I first tried KDE Plasma 5 but tbh I thought it was just a worse experience than Win7, it was really close but all the tiny little annoyances got in the way and it felt like I couldn’t do everything I needed through GUI so I still had to use terminal but it was awkward having to switch between using the keyboard and mouse and I would navigate through the GUI to get to directories then open terminal…
After a month or two of that I finally tried a tiling WM (i3wm) and it’s just a way way better user experience than any DE.
I will note though that I’m using Fish for my interactive shell and seeing anything in the tiny dmenu was just way too hard until I used Rofi for drun.
Without Fish and Rofi I might’ve tried more DEs or even gone back to Win7.
I recently used Linux Mint with Cinnamon on a relative’s PC and using Bash and the apt package manager sucks so bad. I even prefer Arch KDE, although I think Nemo is a bit better than Dolphin.
Anyway it’s been about 2 years of daily driving Arch with i3wm for me and I haven’t really gone out of my way to learn things but you naturally pick stuff up along the way just by using it.
Just make sure you’ve got another device with an internet connection in case something happens. I basically haven’t had any issues after I got better but I made a lot of user errors at the start. Nothing that can’t be fixed but finding out how to do the fixing without internet is a million times harder.
Gnome’s window sizing has always looked comical on my display. So I fix it with Orchis gtk compact theme. Also GSconnect is an irreplaceable utility for me.
I don’t know how helpful it is to split stuff out like that. Especially grouping so many things under “default applications and daemons”, which is most of what a desktop distro is. Also depends largely on a PC vs server setup.
should list an init system as its own bullet, which others have mentioned.
“one or more shells” doesn’t mean that much. Yes, every distro includes one but the only difference between a terminal and any other application is that a terminal needs to be able to escalate to root privileges. You can think of it as just another default (but special) application. A lot of stuff that people think about when they think of Linux distros is just various clever mechanisms for supporting the terminal shell. Like the PATH environment variable. If you are using actual desktop applications other than a terminal, there isn’t any interaction with the terminal shell application.
There’s also fwupd, for updating firmware (your hardware is gonna be running out of date/buggy/insecure code if you don’t have this).
The dbus daemon falls under the “daemons” bullet but it’s pretty important, like wayland/x11 it’s another IPC mechanism you need for programs to work correctly.
There’s also the sound system. PipeWire is the modern one that implements the interfaces of various other sound systems so existing applications work with it. pipewire.org (PipeWire also has its own IPC protocol like dbus/wayland/x11).
flatpak, snap, distrobox, toolbox, docker, podman, etc. for running sandboxed PC/server applications. I assume there are some programs that are flatpak-only these days.
gsettings/dconf for Windows-registry like config that many programs use.
There’s also plugging in an implementation of the glibc Name Service Switch, which allows libc to use a mechanism other than /etc/passwd and related files for user accounts, internet service names, DNS resolution, etc. . systemd can provide NSS implementations using its own user account mechanism.
There are already almost a hundred extension protocols and you need dozens of them to implement just barebones desktop functionality. If you look under the surface the Wayland ecosystem is arguably already more complex than X11 ever was and it’s only going to get worse.
On my computer I also use X with my perfect Openbox setup.
I tried Hyprland and labwc but both failed miserably. None of my main tasks worked properly. I could not screen record (yes, I’ve read all the tutorials and tried to do it but OBS simply did not find any screens or windows to record) and Kdenlive just flickers with white and black bars and areas (it actually crashed so hard that it destroyed my config and I need to re-setup it under X11). YouTube had noticeably worse performance and video quality. Also: Gaming was basically impossible (not starting at all or extremely worse performance - and yes, I’ve read the tutorials on this, too).
All of this may be due to the fact that I have an Nvidia card in my computer. But I’m not going to replace perfectly fine working hardware just because my display manager does not support it properly (which is entirely Nvidias fault, but still) so I stick with X11 until my graphics card needs to be replaced.
On my laptop I run Hyprland without any issues except one, but on my laptop I don’t do screen recording or watch 4K videos or do video editing or do gaming. For this use case it’s very well suited.
Wayland may be ready for certain environments/tasks, but in my opinion X11 is just much more compatible with anything.
The list is generally correct but these days, systemd has made quite an impact also. If a distribution uses systemd, it has one software to handle everything from booting (instead of grub), handling start and status of all system services etc. Its probably the largest change to the Linux ecosystem in a long time.
X11 and Wayland are desktop protocols, so things like desktop environments and window managers depends on one or them to be installed. Without them, you don’t get any graphics except for the console. It’s all built on top of one of those.
Installing Arch for the first time taught me a lot about how my system works, since you have to choose all the parts that make up your system yourself.
I remember when it was considered a mark of professionalism for a web developer to have an email on their own domain. At some point that changed. I think after GMail came out it was so good that everybody switched to that.
The domain isn’t all that important, the IP address of the mail server is. I pay an external service that provides a mail server, and my DNS records point to that.
But hosting my own mail server, while possible is not recommended.
Counter point, Ubuntu specifically has so many old posts and answers that aren’t necessary in modern systems, deprecated, or straight up no longer correct. Also a lot of recommendations that can screw up a system in strange ways. I feel like many issues (ie. Bluetooth, USB, Wifi) are due to people stumbling on old posts with configs and tools that have changed and blindly applying them
This kind of question comes up in many areas. And which software you use is less critical compared to politics. Of course you can use google and advocate foss, if your question is to be taken literally. It would not be the best thing you could do, but what would even be the best thing? Using software is not helping anyone (exept for software that takes your data or mines crypto while you use it or something). You would need to donate, contribute or bring people to do these things to really help the software/devs. Use which software/service you are comfortable with using.
thank you, you’re right, I love open source, I will contribute to it. And by the way, this is an extreme opinion, but Discord is an open source hybrid!! It’s mostly open source but it’s got proprietary blobs.
There are other issues with Discord relating to privacy, which would even with a libre client (I think there is one? Bettercord or fosscord or something?) be a good reason to avoid it.
But I understand that there are important communities on there.
linux
Active
This magazine is from a federated server and may be incomplete. Browse more on the original instance.