-jump--internal (other-window file-name)
   (cond
    ((and (bound-and-true-p archive-subfile-mode)
          (buffer-live-p archive-superior-buffer))
@@ -4746,9 +4786,7 @@ dired-jump
               ;; refresh and try again
               (dired-insert-subdir (file-name-directory dir))
               (dired-goto-file dir)))
-        (if other-window
-            (dired-other-window dir)
-          (dired dir))
+        (funcall other-window (dired-noselect dir))
         (if file
             (or (dired-goto-file file)
                 ;; refresh and try again
@@ -4760,14 +4798,6 @@ dired-jump
                   (dired-omit-mode)
                   (dired-goto-file file)))))))))
 
-;;;###autoload
-(defun dired-jump-other-window (&optional file-name)
-  "Like \\[dired-jump] (`dired-jump') but in other window."
-  (interactive
-   (list (and current-prefix-arg
-	      (read-file-name "Jump to Dired file: "))))
-  (dired-jump t file-name))
-
 (provide 'dired)
 
 (run-hooks 'dired-load-hook)		; for your customizations
-- 
2.30.2

Re: Adding missing C-x 5 C-j and C-x t C-j commands

[Eli Zaretskii]

> From: Sean Whitton <spwhitton@spwhitton.name>
> Date: Sun, 22 May 2022 12:53:25 -0700
> 
> We have C-x d, C-x 4 d, C-x 5 d and C-x t d, but only C-x C-j
> and C-x 4 C-j.  My fingers try to use C-x 5 C-j, and find it isn't
> there, at least biweekly.  Here is a draft patch to fill the gap.

We have the "C-x 5 5" prefix for purposes like this one.

Re: Adding missing C-x 5 C-j and C-x t C-j commands

[Sean Whitton]

Hello,

On Mon 23 May 2022 at 05:24am +03, Eli Zaretskii wrote:

>> From: Sean Whitton <spwhitton@spwhitton.name>
>> Date: Sun, 22 May 2022 12:53:25 -0700
>>
>> We have C-x d, C-x 4 d, C-x 5 d and C-x t d, but only C-x C-j
>> and C-x 4 C-j.  My fingers try to use C-x 5 C-j, and find it isn't
>> there, at least biweekly.  Here is a draft patch to fill the gap.
>
> We have the "C-x 5 5" prefix for purposes like this one.

It's a great fallback, but in this case I don't think one should have to
use it.  For everything else under C-x 4 for which -other-frame and
-other-tab variants make sense, we also have bindings under C-x 5 and
C-x t.  If we miss some of them out because we've since added C-x 5 5,
we are asking the user to keep track of which things under C-x 4 have
bindings under C-x 5 and which don't -- one has to remember that C-x 4
C-j exists but C-x 5 C-j doesn't.  I'd prefer that we don't impose that
cognitive burden on users just because we now have C-x 5 5.

-- 
Sean Whitton

Re: Adding missing C-x 5 C-j and C-x t C-j commands

[Sean Whitton]

Hello,

On Sun 22 May 2022 at 09:57pm -07, Sean Whitton wrote:

> If we miss some of them out because we've since added C-x 5 5,
> we are asking the user to keep track of which things under C-x 4 have
> bindings under C-x 5 and which don't -- one has to remember that C-x 4
> C-j exists but C-x 5 C-j doesn't.

I meant: one has to remember that C-x 5 p and C-x 5 d exist but
C-x 5 C-j doesn't.

-- 
Sean Whitton

Re: Adding missing C-x 5 C-j and C-x t C-j commands

[Juri Linkov]

>> If we miss some of them out because we've since added C-x 5 5,
>> we are asking the user to keep track of which things under C-x 4 have
>> bindings under C-x 5 and which don't -- one has to remember that C-x 4
>> C-j exists but C-x 5 C-j doesn't.
>
> I meant: one has to remember that C-x 5 p and C-x 5 d exist but
> C-x 5 C-j doesn't.

For new commands that don't have a C-x 4 binding
it makes sense to use C-x 5 5, but if a binding
already exists in the C-x 4 keymap, it could be synced
with the C-x 5 keymap for consistency.  And indeed
dired-jump-other-frame is missing among existing keys:

C-x 4 C-f find-file-other-window             C-x 5 C-f find-file-other-frame
C-x 4 C-j dired-jump-other-window            !!!
C-x 4 C-o display-buffer                     C-x 5 C-o display-buffer-other-frame
C-x 4 .   xref-find-definitions-other-window C-x 5 .   xref-find-definitions-other-frame
C-x 4 0   kill-buffer-and-window             C-x 5 0   delete-frame
C-x 4 4   other-window-prefix                C-x 5 5   other-frame-prefix
C-x 4 a   add-change-log-entry-other-window
C-x 4 b   switch-to-buffer-other-window      C-x 5 b   switch-to-buffer-other-frame
C-x 4 c   clone-indirect-buffer-other-window C-x 5 c   clone-frame
C-x 4 d   dired-other-window                 C-x 5 d   dired-other-frame
C-x 4 f   find-file-other-window             C-x 5 f   find-file-other-frame
C-x 4 m   compose-mail-other-window          C-x 5 m   compose-mail-other-frame
C-x 4 p   project-other-window-command       C-x 5 p   project-other-frame-command
C-x 4 r   find-file-read-only-other-window   C-x 5 r   find-file-read-only-other-frame

Re: Adding missing C-x 5 C-j and C-x t C-j commands

[Eli Zaretskii]

On May 23, 2022 10:56:47 AM GMT+03:00, Juri Linkov <juri@linkov.net> wrote:
> >> If we miss some of them out because we've since added C-x 5 5,
> >> we are asking the user to keep track of which things under C-x 4 have
> >> bindings under C-x 5 and which don't -- one has to remember that C-x 4
> >> C-j exists but C-x 5 C-j doesn't.
> >
> > I meant: one has to remember that C-x 5 p and C-x 5 d exist but
> > C-x 5 C-j doesn't.
> 
> For new commands that don't have a C-x 4 binding
> it makes sense to use C-x 5 5, but if a binding
> already exists in the C-x 4 keymap, it could be synced
> with the C-x 5 keymap for consistency.  And indeed
> dired-jump-other-frame is missing among existing keys:
> 
> C-x 4 C-f find-file-other-window             C-x 5 C-f find-file-other-frame
> C-x 4 C-j dired-jump-other-window            !!!
> C-x 4 C-o display-buffer                     C-x 5 C-o display-buffer-other-frame
> C-x 4 .   xref-find-definitions-other-window C-x 5 .   xref-find-definitions-other-frame
> C-x 4 0   kill-buffer-and-window             C-x 5 0   delete-frame
> C-x 4 4   other-window-prefix                C-x 5 5   other-frame-prefix
> C-x 4 a   add-change-log-entry-other-window
> C-x 4 b   switch-to-buffer-other-window      C-x 5 b   switch-to-buffer-other-frame
> C-x 4 c   clone-indirect-buffer-other-window C-x 5 c   clone-frame
> C-x 4 d   dired-other-window                 C-x 5 d   dired-other-frame
> C-x 4 f   find-file-other-window             C-x 5 f   find-file-other-frame
> C-x 4 m   compose-mail-other-window          C-x 5 m   compose-mail-other-frame
> C-x 4 p   project-other-window-command       C-x 5 p   project-other-frame-command
> C-x 4 r   find-file-read-only-other-window   C-x 5 r   find-file-read-only-other-frame
> 
> 

I don't think we agreed to have full symmetry between C-x 4 and C-x 5.  I see no special reason for 100% consistency here.  Simple key sequences are at a premium, and we shouldn't usurp them just because we can.

Any user who wants a C-x 5 binding for dired-jump can have such a binding for him or herself.  But why do this by default for everyone?

Re: Adding missing C-x 5 C-j and C-x t C-j commands

[Stefan Monnier]

>> We have the "C-x 5 5" prefix for purposes like this one.
> It's a great fallback,

It shouldn't be a fallback.  It should be "the new C-x 5" and it should
hopefully completely replace `C-x 5` at some point.


        Stefan

Re: Adding missing C-x 5 C-j and C-x t C-j commands

[Sean Whitton]

Hello,

On Mon 23 May 2022 at 10:56am +03, Juri Linkov wrote:

>>> If we miss some of them out because we've since added C-x 5 5,
>>> we are asking the user to keep track of which things under C-x 4 have
>>> bindings under C-x 5 and which don't -- one has to remember that C-x 4
>>> C-j exists but C-x 5 C-j doesn't.
>>
>> I meant: one has to remember that C-x 5 p and C-x 5 d exist but
>> C-x 5 C-j doesn't.
>
> For new commands that don't have a C-x 4 binding
> it makes sense to use C-x 5 5, but if a binding
> already exists in the C-x 4 keymap, it could be synced
> with the C-x 5 keymap for consistency.  And indeed
> dired-jump-other-frame is missing among existing keys:
>
> C-x 4 C-f find-file-other-window             C-x 5 C-f find-file-other-frame
> C-x 4 C-j dired-jump-other-window            !!!

Thank you for the analysis, I hadn't realised that C-x 5 C-j is
literally the only one that's missing!

-- 
Sean Whitton

Re: Adding missing C-x 5 C-j and C-x t C-j commands

[Sean Whitton]

Hello,

On Mon 23 May 2022 at 12:16pm +03, Eli Zaretskii wrote:

> I don't think we agreed to have full symmetry between C-x 4 and C-x 5.  I see no special reason for 100% consistency here.  Simple key sequences are at a premium, and we shouldn't usurp them just because we can.
>
> Any user who wants a C-x 5 binding for dired-jump can have such a binding for him or herself.  But why do this by default for everyone?

I certainly agree with you that we shouldn't bind things into C-x 5
willy-nilly.  In this case, however, it's not just because we can, but
because it makes things consistent with C-x 4 in a way that's helpful.
As Juri has determined, it's almost the only one that doesn't match
right now.

Here's a different way to look at it.  Given the existing symmetry
between C-x 4 and C-x 5, if we later bind anything *other* than
dired-jump-other-window to C-x 5 C-j, we would be introducing a special
exception that users would have to remember.  That is, the existing
symmetry has already implicitly semi-reserved C-x 5 C-j for
dired-jump-other-window, by giving us a strong reason not to put
anything else there.

So, given that I'm unlikely to be the only person who has naturally
tried to reach for C-x 5 C-j, let's make the implicit semi-reservation
explicit.  This command is as useful as C-x 4 C-j.

-- 
Sean Whitton

Re: Adding missing C-x 5 C-j and C-x t C-j commands

[Eli Zaretskii]

> From: Sean Whitton <spwhitton@spwhitton.name>
> Date: Mon, 23 May 2022 06:54:20 -0700
> 
> I certainly agree with you that we shouldn't bind things into C-x 5
> willy-nilly.  In this case, however, it's not just because we can, but
> because it makes things consistent with C-x 4 in a way that's helpful.

My point is that it may be useful for you, but is not necessarily
useful enough for others to justify a global "C-x 5" binding.

> As Juri has determined, it's almost the only one that doesn't match
> right now.

"C-x 4 a" is also unpaired, as are "C-x 5 u", "C-x 5 2", and "C-x 5 o".
That's hardly "almost".

> Here's a different way to look at it.  Given the existing symmetry
> between C-x 4 and C-x 5, if we later bind anything *other* than
> dired-jump-other-window to C-x 5 C-j, we would be introducing a special
> exception that users would have to remember.  That is, the existing
> symmetry has already implicitly semi-reserved C-x 5 C-j for
> dired-jump-other-window, by giving us a strong reason not to put
> anything else there.

See above: we already have asymmetry.

Re: Adding missing C-x 5 C-j and C-x t C-j commands

[Juri Linkov]

>>> We have the "C-x 5 5" prefix for purposes like this one.
>> It's a great fallback,
>
> It shouldn't be a fallback.  It should be "the new C-x 5" and it should
> hopefully completely replace `C-x 5` at some point.

Should the prefix `C-x 5` take keys from the `C-x` keymap?
So any key in the `C-x` keymap will be automatically
available in a key sequence starting with `C-x 5`?

For example, `C-x C-d` is bound to `list-directory`,
then the inferred `C-x 5 C-d` will show the output
of `list-directory` in a new frame?

Re: Adding missing C-x 5 C-j and C-x t C-j commands

[Sean Whitton]

Hello,

On Mon 23 May 2022 at 05:07pm +03, Eli Zaretskii wrote:

>> From: Sean Whitton <spwhitton@spwhitton.name>
>> Date: Mon, 23 May 2022 06:54:20 -0700
>>
>> I certainly agree with you that we shouldn't bind things into C-x 5
>> willy-nilly.  In this case, however, it's not just because we can, but
>> because it makes things consistent with C-x 4 in a way that's helpful.
>
> My point is that it may be useful for you, but is not necessarily
> useful enough for others to justify a global "C-x 5" binding.
>
>> As Juri has determined, it's almost the only one that doesn't match
>> right now.
>
> "C-x 4 a" is also unpaired, as are "C-x 5 u", "C-x 5 2", and "C-x 5 o".
> That's hardly "almost".

I wrote "almost" because only C-x 4 a and C-x 4 C-j satisfy both of the
following conditions:

  1. Obvious -other-window and -other-frame variants exist
  2. There is already a binding under C-x 4.

For commands satisfying these two conditions, symmetry is worthwhile --
otherwise, we're asking the user to remember which of them is missing.
All of C-x 5 u, C-x 5 2, and C-x 5 o fail condition (1).

In addition, I think the existing symmetry makes it such that binding
anything else to C-x 5 C-j would be hard for people to remember --
that's what I mean by an implicit semi-reservation.  If I'm right that
the only thing we would want to bind there is what I'm proposing to bind
there, then it makes sense to do so for everyone.

-- 
Sean Whitton

Re: Adding missing C-x 5 C-j and C-x t C-j commands

[Howard Melman]

Juri Linkov <juri@linkov.net> writes:

>>> We have the "C-x 5 5" prefix for purposes like this one.
>>> It's a great fallback,
>>
>> It shouldn't be a fallback.  It should be "the new C-x 5" and it should
>> hopefully completely replace `C-x 5` at some point.
>
> Should the prefix `C-x 5` take keys from the `C-x` keymap?
> So any key in the `C-x` keymap will be automatically
> available in a key sequence starting with `C-x 5`?
>
> For example, `C-x C-d` is bound to `list-directory`,
> then the inferred `C-x 5 C-d` will show the output
> of `list-directory` in a new frame?

I have a question the other way.  I forgot that `C-x 5 .'
exists and tried `C-x 5 5 M-.' and was surprised to find it
didn't work (it opened in the same frame as if I just did
`M-.'.  Is it supposed to work or should I open a bug?

-- 

Howard

Re: Adding missing C-x 5 C-j and C-x t C-j commands

[Eli Zaretskii]

> From: Howard Melman <hmelman@gmail.com>
> Date: Tue, 07 Jun 2022 16:17:29 -0400
> 
> Juri Linkov <juri@linkov.net> writes:
> 
> >>> We have the "C-x 5 5" prefix for purposes like this one.
> >>> It's a great fallback,
> >>
> >> It shouldn't be a fallback.  It should be "the new C-x 5" and it should
> >> hopefully completely replace `C-x 5` at some point.
> >
> > Should the prefix `C-x 5` take keys from the `C-x` keymap?
> > So any key in the `C-x` keymap will be automatically
> > available in a key sequence starting with `C-x 5`?
> >
> > For example, `C-x C-d` is bound to `list-directory`,
> > then the inferred `C-x 5 C-d` will show the output
> > of `list-directory` in a new frame?
> 
> I have a question the other way.  I forgot that `C-x 5 .'
> exists and tried `C-x 5 5 M-.' and was surprised to find it
> didn't work (it opened in the same frame as if I just did
> `M-.'.  Is it supposed to work or should I open a bug?

I suggest to file a bug report.  If this is supposed to work the way
it does, someone will explain there why.

Thanks.

Re: Adding missing C-x 5 C-j and C-x t C-j commands

[Juri Linkov]

> I have a question the other way.  I forgot that `C-x 5 .'
> exists and tried `C-x 5 5 M-.' and was surprised to find it
> didn't work (it opened in the same frame as if I just did
> `M-.'.  Is it supposed to work or should I open a bug?

It does the right thing after customizing
switch-to-buffer-obey-display-actions to t.

Maybe its default value should be changed to t?

Re: Adding missing C-x 5 C-j and C-x t C-j commands

[Howard Melman]

On Jun 8, 2022, at 2:40 AM, Juri Linkov <juri@linkov.net> wrote:
> 
>> I have a question the other way.  I forgot that `C-x 5 .'
>> exists and tried `C-x 5 5 M-.' and was surprised to find it
>> didn't work (it opened in the same frame as if I just did
>> `M-.'.  Is it supposed to work or should I open a bug?
> 
> It does the right thing after customizing
> switch-to-buffer-obey-display-actions to t.

Yes I see that too.

> Maybe its default value should be changed to t?

At least the docstring of other-frame-prefix (and friends) should
be clearer about it's effect. I don't think I'm supposed to know 
which commands use switch-to-buffer internally.

If the default value isn't changed, maybe other-frame-prefix
should let bind it to t since if you're using it you presumably
want a new frame.  Same for the window and tab variants.
project.el seems to do this in project--other-place-command 
which the project-other-{frame,window,tab}-commands use.

Howard

Re: Adding missing C-x 5 C-j and C-x t C-j commands

[Juri Linkov]

[-- Attachment #1: Type: text/plain, Size: 1376 bytes --]

>>> I have a question the other way.  I forgot that `C-x 5 .'
>>> exists and tried `C-x 5 5 M-.' and was surprised to find it
>>> didn't work (it opened in the same frame as if I just did
>>> `M-.'.  Is it supposed to work or should I open a bug?
>>
>> It does the right thing after customizing
>> switch-to-buffer-obey-display-actions to t.
>
> Yes I see that too.
>
>> Maybe its default value should be changed to t?
>
> At least the docstring of other-frame-prefix (and friends) should
> be clearer about it's effect. I don't think I'm supposed to know
> which commands use switch-to-buffer internally.

Maybe more commands should use pop-to-buffer-same-window
instead of switch-to-buffer.  But OTOH, too much commands
already use switch-to-buffer.

> If the default value isn't changed, maybe other-frame-prefix
> should let bind it to t since if you're using it you presumably
> want a new frame.  Same for the window and tab variants.
> project.el seems to do this in project--other-place-command
> which the project-other-{frame,window,tab}-commands use.

I'm not sure if the default value can be changed,
but you are right about other-frame-prefix.
I didn't know that project--other-place-command
let binds it to t.  But doing the same is not possible
in other-frame-prefix, that just sets a display action
for the next command.  So a more elaborate fix is required:


[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #2: obey-display.patch --]
[-- Type: text/x-diff, Size: 1294 bytes --]

diff --git a/lisp/window.el b/lisp/window.el
index 1b8fe2b262..eba888a89d 100644
--- a/lisp/window.el
+++ b/lisp/window.el
@@ -8931,6 +8931,7 @@ display-buffer-override-next-command
   (let* ((old-window (or (minibuffer-selected-window) (selected-window)))
          (new-window nil)
          (minibuffer-depth (minibuffer-depth))
+         (obey-display switch-to-buffer-obey-display-actions)
          (clearfun (make-symbol "clear-display-buffer-overriding-action"))
          (postfun (make-symbol "post-display-buffer-override-next-command"))
          (action (lambda (buffer alist)
@@ -8955,6 +8956,7 @@ display-buffer-override-next-command
               (funcall post-function old-window new-window)))))
     (fset clearfun
           (lambda ()
+            (setq switch-to-buffer-obey-display-actions obey-display)
             (setcar display-buffer-overriding-action
                     (delq action (car display-buffer-overriding-action)))))
     (fset postfun
@@ -8971,6 +8973,7 @@ display-buffer-override-next-command
     (add-hook 'post-command-hook postfun)
     (when echofun
       (add-hook 'prefix-command-echo-keystrokes-functions echofun))
+    (setq switch-to-buffer-obey-display-actions t)
     (push action (car display-buffer-overriding-action))
     exitfun))
 

Re: Adding missing C-x 5 C-j and C-x t C-j commands

[Howard Melman]

On Jun 8, 2022, at 12:08 PM, Juri Linkov <juri@linkov.net> wrote:
> 
> I'm not sure if the default value can be changed,
> but you are right about other-frame-prefix.
> I didn't know that project--other-place-command
> let binds it to t.  But doing the same is not possible
> in other-frame-prefix, that just sets a display action
> for the next command.  So a more elaborate fix is required:

Seems to work for me.  Any chance this could make it into a 28.2?

Howard

Re: Adding missing C-x 5 C-j and C-x t C-j commands

[Howard Melman]

Howard Melman <hmelman@gmail.com> writes:

> On Jun 8, 2022, at 12:08 PM, Juri Linkov <juri@linkov.net> wrote:
>> 
>> I'm not sure if the default value can be changed,
>> but you are right about other-frame-prefix.
>> I didn't know that project--other-place-command
>> let binds it to t.  But doing the same is not possible
>> in other-frame-prefix, that just sets a display action
>> for the next command.  So a more elaborate fix is required:
>
> Seems to work for me.  Any chance this could make it into a 28.2?

Also, should the equivalent be done for other-window-prefix
and other-tab-prefix?

-- 

Howard

Re: Adding missing C-x 5 C-j and C-x t C-j commands

[Eli Zaretskii]

> From: Howard Melman <hmelman@gmail.com>
> Date: Wed, 8 Jun 2022 15:17:02 -0400
> Cc: emacs-devel@gnu.org
> 
> On Jun 8, 2022, at 12:08 PM, Juri Linkov <juri@linkov.net> wrote:
> > 
> > I'm not sure if the default value can be changed,
> > but you are right about other-frame-prefix.
> > I didn't know that project--other-place-command
> > let binds it to t.  But doing the same is not possible
> > in other-frame-prefix, that just sets a display action
> > for the next command.  So a more elaborate fix is required:
> 
> Seems to work for me.  Any chance this could make it into a 28.2?

No, it's too late for adding such non-trivial behavior changes.  We
could have unintended consequences, and we don't want to risk that on
the release branch.

Sorry.

Re: Adding missing C-x 5 C-j and C-x t C-j commands

[Juri Linkov]

>> Seems to work for me.  Any chance this could make it into a 28.2?
>
> No, it's too late for adding such non-trivial behavior changes.  We
> could have unintended consequences, and we don't want to risk that on
> the release branch.

Therefore pushed to master.

Re: Adding missing C-x 5 C-j and C-x t C-j commands

[Juri Linkov]

>> Seems to work for me.  Any chance this could make it into a 28.2?

In 28.2 you can just customize switch-to-buffer-obey-display-actions to t.

> Also, should the equivalent be done for other-window-prefix
> and other-tab-prefix?

Actually the same change should handle other-window/tab as well.