Linux users with uncommon or unusual setups: tell us about it
I'll start with mine. yes part of this was to brag about my somewhat but not too unusual setup. But I also wanna learn from your setups!
Anyways: I primarily use Gentoo Linux.
I have two headless servers: a Raspberry Pi 4B and a Oracle cloud VM (free tier). Both running OpenRC, and both were running mainline kernel with custom config (I recently switched the Pi to PiFoundation kernel due to some issues). The raspberry pi boots from SSD and has no sd card inserted.
Both servers were running musl libc instead of glibc for a while. This gave me a couple of random issues, but eventually I got tired and switched back to glibc.
I have a desktop running gentoo and a laptop running arch, but hoping to switch the laptop to gentoo soon.
Both are daily driving wayland (the desktop had nvidia card and used for gaming). The desktop is running a kernel with a minimal config that compiles in 2-3 minutes.
My work machine isn't too unusual, apart that it has 52 USB devices connected. And here's something you may not know: Linux can't enumerate more than 16 USB ports if the root is configured as USB3, so I had to force all the ports to run in USB2 mode - which is fine in this case, since most of them are serial ports.
I'm not sure it's a kernel limitation or a hardware limitation. But it does throw an error in syslog when you connect the 17th device. Not as USB2 though.
This is caused by your root controller's limited bandwidth and it's inability to handle that many 3.0 devices at the same time. Some of the newer motherboards with USB C PD have controllers in them that can do a lot more.
It's basically a hack on part of the company that made the root controller IC. They know they only have enough internal bandwidth to support 16 USB 3.0 devices so they intentionally bork things when you plug in more than that since their Transaction Translator (TT) can't handle more and they were too lazy to bother implementing the ability to share 2.0 and 3.0 properly.
I'm guessing the decision went something like this...
"We have enough bandwidth for 16 3.0 devices... What do we do if someone plugs in more than that?"
"Only a few people will ever have that many! We don't have the budget to handle every tiny little use case! Just ship it."
So it's not Linux fault in this case. Or at least, if it is (a problem with the driver) it's because of some proprietary bullshit that the driver requires to function properly 🤷
Yeah I figured it might be something like that. But I wasn't sure it wasn't a kernel limit - or even a limit in the USB3 specification - because I actually only have one USB3-capable device connected (my cellphone). All the other devices are low-bandwidth USB2 FTDI USB serial converters. I thought it couldn't be a bandwidth issue when all but one device can only use a fraction of what's available.
Knowing some fringe users, your setup is probably ~3 points or so ahead of the middle of the bell curve. You never know. There's probably a guy running kernel 4.12 on a 1990s CPU with his showa era CRT monitor to play freedoom.
I have 100% done this. I have a picture somewhere of my old ass p3 laptop with an rgb keyboard. So not showa era crt, it's heisei era LCD (more appropriate considering it's a Sony vaio f series)
This is apparently very unusual since a lot of games completely break with it but two monitors with the main monitor on the right
This is unusual? I use the same monitor configuration, and I didn't notice any problems with it. Or at least I didn't figure out they could have been caused by monitor setup. Could you give me an example of what problems have you encountered?
Either games spawning on the wrong monitor and not reacting well to you moving the fullscreen window to the other monitor, or mouse input issues. Latest I've had was L.A. Noire, which locks the mouse to a portion of the screen and doesn't allow you to freely turn the camera. (I just tested it again and now it seems to work fine though! I hope that persists.) Quake II doesn't allow you to move the mouse at all, or rather only in what seems in like a 2 pixel wide boundary in the middle of the screen. No such issues if the other monitor is turned off or configured to be on the right side. I've encountered more games that had issues with this in the past but these two are the recent ones I've had trouble with since setting it up like this again.
I've been kinda dipping my toes on NixOS but the flakes are really throwing a wrench my way.... Yet they are apparently NixOS' future so I'm just kinda stuck
Flakes on the system level aren't too bad. You can pretty much just keep your configuration.nix, but now you call that from a flake.nix. The difference is you remove all your nix-channels and you specify your nixpkgs in your flake.nix. So its really using a flake instead of nix-channels.
The cool part is when you nixos-rebuild the first time, it will save your nixpkgs version in a flake.lock. Then it will stay that way until you choose to upgrade with nix flake update. Nice and stable.
I felt the same, but I'm reading through this book and so far it's been helpful for understanding and setting up system flakes: https://nixos-and-flakes.thiscute.world/
NixOS can boot from a file system that only has /nix, since essentially the kernel command line has init=/nix/store/.../init. Everything else will be created during boot by that if it isn't already there. So technically you could only mount /nix and you would get a blank system every time you boot (but that wouldn't be very useful in most cases). Mounting these is done in the initrd.
A lot of people have a setup where only select files are mounted from a persistent partition, such as /var/lib/postgresql, basically anything they want to keep across reboots, so that the rest is discarded when they reboot. This prevents the system from accumulating junk over time, from services you once used to have but no longer have running, and so on. Personally I found it too much of a hassle to keep track of what files I want to keep, so I save the entire /etc and /var. I still keep the tmpfs though because it's pretty cool.
Not mine, but while I was an intern for a lab I enjoyed using a very normal-looking desktop with a casual 4TB of DDR4 and no SSD or HD, dual Xeon configuration. Rather, it did network boot and pivot root into an in-memory filesystem. It had a UPS and typically ran for months entirely from volatile storage and was used to run experimental photo and video processing. This was about ten years ago.
It's amazing how well Linux performs on older hardware. Wayland seems to reduce the resource utilization a fair bit as well. The screens on the 980Ti are quite a bit slower than the RX480 so I arrange my workload accordingly and throw some windows over to an activity to increase my available higher speed screens. But the CPU rarely pegs out, it's not like I'm doing ML shit, just building software for telemetry and automation, or working in spreadsheets.
I have Void Linux running on a GPD Win 4 (6800u). It performs well enough to emulate Demon's Souls through rpcs3 at 720p 40-60 fps.
It has a button on the side which toggles the built in controller between a "kb+m" mode and a normal controller mode, so I wrote a udev rule which opens Steam in big picture mode if its not running already when I switch to the controller mode.
I also sandbox a bunch of applications installed from the repos (including Steam and Firefox) using bubblewrap instead of using something like Flatpak.
I have a custom (half-working) version of slurp which allows starting selection immediately, which in turn allows me to immediately get the position of the cursor, which I use to launch tofi under the cursor (I don't know of any other way to do this on river or even Wayland in general).
I use secureboot with custom keys (using sbctl), and I build a unified kernel image from which I boot with dracut, into a fairly standard LVM-on-LUKS setup, all flicker-free (by manually turning off Plymouth at the right time).
UKIs allow me to boot from an efi shell very easily if thing go very wrong.
I run dnsmasq for caching, together with stubby for DoT. I highly recommend at least dnsmasq if you use Steam (fixes weird issues with their downloads).
I toggle running Qt apps' dark/light mode by modifying the qt5/6ct config file with a perl script which darkman runs. I switch the wallpaper in a similar way.
I don't use a status bar, I put most of what should go there into the Emacs tab bar (with custom dynamic icons and everything). It has volume, battery, temperature, wifi, system load, incoming mail, playing music and time display. Everything but temperature display works on both Linux and OpenBSD (and some on Android too).
Honestly there's a bunch more weird stuff but this is getting pretty long.
These are from my main PC, which use the same setup (no battery/wifi, but you can see that the volume icon only has a single "bar"). The name of the song is truncated to the maximum possible width and "cycles" through the name. Second picture also shows tofi which was opened under my cursor in an arbitrary position.
BTW the header bar only appears for the floating Emacs window (so I can drag it with a touchscreen) using a GTK css hack.
Edit: Wow, I took the pictures 9 minutes apart and both the CPU and GPU temperatures are exactly the same (they both fluctuate slightly while idling).
I'm not too sure how unusual it is, but I have a satellite tracker on a pi 3 b+ based on satnogs. It helps other scientists get data out of cutsats and other satellites. It's pretty easy to set up once you know what to set up.
I once had a butler program on a pi 1 with WiFi chip back around 10+ years ago. No ai, just a bunch of batch scripts + espeak. It was a cool project that would tell us the weather, time, any to-do items, and internet usage ( att had a hard limit of 100gb and I used a script tu tell how much we used per month). Ran for a couple of years and then disassembled it. Still have the GitHub repo. This was many years before Alexa, Google, and the other such projects. It wasn't better at all (espeak sounds so robotic, even when tweaked).
I ran a Bitcoin miner on a pi and made -$4.50ish a month back a decade ago. It was my most popular wiki pages back when I self hosted one. People were really interested, but it never made any money. It was more of a proof of concept . It's pretty easy to compile, but hard to track down all the dependencies. That was waaaay before the asci miners came into play.
I use Wayfire (which not many people use for unknown reasons), and one of the things I like to do with it is have a fiery drop-down Kitty terminal. :)
I haven't seen anyone else do a drop-down Kitty in Wayfire before, so I'd like to boldly claim I'm the first one to do so. :) Yes I know it's pointless, but it's also cool, and it's fast thanks to being fully GPU-accelerated, so why not?
And no, I don't use the fire effect for other windows - that'd get real old, real fast. Thanks to Wayfire, I can define window rules so the effect only applies to my drop-down kitty. Also, my regular kitty windows open normally, without any fancy effects - and it's possible to differentiate this thanks to kitty allowing you to specify an custom appid.
I also use doas instead of sudo. I just got tired always fighting with sudoers, doas is so much more easier to setup and work with.
Finally, I use grc to colorize all my log output. Makes my journactl looks nice. :)
Wayfire is not tiling right? I imagine its a similar reason people didn't use open box much. It's a non-tiling window system, and people who go that route tend to to full DE. But I am with you. I wonder why not many more people use them.
Doas is cool. I actually switched to it shortly before ditching both it and sudo, and deciding to rely on a users/groups system.
Yes Wayfire is a floating WM, but it does have a tiling module which supports simple tiling (up/down/left/right/top/bottom), and you can even define window rules for automatic tile and workspace assignment.
Of course, this is probably not enough for hardcore tilers, but personally don't think I'd ever need more than 4 tiles per screen - and if one or more of those tiles is a terminal, I can easily multiplex it using Kitty's built-in splitter, or Zellij.
It can play all my games at 1440p and ultra settings (RDR2, GTA V, etc.) I've never had a time where I've wanted to upgrade from it. I built most of this computer for about $450-$500, all used parts I got off eBay plus some other parts that I pulled from my other computer
The Optiplex gaming setup is quite bizzare. Isn’t that CPU a bit of a bottleneck for this relatively powerful GPU?
Btw I used to own the same Thinkpad but it was supplied with a dedicated NVIDIA GeForce 6xx-something, but never tried Libreboot on it. Given that I sold it in 2020, not sure if libreboot was even doable on it back then.
A bit, it's actually not too bad. Rarely any micro stuttering on ultra settings in RDR2, I am actually planning on buying the AMD 7900XTX graphics card to put in this machine. I want to run local LLMs on it, I'm not too much of a gamer as I used to be. Anyways, this thing rocks! I love it. Eventually, I'll plan on buying a MSI Z690-A DDR5 motherboard and install Dasharo firmware onto it.
On my desktop I use 2 virtual audio devices that are linked to my real audio card with qpwgraph in order to split audio between VoIP applications and desktop/game audio.
I tried to set this up on a mac using soundflower so I could share my screen with an edit project with the director during lockdown and still chat to them at the same time. Didn't work for some frustrating reason relating to Skype.
I believe while I was figuring this out I discovered one of rogue amoebas apps that I could use in conjunction with sound flower and I was nearly certain I had it, it was something to do with how Skype worked that sabotaged me, I couldn't believe how stubbornly persistent Skype was despite how hard I tried to workaround it. I believe I was trying to make a single virtual sound device that combined my mic output with the system Audio so I could choose that as my microphone in Skype but SOMEHOW it was always able to fuck me over don't remember how, only that I was extremely angry.
Funny, I was going to suggest Audio Hijack, but a search for soundflower revealed that Rogue Amoeba were the stewards of that app for a few years. Anyway, Audio Hijack might be what you need if you're still in the market for such a tool. Rogue Amoeba are the experts in Mac audio-routing.
I have a laptop with an easily accessible m.2 slot, which I use with an m.2 to pcie x16 adapter to connect an external desktop grapics card to game and run ai.
apart from that, a diy nas running opensuse and a couple vms for dns, remote nas access, etc
These adapters only exist for pcie 3.0 and m.2 slots for ssds have 4 lanes at best, so you should expect performance to be between thunderbolt 4 and a full x16 slot in a desktop, but it's been working very reliably for me and is absolutely faster than the iGPU in my laptop.
You do typically need a desktop power supply connected to the adapter though, but since it only needs to power the card, it's fine to go with a lower power one.
Lenovo support seems to think I have an unusual setup since I run Linux on their Thinkpad & while the NVMe even after an RMA fails under heavy IO despite their partner WD, who sent me an email response saying they never test or certify drives for Linux or BSD. Many users have been experiencing similar failures with their controllers WD proudly boasts as in-house. Note that Lenovo also has a support PDF about running the device on Linux, but the support is ran by a bunch of clowns. Also not that when you purchase, the hardware brand is never mentioned so there is na room for due diligence.
Tl;dr: if you want a working Linux system, don’t purchase Western Digital or Sandisk drives.
I have been running my linux installs off of wd drives for years without any issues. Most of the devices I run are Asus laptops, maybe it is a Lenovo issue?
It seems to be an issue in the last year-ish but across a range of WD/Sandisk drives. They build controllers in house &, if the support is to be believed, don’t test or support Linux/BSD.
I play games on PopOS, and host FoundryVTT on my micro PC running Windows for DnD. I also stream games from the PopOS gaming rig to the Windows PC so I can play them from the couch on the weekend.
Waiiiit… what is FoundryVTT and why windows? Aside from adobe I dont know a lot of things that a linux pc cant do, especially with pop os (damn i love it).
Not my main rig, but my most unusual is 32-bit Yocto Linux on an Intel Edison that I got for free from a college professor that worked for Intel.
Yocto is awful. I mean it has a niche I guess, but there is basically no package manager. Somehow I managed to install a Rust toolchain on it, but it couldn't build the web server I wanted to run on it.
Oh, I remember having to use Yocto when I started experimenting with the BeagleBone Black SBC back in 2015. Yes I remember it being very hard to use. I remember I had need to rebuild the kernel to include a disabled kernel module. The cross compilation on my desktop PC didn't work, so I had to build it on the BeagleBone. That was an awful process, it took about 6 hours.
For anyone not familiar, the BeagleBone Black was an SBC that came out as competitor to the Raspberry Pi 2. The main difference was the BeagleBone used an open source design, based on a non-NDA CPU unlike the RPI, so it meant they published full kernel sources. But in my experiments I found the BeagleBone CPU was much slower than the RPI, and it's graphics hardware was almost non-existent compared to RPIs integrated graphics.
Yah, BBB was horrendously slow. Lots of neat features like realtime PDU and tons of IO, but the only way to use it was Debian headless because a full DE was painful. I bought a 7" touch display to use as an HMI that the BBB mounted on the back of and it duplicated the pins so you could put a cape on it (that didn't conflict with the HDMI pins it used), but I never used it because it was so slow.
BBB is actually really good nowadays compared to where it started. I've got quite a few deployed hardware appliance designs with them baked in. The real time IO and subprocessor was a nice quick and dirty way to get a little psuedo FPGA
Sometimes I'll start up ConnectBot, which is an android ssh client, on my meta quest. Then I connect to my laptop and attach to a running tmux session so I can use the laptop keyboard but see the text in a virtual window.
I used to have Gentoo running a Libvirt hypervisor, which would then run multiple short lived isolated windows and Linux machines with GPU passthrough for all the different companies and projects I was working on.
Spent far too much time keeping the guest machine images up to date, and all the configs and stuff managed and synchronised.
Then my laptop died that I was using to manage everything so I gave up.
The host was stable. And I was compiling the kernel for hardware and vfio reasons anyway, so why not compile everything and it’s not like there was a lot to compile.
i have a gentoo system with a custom s6-rc service tree that fully replaces openrc and boots via s6-linux-init.
instead of a display manager i have tinydm (from postmarketos) and autologin setup. since i use full-disk encryption and suspend-to-disk i find that i don't need the extra login step into my user session.
i have a bunch of bemenu-based helpers for wifi, bluetooth, vpn, audio, passwords, mounting drives, etc.
i don't have polkit or sudo installed. i use doas.
Do you have those bemenu helpers in a repo somewhere? I've been really getting into bemenu scripting recently and I'm really enjoying the results (case in point: when I do a logout/shutdown via my bemenu logout script I feel like an absolute boss).
I've got a thunderbolt chip on an AMD motherboard, which doesn't usually happen, and I'm running an LG 5k monitor through it. I use an IBM model M over native PS/2. I've got a Ryzen 7, but a GTX 1060 cuz it still works. It's running Ultramarine Linux, based on Fedora.
Is it that Asus ProArt Creator motherboard? To my knowledge that's the only AMD board that shipped with the special Intel chip required to use Thunderbolt.
I've been thinking of picking one up, but I can't justify the crazy price for it.
Not THAT unusual, but... I have a Dell R520 server that was leftover/retired from work. I mostly use it for storage due to the amount of disk trays it has. I have all of these disks in a ZFS pool, leaving no actual drives for the OS. However, this was an old VM server, so it has an internal USB 2 port and a ridiculous amount of RAM, so the OS is booted from USB, and I don't use swap.
Boot performance is abysmal (on the rare occasion where I actually need to reboot), but once booted I notice no real downside to having the OS itself on really slow storage. Sure, it's somewhat slow to do os-related stuff such as apt-get, but it's not like I'm in a hurry when doing it. Plus other than updating stuff, the OS storage doesn't see a whole lot of changes/writes.
Now I just need to figure out how to economically attach these 40 additional SAS drives I have. It doesn't have to look good (i.e. fit in the same chassis. Or any chassis at all, for that matter), it just have to work. These additional drives are only 4TB each, but they were free.
Just wondering what’s the power consumption and how long have you had it? I just got my electricity bill after running an R720 for a year and… let’s just say it wasn’t worth the low price after all
Now, not so unusual, I have pretty dull and standard "gaming" type PC running stock Debian, but about 20 years ago as a broke mofo I was running a phpBB forum off a wheezing Pentium MMX laptop with no screen (got ripped off a year prior) on Mandrake Linux. The whole thing was just loosely sitting under my bed. Managed to get a userbase of just under a hundred people before I lost interest. I was using Webmin to manage it from another PC.
I had to connect up an external monitor every time I needed to do something I couldn't do remotely. I learned so much from that laptop. "./configure, make, make install" became muscle memory.
Both servers were running musl libc instead of glibc for a while. This gave me a couple of random issues, but eventually I got tired and switched back to glibc.
My casual-browsing-only netbook is currently running on a RAID0 setup between the internal eMMC and the microSD card because I think it's funnier that way. Nothing useful's stored on there and it's one nixos-rebuild away from being reinstalled so I don't mind the inevitable breakage.
Just started running Arch + KDE on a Kingston Traveller to experiment with setup. Installed from live usb iso and then ran archinstall to the same device.
Runs nicely on my dell xps laptop and my desktop with 3 monitors connected to an Nvidia 1070Ti.
Debian testing on a MacBook Air 6,2 (2013). I guess that's kinda weird. Works fine as a netbook: Firefox, Thunderbird, TigerVNC (handles the low resolution well) and SSH. That's all I ask of the thing and it works fine. The only hardware that doesn't work is the webcam, everything else is 100%
It was a free hand-me-down and I put a $45 battery in it so I can use it on the couch. I think what will kill it is when the proprietary charger dies, they cost more used on ebay than the battery did.
Nice I have a MacBook pro 2016(?) that runs a flavor of Ubuntu over at a local makerspace. It was hard as heck to find and customize the driver's to get it working, but it does!
I've only met one other person that knew who/what Dvorak was/is, and also reportedly used that keyboard layout.
I struggled with getting lost on the keyboard (several family members have dyslexia and ADHD--I'm not sure if that is related or not), and as an experiment spent 4 months exclusively using that layout to force myself to learn.
They never told me how my brain was also only big enough for a single keyboard layout. Usually in windows, games map to the same keys automagically. On Linux, not so much. I'm constantly remapping controls because I can't be bothered to just have two keyboard layouts I swap between for games /facepalm
I’ve only met one other person that knew who/what Dvorak was/is, and also reportedly used that keyboard layout.
I experimented with it in University--I actually got a screwdriver and pried up and rearranged all of the keys on my keyboard within a week or so of starting--but after graduating I noticed that I was still slower at typing on Dvorak than I was on QWERTY so I gave up and changed back.
Thank you! It's so much more comfortable to typ on. Not faster, but Comfortable.
I hate the awkward and annoying questions from colleges tho: wHY iS yOuR nOt woRkinG NoRmAllY?
And the mess that ctrl-c ctrl-v becomes is also super annoying. Mostly on windows its annoying. Linux is a bit more consistent.
Qubes OS counts as an unusual setup, right? Maybe even more unusual, I used to use Proxmox on my desktop PC, and I ran Debian and Arch on top of that. Also a little unusual, I use a MacBook Pro with Asahi Linux (actually the Fedora Asahi Remix).
I use a chromebook which has two Linux containers running on it. One of them I'm experimenting with learning Docker and possibly selfhosting some things there. Only running one thing right now, and it seems to be going fine.
The other container is my main Linux "install", which has all my apps like Inkscape, VSCode, Kdenlive, etc. The container uses a mix of nix, flatpak, and apt for installing things, which I do want to try and consolidtae eventually.
Probably not the weirdest of them all, but I do think it's pretty cool to run all this on a chromebook.
Yeah, the official way from what I understand is basically a container running inside a VM (with multiple containers behind a flag). They run Debian, but I believe there are tutorials for other OS's if you want to try.
I've also heard of people who just replace chromeos with a Linux install, but I think that puts you at more risk of bricking your device.
Well my unusual setup I spent years thinking about it before I was even able to have the money to achieve it. It's based on portability and versatility and since I'm now working remotely now it makes even more sense. The plan was to run something portable with less power and smaller when outside, and leave the powerhouse to be accessed remotely. So for that reason I have a dualboot Oneplus 6 with LineageOS and Droidian, Waydroid container on Droidian and Debian proot-distro on LineageOS. That so i dont have to totally reboot for some tasks i might need on android or linux. 4 media folders shared between both of them as well as their containers. This makes sense now cause i long thought of running a Lapdock with it even if only wireless, and I got it recently! It works really nice on android but cant transmit over miracast on linux yet, still figuring that out. Nevertheless thats not the main device that is on my mind. A pinephone pro is a good fit too, but im leaning towards something like the gpd pocket 3, a real portable and modular mini pc that could be connected with just a cable to work better on the lapdock (also can be used as a tablet which is dope).
The powerhouse itself is a server with 16 threads of cpu and 64gb of ram and 2 gtx 1060s for graphics that i plan on configuring with vgpu to split graphical load between the vms with. It is also my remote gaming server :D with moonlight and sunshine, and i spent quite some time configuring all of it to be easily almost plug and play with controllers to have no issues if i disconnect or using multiple different controllers, with a good game launcher (Playnite) to host all games from it.
All of this just to someday achieve my dream of working wherever I want with a camper van to explore the world!
I've developed an install alias that automatically configure a wide variety of things really easily for arch, I had a bunch of people use my setup and logged the usage of each different keybind, then sorted them by most used and put those on the strongest fingers
I've spent more than a few hundred hours configuring stuff, you can check it out here if you want:
I guess my macbookpro from 2009 with Legacy NVIDIA grafics running Arch with GNOME on Wayland is pretty uncommon, lol
(Of course using nouveau derivers)
Alpine Linux on my desktop and laptop, Alpine on a Raspberry Pi 3 working as a network/Bluetooth speaker for 5.1 surround speakers, postmarketOS on 2 RockPro64's which I'm currently replacing for a single x86 NAS running Alpine.
Gentoo + OpenRC + TDE (therefore X) on both a first-gen Threadripper desktop with 96GB RAM and a laptop from 2008 with an Athlon64x2 processor and 2GB RAM. Updating gcc on the laptop can take a while, but it still serves well enough. Plus a couple of headless Pis that are also running Gentoo. Not overly unusual, but I may well have the only Threadripper of that gen running that specific distro and DE combination anywhere in the world, since each individual item is kind of low probability.
Not sure if this counts, but on my install of nobara whenever I hover over an icon in the home bar it relocates to the bottom left corner of the screen, leaving an empty space. I can still interact with apps by clicking on the empty space where the icon used to be so it doesn't brick my home bar, it's just really annoying.
I could mention that my bare metal server runs a rather unusual setup in that I use Arch Linux on ZFS headless as a kvm hypervisor and lxc containerisation host. I maybe want to migrate it to something else like NixOS at some point since I use nix on Arch on my desktop already but since I know Arch the most of any Linux distro I just went with it and it's running rock solid for quite a few years already.
I had the same issue of liking the UX of Linux admin too much to really use nix fully for system administration. I will say now being forced to work with an MacBook I decided now was when to start that journey. I like it, its been very clean over all but I feel like I am very much "developing" my system instead of administratering it, if that makes any sense.
Gentoo gaming and music production rig working through mostly tty with dwm as a graphical display. I typically stay on tty until I want to play a game, use modern web, or record a song. Otherwise tty with Links browser.
Irubn Bluefin which is a downstream of Fedora built to be more Dev focused and "cloud native". Desktops apps are flatpak first, my terminal just opens distrobox containers, system CLI tools I get from nixpkgs, services i run on a k3s service on it, and I have use vscodiunm with gitpods to support devcontainers hosted on the k3s cluster. I sometimes pxeboot a raspberry pi or another laptop or a server from my openwrt router to add compute to the cluster if I need it.
Been tinkering with, well, Tinkerbell to do the pxe booting from the k3s cluster but I may go back to Metal3 so I can just used the servers BMCs and do the extra work to config the pxe boot from there.
I really want to get it too full distributed desktop OS at some point, either using moonlight or some real systems work with RDMAoCE and tricking the processes into thinking they are on the same system. That one feel very RnD though.
Fedora Hyprland, with Floorp and Emacs. Not very unusual, especially when compared to what people are saying here. Umm... floating Waybar and EWW as a conky replacement? A customised Neofetch?
Oh, I got it! I'm using my own handcrafted colourscheme! It's not perfect but it looks very good and is quite nice and blue! And I use Bemenu for a logout menu in a homebrew script.
I'm getting into using Syncthing to synchronise my Notes directory between my devices, which I use on my phone to access my orgmode notes and todo items via an app called Orgzly.
I use a Launcher called Olauncher on my phone which runs LineageOS rooted with KernelSU (that's quite unusual I guess).
My backup solution is 2 USB sticks and Syncthing.
I run the teams-for-linux flatpak for education-related purposes.
Even with all that I still feel like the most Plain Jane user when I'm seeing people using servers and niche distros, even though I'm sure combining it all together will leave us with only 1 user in the world that does things in that exact way: me.
Not mine but my partner’s machine (which I build and largely maintain for her) is a custom Debian install on ZFS root using ZFS boot menu and running a custom minimal i3 desktop environment.
Nixos with xmonad and with xfce in no-desktop mode. Xfce gives me monitor positioning since I have two monitors and one is vertical. On a desktop, and on two laptops. Oh and I swapped my esc and capslock keys. Crazy I know.
Also I have nixos on my pinephone, ha. But I don't use it.
I was unaware of no desktop mode, I'm using dwm but I find occasionally if a monitor gets jostled it will lose the input. Going back to login screen on my work PC fixes the problem but on dwm only it's a reboot. I suppose this allows fixing that problem...
My old procedure with the monitors, pre xfce, was an xrandr script. But I didn't bother with positioning so the mouse goes straight across going from one monitor to another... you can probably do it but that was enough xrandr for me.
xfce-no-desktop also gives me media keys, although the audio keys are broken in pipewire now. At least screen dimming keys work.
I have a stateless and immutable system based on arch images that get built every night. I also have two different images for the same machine: one for general use and one optimised for music production. I just rebooted into the other image and I'm set.
Just a reboot. There's an EFI menu listing all the available images - I keep a couple of old ones around to roll back to in case of problems - and I just select the one I want.
I have a ZFS Ubuntu deaktop install que on an all amd PC and use snap. Given that everyone else has something completely extravagant and hate snaps, I'd say my standard build is unusual.
It was not too hard if you are already familiar with Nix. The features supported (and the custom Linux kernel) can be found in surface-linux. For NixOS I used the nix-hardware flake to simplify things.
The worst part was the compilation of the Linux kernel, that took hours on the surface. Eventually, I used the remote nix build feature to compile on a more capable computer.
I have a 2010 iomega arm board NAS. It is a board and a 3.5 HDD in an extruded enclosure. Lenovo bought them and quickly trashed the OS with google ads in the web interface, then dropped support. The HDD was aging and a bit noisy for my liking.
So I found an industrial sata SSD and swapped out the 3.5 HDD. I found an OXNAS kernel online, and installed debian (only supports older Kernel 3.xx due to limit of memory )
Runs older version of OpenMediaVault for Samba shares, and daap server plugin. But recently setup MiniDLNA.
Streams music mostly to our sound system, to my phone or PC. Does that on 256MB RAM