The Joy of Linux Theming in the Age of Bootable Containers
blues.win
blues.win
Many years ago I added an install script to https://github.com/nickjj/dotfiles to get set up in basically 1 command because I wanted a quick way to bootstrap my own system. I used the official Debian and Ubuntu images to test things.
Over the last few days I refactored things further to support Arch Linux which has an official Docker image too.
This enables being able to do full end to end tests in about 5 minutes. The container spins up in 1 second, the rest is the script running its course. Since it's just a container you can also use volume mounts and leave the container running in case you want to incrementally test things without wiping the environment.
Additionally it lets folks test it out without modifying their system in 1 command. Docker has enabled so many good things over the last 10+ years.
Never stop tweaking. No computer can be called home until it runs your own set of aliases/commands.
I just pushed an update to remove VcXsrv at: https://github.com/nickjj/dotfiles/commit/fdc1ddd95c2defb791...
As for why I was using it:
I've been using WSL since nearly the beginning (2017 / 2018) and used VcXsrv back then to get bi-directional clipboard sharing before WSLg was available. For a brief time I even ran Sublime Text in WSL 1 way back in the day.
Then I used WSL 2.
Then I tried WSLg when it first came out and it was really bad. Clipboard sharing didn't work for me which was the only reason I wanted to use it. I set `guiApplications=false` and never looked back.
I just tried it again now by closing VcXsrv and removing any DISPLAY related settings I had in my zsh profile. Then I shutdown WSL and started up my instance.
Bingo, clipboard sharing "just works" and I also installed xcalc which ran flawlessly. This simplifies things so much.
This unfortunately makes pasting text from Windows into WSL a problem. The ^M isn't present when pasting with CTRL+SHIFT+v because I guess the Microsoft Terminal converts it but that's a big pain to hit compared to "p".
I opened a bug report, let's see: https://github.com/microsoft/wslg/issues/1326
I never used Incus before to know if it has other benefits but Docker solves my use case. Before Docker existed I used to do this with LXCs back when I ran my entire dev environment in a Linux VM.
There's something nice about running Debian and having confidence in all the packages because they're built and maintained by the Debian team. Of course there are exceptions, but in my experience they're rare. The only non-standard repo I regularly use is fish shell, and the updates are so few and far between (and very public) I think the risk is low.
I suppose this isn't strictly a container-specific problem; you could add the repos and install / update all those packages yourself too. But being able to package everything up into a single file that you can then boot into as your OS means you're also packing all the supply chain risk.
Curious if anyone else shares my concern or if I should just put my tinfoil hat back on...
1. https://github.com/ublue-os/bazzite/blob/main/Containerfile 2. https://github.com/ublue-os/bazzite/releases/tag/42.20250417
From your link, everything is pinned? So a theoretical exploit in a future release of package is not going to exist in this immutable release https://github.com/ublue-os/bazzite/releases/tag/42.20250417
Every system which tells me how immutable it is then shows me it's automatic version bump script or something.
Contributor here, we've been working on this diligently over the past cycle (the rest of the org is mostly done, Bazzite is largest so we're only getting to it now). We're hoping to be done over the summer with published SBOMs and all that good stuff.
I looked over their code, saw some things (I believed) I would do differently, and it was very easy to make my own personal spin to use.
After doing that, maintaining it, and using it daily for the last year I went back on some of my original choices. I feel much less critical of the decisions Jorge Castro made and it's probably time to compare and contribute if I can. Like, Homebrew on Linux ended up being way better than I expected. But some things I liked better my way. Say, including the signing keys for Chrome's 3rd-party repo statically instead of fetching them over the network. (Writing this from my phone I don't exactly remember how they do/did it.)
Overall, I'd recommend trying it yourself! It's been a ton of fun.
This is a fantastic idea, it sucks to have an upgrade blocked by a slow repo, if you wouldn't mind filing an issue or sending a PR I'd love to have this. Thanks for the feedback!
[0] https://github.com/grassmunk/Chicago95 [1] https://garudalinux.org/editions (screenshots don't do it justice)
Even on Mac OS there used to be solid theming support (thanks, Unsanity) because software all used the system UI widgets. Not nearly as common anymore.
Personally I believe CSS to be quite ill-suited for the purpose. It’s ok if you’re writing a theme for a bespoke one-off app but breaks down in the system theme use case. In particular, CSS inheritance makes for a lot of unnecessary trouble for both third-party themes and accessibility affordances.
Last I knew there was something of a disinclination away from paramaterization in the GTK dev sphere too, which is another significant problem for third party themes and accessibility. Hardcoded fonts, colors, etc makes for pointless brittle rigidity.
But its terrible for themers, it's like running a CSS override on every site that runs Bootstrap and expecting it to work properly. It won't.
I don't run any themes anymore so it doesn't bother me.
Once they finish sucking donations and other forms of financial support they'll probably announce it's time to "sunset" Gnome/gtk because it sadly didn't met unspecified expectations of unspecified group of people.
Gnome team, what they did and what they still want to do, their attitude towards users - especially those who dare to criticize them is THE result of polluting FOSS with corporate style of software development.
Theming and customization of Linux is half-dead because of what happens at Gnome.
To make my biases clear: I'm a software developer that uses Gnome daily, and is developing a GTK/Adwaita app. I used to rice a lot back in the i3 days, but I don't particularly care about that nowadays, and stick to the defaults when I can. For my purposes, GNOME and Adwaita is perfect since it's very opinionated by default, and you can make good looking apps with minimal effort. Since all Adwaita apps are supposed to look similar and follow the same HIG, most of my desktop apps have the same look - but more importantly, the developers of the apps can also be confident that their apps look correct on my desktop. This is something that developers in the GTK space generally want, and for good reason[0].
One argument is that you as a user of the desktop should be able to have the final say on how your apps look, which is a totally valid take! And there are DEs, WMs, and apps which give you this freedom like Hyprland. But this doesn't guarantee that those apps will look good, or look consistent with each other, or even act consistently across apps. On the other hand, I as an app developer want to guarantee that my app looks good on your desktop, and the easiest way to achieve that is to target a single desktop environment, rather than an infinite combination of possibly-similar-but-maybe-completely-different desktops. Every preference has a cost[1][2], and when you take this philosophy beyond just preferences and expand it to color schemes, padding, margin, iconography, typography, it becomes unmanageable.
This isn't to say that GNOME is perfect, and I disagree with the project on some fundamental technical things like not supporting xdg-layer-shell[3], and refusing to accommodate server-side decorations for apps which don't want to render decorations themselves. (On the cultural side I can't comment, since I have no experience with that.) But in my opinion, this is the project that can deliver a usable and consistent Linux desktop to the average person the most effectively.
[0]: https://stopthemingmy.app/
[1]: https://blogs.gnome.org/tbernard/2021/07/13/community-power-...
[2]: https://ometer.com/preferences.html
[3]: https://gitlab.gnome.org/GNOME/gnome-shell/-/issues/1141
Aside from that, consistency and themability are not at all mutually exclusive. Back in the early days of OS X, theming by hacking system resource files (or patching them in memory via haxies[0]) was quite popular and for the most part, worked very well — generally, the only apps that didn’t play nice with themes were those sitting in the uncanny valley between native and custom, using bits of both, which tended to not be the highest quality applications anyway. This was way before Apple started pushing devs to parameterize their apps, too, and so similar theming capabilities today would work even better since themes can just tweak the parameterized fonts, colors, etc as needed to maintain coherence and usablity.
The real problem with GNOME/GTK is simply that it wasn’t designed with user customization in mind even as a remote possibility. A UI framework that did keep these things in mind combined with a strong dev culture of parametrization would make for a desktop that’s both customizable and consistent.
When your app doesn't follow how my desktop looks it doesn't look good on my desktop. And unsurprisingly most modern Gtk3 and especially Gtk4 apps do not look good on my desktop.
What you actually mean here is that you want to guarantee that your app looks good on your desktop, not mine.
The counter argument to that is "so let the user theme the app, to suit their own desktop", which would be a decent solution, but:
1. My vision for my app might conflict with your vision for your desktop. Maybe I want this button to be a light blue because it meshes well with some other elements in the app, but you want it to be a darker blue because it fits with your desktop's color scheme. What happens then?
2. This still doesn't guarantee that the app will look good. If you theme my app's home page, but don't theme the rest of the pages, then sure it'll look good at the home page - but as soon as you start using it, the look will fall apart. Or, what if I push an update to my app which adds a new page with a new kind of UI element? Do you really want to be maintaining your desktop theme for every single app you have?
3. This adds a burden on me as the developer to make parts customizable. This is the least convincing argument in this list IMO, since if there was better tooling and infrastructure for theming in GTK this wouldn't be a problem - but there isn't, so it is still a problem.
As a practical example, my app makes use of a WebViewGTK to display some info. I inject some custom CSS into this web view to make it look like Adwaita. This touches on points 2 and 3:
2. The webview has some UI widgets which aren't present in the rest of GTK, like a sticky header bar. You would have to manually maintain a stylesheet for this single element.
3. I now need to write a way to let users theme the custom CSS inside the webview, rather than just the CSS of the GTK widgets themselves. (I have already written this, but it's still a maintenance burden.)
The user trying to make your app match their desktop should 'win'. Your responsibility is to ship out an app and make sure it works in the way you want it to work.
If the people need to do more work to make it look good on their desktop (as I likely would running awesoemwm), that shouldn't be prevented, but it also need not be encouraged. It should at the least though be facilitated, certainly to a better extent than it is.
Probably something similar to how Apple platforms handle colors. Instead of providing a single static light blue, you have a couple options:
1. Use a “system color”, which is pre-tuned for optimal contrast, appearance, and usability and adjusts automatically when e.g. the user switches between light/dark mode or enables an accessibility setting related to color or vision
2. Define a light blue that’s actually multiple variants of the color bundled together, with each being optimal to various environments, with the UI framework choosing the right one depending on the situation
Arguably developers should be doing these things anyway for accessibility reasons. It’s not been good practice to use e.g. bare color hexes for quite some time now.
I suspect the thing that rubs a lot of people the wrong way with Gnome isn't so much the lack of customization or theming, but actual decisions behind how the user interface should work. People simply complain about the inability to customize Gnome as a reflection of how powerless they are to do anything about those decisions.
But also that the Gnome philosophy "leaks" into the wider ecosystem, thanks to the dominance of Gtk.
(I use GtkWave as a waveform viewer. I have it installed systemwide, and it has proper menus. If I activate the oss-cad-suite environment [which is superb - this isn't a dig at that project] I have to remember to specify its full path when running it, otherwise I get a newer build shipped with oss-cad-suite, which has hamburger menus.)
It's near impossible to have a Linux GUI environment without GTK applications, while the opposite is not true for QT. I have a full desktop setup with GNOME and my machine doesn't even have the QT libraries installed.
https://www.gamingonlinux.com/index.php?module=statistics&vi...
It's totally expected for me, KDE is just better if you want any kind of customization.
It also comes preinstalled on Steam Deck for instance.
From some of the big GTK applications I use are Firefox and Gimp. That's about it, most of everything else is using Qt or is in the process of switching to Qt in my experience (like Audacity).
And from the above, Firefox isn't really using GTK due to some integration with Gnome, they just didn't want to write their own Wayland handling bits. It works on KDE all the same.
My guess: because it is difficult to develop software that can be themed and it is difficult to create themes that look good. Not only is it high effort, but it has relatively low returns. Themes mostly affect how things look and, ideally, have very little impact on functionality. I say ideally since, when there is an impact on functionality it is usually a negative one (e.g. buggy behaviour). Contrast that to a window manager or compositor: while it won't affect the functionality of individual applications (ideally), it does have a fairly significant impact upon how one interacts with the desktop as a whole.
Seriously, just make GUIs. That is the solution to ALL of Linux problems. MAKE THE GUIs!!! I can't select the background color of panes from a color picker and instead I have to manually edit text config files and create folders inside dotfolders. Ridiculous. It's 2025.
After a while it loses the appeal, we decide to just use whatever defaults get offered, finetune one or two options and that is it.
I can speak to this personally. I used to always tinker with various Linux desktops, themes, etc but nowadays I just use vanilla Ubuntu with zero theming modifications. There are two reasons for this:
1. Like others have said, theming is easy but consistency is hard. I've found that anything besides Gnome just turns into a shitshow where half your apps just don't theme properly.
2. It's a massive time sink. While I could create a very consistent theme, it would involve a massive time sink into dealing with all the edge cases. When I was in college and just used Linux "recreationally", I could justify spending a ton of time tinkering with my system and getting everything perfect. But these days I use Linux professionally so it's less about having a beautiful desktop and more about something that just works and gets out of my way so I can get my actual work done.
I should note that I still play around with other DE's and themes though I now do it all in VM's. I'm slowly building up my own theme stack on a Debian VM and once I get everything buckled up I might actually deploy and it use it on my primary machine.
Surprised it's still going https://www.enlightenment.org/
I used to love theming my desktop environment, but the joy faded when I realized the UI felt much more magical than anything I was using it for. Wonderful application of the tech, though.
I guess the equivalent in the NixOS world would be its impermanence module, which erases root on every reboot to keep things as stateless as possible.
To put things into perspective, GTK themes, unless they bring lots and lots of bitmap images (which doesn't happen nowadays), rarely exceed a megabyte in size.
I guess one could spend less time learning how to package these as RPM packages and set up COPR to do just that, making OCI + bootc entirely optional (and yet you could build an OCI + bootc installing this package if you so wished!).
Side note: Judging by what I see on Reddit, the ability to theme a desktop is one of the top reasons someone develops a personal interest in Linux to begin with, so no need to justify that in my book.
I looked into it, but it looks like that you need to manually build the image and fiddle around with qemu.
Containers are so easy so people just started using them for every use case, even when it doesn't necessarily make the most sense.
One of the rare examples where "Dark Reader" not only failed but actually made it more light; there must be some funky CSS shenanigans going on.
Perhaps there's a nicer way to say it? "If the author so pleases, he may consider improving accessibility by making it compatible with Dark Reader."
Then it is phrased as a request, rather than a complaint or critique.
Also discovered that for me it’s less the OS or paradigm or theme/look and more that the windows manager is tiling type.
I also don't think the distinction between distro and container is murky at all.