Customizing tmux
evgeniipendragon.com
evgeniipendragon.com
I was inspired by the "How to Configure tmux from scratch" post [1].
I came up with my use cases:
- I want to create sessions
- I want to open and close windows
- I want to create split panes
- I want to use vim-style text select
- etc.
Then I made mappings for the things I care about for my workflow specifically.
Before this, I would accidentally hit a keybinding while doing something else and not know how I got to that state, taking me out of flow to troubleshoot.
Afterwards, only the keybindings I have defined take action.
Now tmux fits like a glove. Because tmux is so stable, I haven't had to touch my config in years. It's worth the one time effort.
I used to accidentally background my tmux frequently and the first couple times made my heart sink in that “did I save??” way lol, though these days I’ve largely stoped using it since splits feel smoother in Kitty and I don’t do any remote sessions.
EPendragon elaborate on this. It seems like you understand the shell enough to use it productively and customize it to be how you like it. The screenshot this quote is talking about looks like a shell? In what way is it dreadful, and in what way do your customizations make it not dreadful?
For example, it seems like a valid criticism that shells are not very discoverable and a prompt and blinking cursor can be intimidating to a noob, and that surely applies to tmux as well. But then changing the shortcuts doesn't really address that, does it?
To me it seems your article is either saying, "tmux is a pig, and here's me putting some lipstick on it," :) or the problems are not insurmountable, and the solutions are nice but not life-changing.
It's great that you like it better now, and we're all having fun chatting around it about how we configure our tools here on the Hacker News :) But I don't see why it's "dreadful". It's just.. computers, no? Tmux seems no more gatekeepy than any other aspect of computers. And I say this as someone who stopped using tmux a few years ago after getting super annoyed with it.
The initial experience where I saw a lot of elements on the screen without knowing what's what was a lot. And sure enough - a lot of it had to do with just learning how it works and what's what. But in addition to the UI that is unintuitive for the beginner, most of the keybindings either didn't make sense or were not convenient (talking about prefix).
For example, splitting panes in my opinion are much better done with | and - symbols rather than " and %. I can see what they were going for with double quotes and the percent sign, but IMO vertical line and a hyphen make more sense. I do not need to be actively thinking about which one should I use for a vertical vs horizontal split.
There was some aspect of "putting a lipstick on it" - for sure. Making it more pleasant to look at didn't improve function - just looks. I just think making your tools look the way you like and enjoy them will create a more pleasant experience and help you learn the tool faster too.
The dreadfulness did come from not understanding it completely and being a noob early on. Fixing the style and function also put me into a process of being intentional about learning how the tool works and adjusting things that I didn't like.
And you're right - it is not more gatekeepy that other tools per se. I was simply describing my experience and what I ended doing about that.
What were the "lot of elements" in the screenshot? Ignoring shell stuff, is it the bar at the bottom? Is it the "[0]" and the "0:bash*"? The "Macbook-pro" might seem familiar to someone familiar to the shell, who is used to seeing it in the prompt of an uncustomized shell. The date and time likewise seem obvious what they mean. Am I missing anything?
I think the keybindings being non-mnemonic there's only so much you can do about. Vim's hjkl were mnemonic on the VT-100 terminal, but keybindings outlast mortal hardware. So the ability to rebind is in some ways the best you can get IMO.
Zooming out, the best power UI I've been able to come up with is a combination of a Wordstar-like always-on menu showing common keybindings, and a command palette. Here are a couple of UIs I've made, I'm curious what you think of them:
https://merveilles.town/@akkartik/108766067153506592
https://github.com/akkartik/mu/blob/main/html/20210624-shell...
If anyone is curious, there is a Linux port of Lotus123:
The bottom elements were part of it. Not the host name and time, but the window information. Also, it took a while to remember the default keys to use to get to session management window where tmux displays everything, among other things.
Vim's hjkl make sense as you start using the tool. tmux's Ctrl-B for a prefix doesn't make sense, so that leads to remapping.
I have checked out the UI's you have made - they look great - similar to what I would expect a custom tmux configuration to look like :)
Hmm, I'm still confused:
* How do Vim's hjkl start to make sense? They've always seemed arbitrary. 'l' should stand for 'left', but it takes me right!
* How is a different keyboard shortcut less confusing than ctrl-b. They all seem arbitrary. You remapped it to ctrl-a. Maybe you were used to GNU Screen?
Ctrl-B is a mechanically more difficult key binding than Ctrl-A.
Here is a codepen example
https://codesandbox.io/p/sandbox/yup2o
of a word cloud using @react-three/fiber for example. Imagine if I could have a little cloud floating for navigating tmux. A cloud not of words but sessions, windows, panes or whatever. If I could then I move around the 3d cloud from the keyboard .... Just an idea
In the context of this thread my mind was mostly on what it takes to make some arbitrary power UI "not dreadful", i.e. easy to understand for newcomers. Since it's a power tool, assume there'll be a learning curve. But the initial experience doesn't intimidate someone to start on that journey of a hundred miles of understanding how to use it with fluency and thereby augmenting themselves.
There have certainly been many intuitive 3D UIs. 3D games devote a lot of attention to being discoverable and friendly to newcomers, and they have a lot of success. However, my impression is that they achieve this success by dint of hard work. It's not clear that 3D makes the problem easier. If anything I'd be more likely to believe it makes it harder.
So if you have a product and are thinking about how to make it more intuitive (as I am), "make it 3D" is probably not the best suggestion. Does that seem reasonable? Let me know if I'm missing something.
I like your codepen. I wonder what it would feel like to use a whole note-taking app or digital garden with this metaphor. Rotate to a particular note until it's in front, then click on it to edit it. Perhaps you should stop moving in 3D while editing.
My guess, and only a guess, is that 3D is especially (only?) useful for navigation. As in moving around documents, panes, windows, etc.
As for "I wonder what it would feel like to use a whole note-taking app ..." The note taking is in some ways already done pretty well. So my current guess is that it is more the navigation that would benefit.
I don't really have answers, but I often feel like I live in 'flatland' these days.
Especially something like keyboard shortcuts and leader keys doesn't make sense to change in my mind. It just confuses you when login in to a remote host.
However, I think that creating a simple git repository with rc files and configuration can be used to create a uniform experience across machines.
Also the machine may not actually able to reach out onto the wider internet.
I do like the repo idea, but I wonder if there’s an easier and maybe faster way to do it over the SSH tunnel
Tmux is more customisable, but a lot messier out the box. It's much older so has picked up a little cruft along the way.
Zellij is newer so has the benefit of being a clearer implementation. But it's designed as an "out the bix" experience, not one you can customise to your hearts content.
Edit: if you did want to try out zellij, I should point out there's a "no install, try in shell" option on their website's main page which is a super quick way to get a taste!
I have definitely felt that. Customization wasn't necessarily straightforward - just required some time spent reading the docs.
I appreciate the heads up about the "no install, try in shell" option. I will make sure to take a look!
Last I tried it the bind issue and performance were sort of blockers for me, but it looks like both of those are moving forward.
I don’t know what it is about how my brain works, but I have absolutely no memory for keybinds (and envy those that do), so I love that it just lays them out in the UI.
Only customisation I make is to turn off the borders.
Would love if built a similar, visually guided, experience for nvim.
Definitely would recommend checking out Helix. My brain is similar in that keybindings often don't stick - Helix has a grammar leading to a bit more of a visual interface (you select things first, and then edit, so you can see how things go as you carry out multi-cursor edits etc), but also, it'll show pop-up menus for shortcuts.
In normal mode for instance, "g" is "goto", pressing g also displays a little window that shows the different options and their keys. It's easy to ignore when you know what your doing, and super helpful when you can't remember something!
Should say that it's really similar to the tmux/zellij distinction, in that, helix works nicely out the box, but is a lot more opinionated, there's LSP support for pretty much all languages you'll use, but no plugin eco-system, so if you have a really customised, or AI based workflow for instance, your kinda stuck.
In particular, I'm curious about libtmux/tmuxp and how well it works to have a more declarative bringup than just a bash script launching things into the background and then attaching tmux windows to them.
And I guess the broader question of whether this approach is an evolutionary dead end— like does it just become a pile of hacks as one discovers the need for restart policies, backoffs, ordering, whatever other stuff that is built into a "normal" process manager like systemd or runit? If you do end up wanting to send process stdout/stderr elsewhere or do alerting, is that now annoying if those streams are "trapped" in tmux sessions vs being written into log files or the journal?
All the other examples I’ve seen where people have used it as a shortcut for daemonizing a service, it’s felt like just as much effort as writing a systemd service file.
I guess I could forgive people more back in the days of sysv-style initd. Writing those shell scripts was seldom fun. But with systemd, it’s literally a 5 minute task. And I say that as someone who dislikes numerous decisions made regarding the approach systemd takes.
It's tougher to make the case there that systemd units are "worth it", particularly given the other ergonomic benefits of a pure tmux solution, such as being able to ctrl-c a pane and trivially re-launch it.
And once you're launching services in a container, that's another world again, since systemd in that case is quite a bit worse and probably wants CAP_ADMIN.
I dont agree that a bunch of services makes things harder if it’s already easier as one service. The set up here is linear in effort. Plus, if anything, being able to restart the host and have all services automatically come up makes systemd much more convenient.
> and then if you want the dashboard experience you still have to launch all the panes and get each one tailing the respective unit journals.
logs are a literal one liner with systemd. Though if you care about dashboards then you should be exporting your logs to a proper observability platform instead of SSHing into your service.
> such as being able to ctrl-c a pane and trivially re-launch it.
systemd restart …
tmux attach …
[prefix]+w (select window)
ctrl+c
[up]
[return]
If you’re launching a service in a container then you don’t need either systemd nor tmux in that container because the container itself provides all that scaffolding for you.
Whether it’s Docker/podman, LXC, FreeBSD jails, Solaris Zones, etc, you just set your service to be the entry point of that container.
Docker makes this even easier still with a literal directive called “entrypoint”.
Trying to start services using tmux inside of a container is massively more effort than calling it directly from the entrypoint.
———
I do get the appeal of using tmux for these usages. I honestly do. And if your services are constant in flux, for example inference where you’re constantly tweaking parameters then tmux does make sense (a proper ML pipeline makes more sense, but that requires knowledgable MLOps to build). However for anything consistent, the effort in doing things “properly” is usually less than you think it is.
Sure! I use tmux on most of my OpenBSD systems and copy the same .tmux.conf file around, but rarely tweak it other than to set a hostname.
My status bar is at the top and looks approximately like this:
--~ msg auth pf hostname. Mon Aug 04 15:05
* The tilde window (~) is where I do stuff as root.
* The message window (msg) tails `/var/log/messages` and `/var/log/daemon`, or `journalctl -f` and `journalctl -k -f` on Linux.
* The auth window (auth) tails `/var/log/authlog` to monitor login attempts and usage of doas and sudo.
* The pf window (pf) runs a script to monitor the packet filter rules (pass, block, match).
* hostname is just the hostname. The period (.) means no mail, it turns to an exclaimation mark (!) when I have mail.
* Simple date/time.
This setup gives me quick access shells, important logs, mail status, date/time, and works the same on every server I manage. I add extra windows to tail other app logs as needed.
In case helpful: You can use `#h` in a status line string to get the hostname
Note that the dashboard is read-only and mostly just shows journalctl for different units (we use systemd for process management). And we assume it's only used by a single user at once and everything is stopped on detach.
Here's a simplified sample of setting up a 3x2 layout with some placeholder content:
#!/bin/sh
set -e
SESSION=example
WINDOW=dashboard
tmux new-session -d -s $SESSION -n $WINDOW "journalctl -fk"
tmux set-option -t $SESSION -g mouse on
tmux split-window -h -t $SESSION:$WINDOW "journalctl -fu nginx"
tmux split-window -h -t $SESSION:$WINDOW "journalctl -fu postgresql"
tmux select-layout -E
tmux split-window -v -t $SESSION:$WINDOW "echo '<Ctrl-b d> to detach' && read"
tmux select-pane -t $SESSION:$WINDOW.0
tmux split-window -v -t $SESSION:$WINDOW "watch uptime"
tmux select-pane -t $SESSION:$WINDOW.2
tmux split-window -v -t $SESSION:$WINDOW "watch df -h"
tmux attach-session -t $SESSION
# after detaching, close session
tmux kill-session -t $SESSION 2>/dev/null || trueWell, they'd still need to bring their own ascii teacup.
It's free https://iterm2.com/
As someone who lives in terminals, iTerm2 may be the Killer App for Macbooks.
You see, it has Tmux Integration [1], which makes tmux windows and panes become native tabs and panes in iTerm, which means I don't need to use weird chords to manipulate tmux, but most importantly, most UX-ily, it has native scrollback. No need to escape into a weird secondary mode just to do the most basic thing that all graphical terminals are good for.
If you work with terminals on remote machines, you owe yourself to use tmux. If your client is a Mac(book), you owe yourself to use iTerm2.
The weird thing is that iTerm2 is the only Terminal Emulator to leverage Tmux's Control Channel mode, for the integration. For all the breadth of copycat terminal emulators on Linux, none of them perform this one killer feature.
Other than that, I always use the default CTRL+B (MOD) and rarely if ever do anything other than the following:
- MOD + d to detach, "tmux a" to attach
- MOD + c to create new window
- MOD + % to create new split pane, " for horizontal
- MOD + space to change between pane layouts
- MOD + pgup to scroll in buffer
Does it ever create issues for you if you need to type the backtick in the terminal?
I found that Ctrl-A does a great job: it is conveniently located in relation to other keys that I need to interact with after I activate the prefix, and is in general easy to use.
I've been using Ctrl+Q, it replaces an almost completely useless key combo (flow control if I'm not mistaken), and is easy to press.
bind C-a send-prefixI also have my own keybindings, styling etc. - but this is an awesome base to get started.
I'm puzzled how we get from unintuitive interface to "gatekeepy" (which I'm assuming implies a deliberate attempt to make the set of users small). Things can be difficult to use without anyone setting out to exclude new users, some things _are_ difficult to use, and other things are written by people who haven't been paid to make them address the needs of a wider range of users than they want to.
Glad that that it supports XDG though.
It's easy to install via uv:
uv tool install claude-code-tools
For those of you who use a tiling or scrolling window manager and tmux, what’s your selling point?
The down side is that it's really sensitive to versions, right now I'm struggling to get it to work on NixOS in the version that I have running on my Ubuntu dev box and MacOS laptop.
But only customization so far is metahar = C-b, because C-a needed on Emacs.
But the after moving to Emacs and using Tramp for remote sessions, followed by VS Code with its built-in session management, I never feel the need to even leave the editor let along manage sessions
Host myhost
Hostname host
User user
RequestTTY yes
RemoteCommand tmux new -A -s foobar alias ssh="ssh -o RequestTTY=force -o RemoteCommand='tmux new -A -s foobar'"Though I agree, probably not the best solution to start all sessions with tmux. There's other issues it can cause.
The only reason to still use tmux (or screen) is because you use remote sessions. All modern terminal emulators are already capable of doing tabs and panes (okay, maybe not sessions, but some can locally). If you're using tmux for this reason, stop. Go pick up a modern terminal like ghostty, Alacritty, or WezTerm.
But why tmux still exists today is because people are working on computers they aren't sitting in front of. Because I don't want to have to be running nohup or detaching, moving to the background, and resetting the session so error messages don't appear in my active instance. Hell, technically I can do this with vim and get something pretty similar to tmux by using the terminal. But that's a pain.
Tmux is for *terminals*
Like you mention, tmux sessions was one of the biggest features that I enjoyed about this tool: it does the job and does it well. Aside from that I love that I can use Vim-bindings with it, that it feels very similar to Neovim for me, so using them together is a no-brainer. Learning a couple of specific keyboard shortcuts is all it takes to make me productive.
I haven't needed to use a remote session yet like you note. While working with tmux locally I have learnt about all of these other features that I will definitely be looking forward to using once the opportunity arises.
I don't doubt that tools like ghostty, Alacritty, or WezTerm are great. I have heard lots about the first two. And I do intend to check them out.
What would you say is your favorite terminal and why as it relates to local development experience use? Are there features that these tools have that tmux lacks?
> What would you say is your favorite terminal
Btw, vim bindings should already be available in your emulator, without the need for tmux. You can set bash to use them (set -o vi) but I suggest migrating to zsh if you like using the cli. I'll also give you one fun command most people don't know. While holding control, press x then e (if you are already using zsh you might need to add `zle -N edit-command-line` to your rc file but <C-x><C-e> needs no config in bash). I guess I should also mention that in vim you can open a terminal (:help term)
> why as it relates to local development experience use?
> Are there features that these tools have that tmux lacks?
Second, tmux lacks many of these modern features. Doing a passthrough can help but dealing with images is not a great experience. I have found no configuration where I can reliably view images and never have been able to produce images of the same quality. I always drop out of my session if I am entering a file browser like yazi or fzf (I'll use fzf but it limits the capabilities of previewing).
I prefer the idea of having one tool to manage everything: from local flow to remote sessions. And I usually do not view images in the terminal. But I'm willing to put them to the test and see if I like one of those more.
> I prefer the idea of having one tool to manage everything: from local flow to remote sessions.
I'll add that your muscle memory will adapt to whatever pretty quickly. Like I was switching between a mac and a linux machine with a keyboard frequently and it only took a week to forget where my pinky needed to go to hit ctrl. Same is true even if your bindings for whatever are not identical local or remote (though mine are very similar, just differing in the prefix). So I'd say when testing things make sure you include a burn-in time, otherwise any change is going to make all other options feel worse.
> I usually do not view images in the terminal.
> vim bindings should already be available in your emulator, without the need for tmux
They may mean things like the vim-tmux navigator plugin, which makes it so if you hit e.g. the left-most window in vim and hit your go-left keybinding, it will go one tmux pane to the left. It's pretty crazy how seamless it is, and it's made possible by the fact that you can query tmux outside of the process and run commands on your running sessions, which I'm not sure you can do with any of the standalone terminals?
Tmux also doesn't prevent you from using c-x e. I'm aware of the vim terminal, but I'm very picky about my multiplexing and have not been satisfied by any built-in terminal in any editor.
> You just have to remember that a session is the same thing as a window.
A session is certainly NOT a window. It is a buffer. There is a difference, and that difference matters and comes with advantages that are separate from the advantages of windows.
> First off, switch because your emulator is probably more resource intensive
I'm not entirely sure if this is true, but it's utterly uncompelling to me. Holding my binding for 'switch window' will switch through n windows faster than I can mentally digest, and the lag is lighter than any gui app I've used.
> It may also be missing modern features like being able to view images (see sixel, chafa, or the kitty graphics protocol), ligatures, and a lot of other features
I'm not sure why you'd assume this? My setup (wezterm or iterm + tmux + fira code) indeed has ligatures and sharp image viewing passed through the multiplexer
> Second, tmux lacks many of these modern features. Doing a passthrough can help but dealing with images is not a great experience. I have found no configuration where I can reliably view images and never have been able to produce images of the same quality. I always drop out of my session if I am entering a file browser like yazi or fzf
Yazi is exactly what I use, and it works great with iTerm + tmux or wezterm + tmux.
I really don't want my terminal to do that much for me to be honest. I change terminals every couple years, and it's nice to not have to fuss around with them whenever that happens. I haven't changed my tmux config in about 15 years
Do you think it's possible that someone might have a different workflow from you and tmux fits their use case in a way you haven't thought of?
That said, I still use tmux. Almost every day in fact. Because all my work is being done on a machine I'm not sitting in front of. This even includes at home. My main desktop is connected to a TV for videogames and movies. When I want to do work on it there's no difference if I'm sitting in front of my laptop or it other than it sitting in my livingroom keeps it cooler and gives it better air flow.
Edit:
Locally: my terminal emulator (ghostty) can do everything tmux can. I can do sessions (windows), panes, tabs, and all that. But with ghostty I also get images, a lower memory footprint (than stock emulator), lower CPU usage (than stock emulator), ligatures, and everything else. It is strictly better.
But I can't do {widows,panes,tabs} with remote connections. Hence, tmux. Which in that case, I will frequently give up capabilities (like images) for that.
You don't need a server to handle this feature. Using a non-server solution even allows restoring after power cycles!
[0] https://ghostty.org/docs/config/reference#window-save-state
BUT
1) I'm unconvinced the things it solves that emulators don't do are common situations
2) tmux isn't "free". There are trade-offs that you have to make for tmux.
But most users are turning to tmux for the {tabs,panes} feature (and then using sessions like anyone would use a window).
> interrupted/killed if you accidentally close your terminal
If you have a process running that needs to stay alive for long periods of time then does it need to be running interactively? It is pretty rare that long running processes need be running interactively. Not rare in the sense that an average user is going to do this infrequently, but rare in the sense of few users are going to do this infrequently.
- You can always throw running jobs into the background by doing: <C-z> (suspend), `bg` (resumes program in background), and `detach` (so the process is not attached to the current terminal instance).
- If you're launching a job consider using `&`
- Remember, you can always reattach to a running process too. If you still have the terminal open it is trivial with `fg`. If you don't, then you can still attach with strace or gdb, but you'll probably need `sudo` and this isn't exactly trivial.
- If you know your job needs to be interacted with then there's better ways to handle that communication like by setting up a pipe. But that requires forethought.
- or the best option is to actually run it as a service since you'll get a lot more benefits and flexibility through that. It's pretty quick to turn anything into a systemd job.
- You can interact with systemd jobs if you set them up that way. Lots of fancy options here from ttys, pipe, sockets, or containerizations.
- systemd-nspawn or systemd-vmspawn. Containers are honestly the best option for long running but interactive jobs.
--- Aside ---
Batch jobs:
script and should run in the background. Are your batched jobs not scripted? Just copy paste your loop into a file and add the header. I typically add a few more lines for logging and stuff but yeah. I mean if I want that output I usually want that output longer than my session will be alive, so logs. That way they also are organized.
Backups:
unless you're doing a quick rsync then this really should be automated. A service is the solution. If you find yourself needing/wanting to manually run, you can also just call the systemd job. But yeah, if I have a one-time big sync, I'll pull up tmux (or caffeine). But that's different than always running in tmux.
Compiling:
I'll be honest, locally I rarely have long compiles (or anything over a few minutes). Anything that is long and repeated should be incrementally compiled. I mean this is a much more robust strategy anyways and in a long compile there's a lot of potential disruptions that are far more likely than me closing my terminal window. But in those rare instances I usually throw the process into the background (or start with `&`) or will turn on caffeine or whatever to keep my machine awake.
--- /Aside ---
> or you terminal/desktop environment crashes/freezes
I'm not sure how often ghostty calls the window-save-state command but it looks like you should be able to call this frequently, though I haven't tried it and I'll give you this one because this definitely requires more work. Though ghostty will survive a reboot and you need a plugin for that with tmux. But either way, this is related to the lesson of backup often. Or commit frequently.
> While essential for doing dev on remote servers IMO cause of "networks"
I'm 100% supportive of tmux with remote connections.
If you're having that spotty connection then I think the better solution is using something like mosh or the best option is configuring better timeouts in ssh{,d}_config. Look at `TCPKeepAlive`, `ServerAliveInterval`, ServerAliveCountMax` (s/Server/Client/g for sshd_config)
So, even when nested, it's always clear what will happen. One nested session (two total) is manageable. Adding another gets annoying, but the need is rare.
Here's what I use:
set -g status-left "#{?client_prefix,#[bg=colour76]^,#[bg=colour235]_} [#S]"
set -g status-right ' | #[fg=colour76]#h#[fg=default] %H%M%Z'
But yeah, when I didn't have splits and things in my emulator I used something like you did. Make a remote section visually distinct. To a lot of extent I still do, just this isn't necessary through tmux anymore
bind -n S-up \
set -g status-bg white \; \
set -g prefix None
bind -n S-down \
set -g prefix C-a \; \
set -g status-bg green
Locally, I don't get the point. I can do tabs and splits. A session? That's another window. I can also hide that and use keyboard shortcuts to navigate towards it. Am I missing a local benefit?
The session is the thing I use, and I do use tmux every day, but I exclusively use it when in ssh as on a local machine I have everything I need. You mention the hybrid, but isn't the point that the connection is ephemeral? We don't exactly want to keep it alive unless we are actively using the connection, right? This is what I'm not understanding about your mixed scenario. What is the difference between what you do and me just opening another pane in ghostty and doing the ssh there. I can't detach from ghostty, but I can save the window state (or move to a different window in a different workspace or just cycle through windows like I would a session) so when I quit it it resumes, but honestly, I rarely quit it. Keeping a remote connection alive while there's no active usage seems like a waste and it's just going to disconnect when I walk away from my machine anyways. So what is the difference?
I do find sessions tend to get unruley easily, and more easily than windows just do to being hidden another level deeper. It's not too bad when single user, but I'd still suggest organizing tabs over multiple sessions, in most cases. That's because it is easy to just leave something like an editor open and accidentally overwrite files. Someone just logs back in after awhile and types :x into vim (or :x!) without thinking. It's also just too easy to let sessions stale. I haven't worked with a single person (myself included) who did not have a tmux session that was inactive for at least a month more than their last login time. (Better than all the hanging VSCode sessions though) That's like people using srun to run slurm jobs instead of submitting a batch job into the queue. That's totally fine to do when you're sitting in front of the sim but if you walk away and the code errors or just ends before your session ends then you're just hogging a machine that could be used (there wasn't even a benefit to this either!). Locally, it's pretty much the same thing except locally I can do all my tabs and splits and sessions without tmux.
Okay, but sessions are one of the best things about tmux.
Sessions aside, in my opinion tmux's flow is just better than terminal emulators I've tried. It slots into my brain in a way no terminal emulator's tabbing support ever has, and I have never found a terminal emulator who supported keybindings in such a sophisticated and seamless manner as tmux does.
I also concur with other commenters, who mention that having uniform multiplexing of shell windows between local and remote environments is super useful for muscle memory.
Yep, this is why I use the same .tmux.conf on my servers and my local machines. I don't have to care about my terminal emulator - I might be using a fully featured Windows Terminal or iTerm, or I might be using xterm or st on some spartan system. My navigation works the same everywhere.
> My navigation works the same everywhere.
> Okay, but sessions are one of the best things about tmux.
> Sessions aside
> uniform multiplexing of shell windows between local and remote environments is super useful for muscle memory.
Then, there's the fact that some terminals capture too many keybindings and get in the way of some terminal code editors I'd like to use (e.g. recently I was again trying to use flow <https://github.com/neurocyte/flow> and its next-tab and previous-tab shortcuts clash with Ghostty's). If I had a terminal that 1) was nothing but a black box with the capability to display Unicode font glyphs and ligatures correctly, 2) works under Wayland, and 3) captures as few keybindings as possible, I'd use it locally with tmux and live happy.
> I like that I can use tmux locally and remotely and thus have the same interface wherever I do my work
> (e.g. recently I was again trying to use flow <https://github.com/neurocyte/flow> and its next-tab and previous-tab shortcuts clash with Ghostty's)
Like I use vim so frequently hit <C-w> while typing and when I'm not on mac I kill browser tabs (probably for the best lol). But the benefit of all CLI programs is the config. There's no "one-size-fits-all" solution, so having configuration is the next best thing.
> If I had a terminal that
That's precisely my point! Many terminals these days try to do too much and by default end up capturing keybindings that I want to use for something else. The terminal should allow me to do that something else, not get in the way. A simple terminal wouldn't get in my way and I could insert tmux in the more complex scenarios (even for local sessions). I know I can reconfigure most terminals, but not everything is configurable, plus everyone knows the best tools are the ones that don't require to be reconfigured to be great. :)
Thankfully my main editor is Helix and by default it doesn't clash with anything because it's a modal editor where commands are regular keystrokes and not key combinations.
> Foot is a barebones solution, so that might be good for you
I'm aware of foot, and it works fine, but it doesn't correctly display the fancy icons used by foot (and other programs). For a visual comparison, here's foot on the left, Ghostty on the right: <https://v1.imgpaste.net/images/public/21baee04-b4a5-4693-8a8...>
> The terminal should allow me to do that something else, not get in the way.
I mean I really don't understand. Do you expect the terminal developer to know all the keybindings you currently use and will use in the future? That's a really big ask! Or are you asking that the terminal comes with no keybindings? I can understand that one, but I think it is not going to be popular. For most people that means more configuration. You can disable anything you don't like.
> Thankfully my main editor is Helix
> I'm aware of foot
i expect terminal developers to keep keybindings to a minimum.
come to think of it, i can't think if a single keybinding that i need from a gui terminal. the only keybindings i need from a terminal are those for tmux, so if a terminal replaces tmux (like wezterm is able to) then those are ok, but otherwise when i run tmux in a gnome terminal then there isn't a single key binding that i need for gnome, except possibly copy and paste.
> i expect terminal developers to keep keybindings to a minimum.
> there isn't a single key binding that i need for gnome, except possibly copy and paste.
I mean I'm not trying to harp on your preferences, but I think you have unreasonable asks. I haven't used an emulator where you cannot change or remove those keybindings, so I think that solves your asks. But I think your asks are unreasonable because they don't match the requests of the majority of users. You're totally right that "sane defaults" are not going to be "sane" for everyone. But why isn't that fine? I mean that's why we have configs, right? So that we can make them our own? "Sane defaults" are just an attempt to minimize configuration by trying to match average demand, not try to match every person. I mean we all have dotfiles, right?
I'm just really lost in what you're asking for. You're asking that *defaults* match *your* needs, right?
Because it sounds like you're upset that keybindings exist so... disable them? Is there something you can't disable? I'm really lost here...
agreed.
You don't use new tab? New window? You physically click the tab to move tabs?
in gnome terminal i don't use tabs at all. i use tmux for that.
gnome has a number of keybindings, most of which start with shift-control, which is a rare combination not used elsewhere. in tmux all bindings start with a leader which is configurable, making it easy to stay away from other key combinations. wezterm can do that too. i am not the person you are originally responding to. i am happy if all keybindings can be changed with a single setting. i don't know if gnome terminal can do that, but it appears that gnome does exactly what we ask for: choose keybindings that don't get in the way.
also, very annoyingly, when i select something in tmux, the selection doesn't stay highlighted. i find that a serious issue that makes tmux selection annoying. another problem: clicking on urls to open them, which is a terminal feature, stops working too.
These are icon glyphes. You have to set the font setting for the terminal to have them. By default, foot uses whatever you set for the monospace family (you can use the command `fc-match monospace` to find out which font). You may want to use a Nerd Font instead.
Also good command to know `fc-list`. It'll list the fonts on your system (you may want to pipe it into a pager: `fc-list | less`). There's a few fc- commands that deal with fonts for anyone facing these types of issues.
This may be of use to you or someone else reading the thread. You can set your bash prompt to show user@host when connected via SSH:
if [ -n "$SSH_CONNECTION" ]; then
export PS1="\u@h:\w$ "
fiTypical Macintosh user.
I have started using nvim about a year ago, and that one also took three separate tries before it finally stuck with me :) After that, about 8 months ago I found tmux and had a harsh "first impression".
These days I enjoy using both of those tools and get more excited the more I learn about them.
It's modular and customizable.
It has been my starting point, and I have gradually been adding new plugins and reconfiguring existing ones.
I can see its purpose for vim users... maybe tmux can still have some features for neovim users. For emacs.. my answer is simply all-in emacs. When I saw a screenshot of emacs in the article, I thought I was in for some education... but sadly not the case.
Maybe I am missing the point - but I am happy to learn.