Diredc in the same frame

[BossCode45]

I use a tiling window manager, so opening a new frame doesn't work so well for me. Instead of this it'd be nice if it would just save the current frame's layout and then replace it with diredc, and then when calling diredc again it would go back to the original layout. This could be a variable like diredc-make-new-frame or something, where setting it to false would just use the same frame it was called from.

[Boruch-Baum]

I've pushed several commits to a new branch 'single-frame', which might satisfy your request (and maybe fix another bug in the process). Are you able to test it? Extra points for exhaustive brutal testing. Let me know exactly what you put it through.

[Boruch-Baum]

Four months with no response from the only person who expressed any interest.

[rayvermey]

Hi, i was looking for just this.
I put (setq diredc-make-new-frame nil) in my emacsconfig but still a new frame is started....
Did i do something wrong?
Thanks

Ray

[Boruch-Baum]

Hi. Thanks for the interest. That experimental / in-development feature is sitting off the main branch. You need to switch to branch single-frame in order to access the git version that has it.

[yqu212]

Hi, do you plan to merge the single-frame branch?

[Boruch-Baum]

I fulfilled this request almost a year ago, but no-one, not the requestor nor anyone else, ever collaborated with testing or even verification.

…

On 2024-12-08 01:14, yqu212 wrote: Do you plan to merge the single-frame branch? — Reply to this email directly, [1]view it on GitHub, or [2]unsubscribe. You are receiving this because you modified the open/close state. Message ID: ***@***.***> References 1. #10 (comment) 2. https://github.com/notifications/unsubscribe-auth/AAOE3KGW5RSGGN2OUZNM6BT2EQEYVAVCNFSM6AAAAAAYXTFPU2VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDKMRVGU2TENRVGA

-- hkp://keys.gnupg.net CA45 09B5 5351 7C11 A9D1 7286 0036 9E45 1595 8BC0

[yqu212]

I understand how you feel. I have tested the single-frame feature and have not found any issues so far. Diredc is exactly the feature I need in emacs. I have been a long-time user of Total Commander. Thank you for your work. I'm very appreciated.

[Boruch-Baum]

Thanks. Good to hear. I have some questions if you don't mind. On which window managers have you tested the feature? Have you played with manipulating the tile (eg. resize, move to different pane, move to different window)? How does it react to the system font being changed or resized? If I do get substantial feedback over time, I'll be glad to merge the feature. I mean, that is why I did the work in the first place, right? I just don't want trouble due to insufficient testing.

…

On 2024-12-08 07:43, yqu212 wrote: I understand how you feel. I have tested the single-frame feature and have not found any issues so far. Diredc is exactly the feature I need in emacs. I have been a long-time user of Total Commander. Thank you for your work. I'm very appreciated. — Reply to this email directly, [1]view it on GitHub, or [2]unsubscribe. You are receiving this because you modified the open/close state. Message ID: ***@***.***> References 1. #10 (comment) 2. https://github.com/notifications/unsubscribe-auth/AAOE3KH34UXHEV6GJ6PAXIL2ERSK5AVCNFSM6AAAAAAYXTFPU2VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDKMRWGE4DQNBRGE

-- hkp://keys.gnupg.net CA45 09B5 5351 7C11 A9D1 7286 0036 9E45 1595 8BC0

[arialdomartini]

Hi @Boruch-Baum
Sorry for being late to the party. I would also be very interested in reusing the same frame (I do prefer using tab-bars instead of frames).
I volunteer to test it!

So far I'm using this trick. As soon as diredc is run and it creates a frame, the following triggers, that moves everything back to the only frame I have, in a new tab called diredc:

(defun move-frame-to-main-frame-in-new-tab ()
  "Move all windows from current frame to a new tab in the main frame, then delete the current frame."
  (interactive)
  (let* ((src-frame (selected-frame))
         (main-frame (car (cdr (frame-list)))))
    (unless (eq src-frame main-frame)
      (let ((window-state (window-state-get (frame-root-window src-frame) t)))
        (select-frame-set-input-focus main-frame)
        (tab-bar-new-tab)
        (tab-bar-rename-tab "diredc")
        (window-state-put window-state (frame-root-window main-frame) 'safe)
        (delete-frame src-frame)))))


(defun create-diredc-tab-if-needed ()
  (unless (member "diredc" (mapcar #'car (tab-bar-tabs)))
    (move-frame-to-main-frame-in-new-tab)))

(advice-add 'diredc :after #'create-diredc-tab-if-needed)

It's probably a horrible workaround :D

Edit: it's also a buggy code. It creates multiple diredc tabs. I'm fixing it.

[Boruch-Baum]

Sorry for being late to the party.

More pizza for the rest of us...

I would also be very interested in reusing the same frame (I do prefer using tab-bar instead of frames).

Tab-bars sound like a very cool feature. I see that they were introduced in Emacs 27, released in August 2020, which would have been after I had written this package (github says that I published diredc in November 2020). It could well be that had the feature been around earlier I would have used it instead of using a dedicated frame. What would be the advantages to you of a diredc tab-bar over a diredc frame?

I volunteer to test it!

OK. The branch 'single-frame' is still out there on github. It's now 87 commits behind the 'main' branch, but that point was still a feature-rich and stable snapshot, so it should be fine for testing. If you want, we could also continue to discuss the notion of implementing diredc as a dedicated tab-bar.

So far I'm using this trick. As soon as diredc is run and it creates a frame, the following triggers, that moves everything back to the only frame I have, in a new tab called diredc: It's probably a horrible workaround :D Edit: it's also a buggy code. It creates multiple diredc tabs. I'm fixing it.

Your code isn't checking for the existence of a diredc tab before creating a new one. Also, there are points in the diredc code where the diredc frame, window configuration, and buffers are checked / verified / fixed, so those points might also need to be addressed to polish your approach. The code even has a feature to try to intelligently recover a messed up diredc frame (diredc-recover).

…

-- hkp://keys.gnupg.net CA45 09B5 5351 7C11 A9D1 7286 0036 9E45 1595 8BC0

[Boruch-Baum]

@yqu212 : Hi! Do you have any follow-up to report or add about single-frame diredc? The thread discussion that you contributed to has new posts.

[arialdomartini]

What would be the advantages to you of a diredc tab-bar over a diredc
frame?

At work, alas, I have to use Windows. Jumping from a frame to another
with Windows is not as convenient as it is with i3.

Generally, though, on Linux too I prefer to have 1 single frame for Emacs,
and to move from sets of buffers either using tab-bar or registers (C-x r w).
It's really a personal taste, and by no means I claim it's a better choice than
other ones. After all, Emacs is about freedom of customization.
Tabs have names, so I can use fuzzy search. Using multiple frames and using wm
to move would mix Emacs frames with other programs. I just find tabs and registers
more convenient, so I can dedicate i3 keybindings exclusively to non-emacs activities.

Besides my personal, debatable tastes, I like when packages leave me in control
how to display things, either through display-buffer-alist or via dedicated
variables, such magit does with magit-display-buffer-function.

OK. The branch 'single-frame' is still out there on github.

Cool! I'm giving it a try. Thank you.

If you want, we could also continue to discuss the notion of
implementing diredc as a dedicated tab-bar.

I am not sure if it is even possible, but honestly I think diredc should not
impose any way to display itself.
Ideally, diredc could leave the user the chance to define their
preference using display-buffer-alist or some other configuration. I have no
idea how feasible or hard to develop this could be.