When using WSL, be sure to not mention anything about that when reporting bugs because that’ll just confuse the issue for the maintainers. They like having that casually mentioned about 20 messages into the troubleshooting process.
Pff, issue reports should ask for the output of ‘uname -ar’. It clearly shows its wsl as wsl runs a special kernel
I’m a big fan of going on WSL forums and letting them know everything is working well for give or take 20 messages, then I let them know I need help troubleshooting.
I don’t think anyone is a “wsl user” so much as they’ve found themselves in a position where the lowest friction solution is utilizing wsl for a given situation.
Around 2019, even up until like 2022 if you wanted to run docker in windows, that was how to do it.
I learned the shell in wsl before I switched to Linux full time. I wasn’t trying to learn it intentionally. Just didn’t want to develop software on windows. It’s a great gateway drug that reduces friction by a lot.
The terminal is a pathway to many abilities some consider to be unnatural.
It’s not a story the Vibe Coders would tell you
Pretty much my situation. Work stuff, Windows machine, but Linux/Docker workflow and I refuse to let go of my POSIX shell.
All the good stuff available and you choose a POSIX shell? To each their own I guess.
Granted, I still prefer it to PowerShell, but only in how it feels, not conceptually.
Eh I’m not hard set on full spec compliance. I use ZSH, it’s not technically POSIX compliant but close enough that I virtually never have to think about it. Technically correct would probably have been “sh derivative” or something.
I still prefer it to PowerShell
Who hurt you?
Why?
Not enough caffeine. Please disregard.
That’s where I was a few years ago, and then I switched back to proper Linux. I was only keeping Windows at all for games, but then most of the games I played started working fine on Linux (thank you, Valve).
Plus, I tried doing some TensorFlow stuff with CUDA (Nvidia) GPU acceleration. In theory, you can do it in pure Windows, but nobody has bothered trying to do that. You’re on your own if you try it. The usual way is to do GPU passthrough to WSL. There have been three different ways to do that over the years, only one of which currently works. If you happen to Google a page that tells you one of the wrong ways, there’s a good chance you’ll need to reinstall to get it working the right way.
Using pure Linux for this stuff is no problem. Just use Nvidia’s server drivers instead of gaming drivers. All the AI datacenters are using Nvidia GPUs on Linux, so Nvidia is highly motivated to make this work. Someday, Windows might be as easy to use as Linux.
What’s the current best way to run docker on Windows?
I’m still using wsl(2) for that in 2025 because it seems to be the path of least resistance on Win11.
That could very well be the best practice. I haven’t had to run docker in windows since then.
Just install the docker engine in wsl like anywhere else and avoid docker desktop if you can
Look what they have to do to mimic a fraction of our power
Look what they have to do to mimic a fraction of our power
I think that sounds more like Cygwin.
I forgot about Cygwin… ugh
WSL is cool when you first use it then you realize it actually kind of sucks
Apt username
Yum username.
Yay username
Dnf usernam
This sort of elitism really hurts adoption
How? WSL is absolutely awful for adoption. Theres no GUI, it bearly runs GUI apps, and you have to manually configure it. If my first experience with Linux was WSL I would never touch Linux again.
It has proper gui support now?
Let me clearify, you cant run full desktop environments on it. You cant even run a window manager like Sway on it.
You can but it’s a very cursed way to do it. You need to run an X server on windows and set up x11 forwarding. I remember trying it ages ago and it sort of worked. I ended up giving up on it as I ended up just using a live boot.
https://stackoverflow.com/questions/61110603/how-to-set-up-working-x11-forwarding-on-wsl2
That seems like its not stable and the software is likely to either be undermaintained or unmaintained
Vcxsrv has been around for ages. It’s also more documented now. I tried it right when wsl2 released.
The only way it could really do it would be for the Windows shell to get out of the way, which it won’t.
You can with effort have some sort of abomination where you get inflicted with both UI designs at the same time…
Yes
You can sudo apt install firefox. And you will get a firefox (wsl ) icon on your start menu
And i always thought Real™️ linux users dont need a desktop manager? No wait they need arch with a tiled window manager because it looks cool but actually dont do annything besides configure their install.
Real™ Linux users stare at their desktop until they reboot into Windows so they can acturally run software :3
If you actually do work, getting used to a tiling WM is like a drug. I can’t live without it now.
(that’s a lie, I do at work cus I’m forced to use Windows, so WSL with tmux is an acceptable alternative)
Ok but what is your job then? I do software development and in no way would it make my work faster if i can type 2 more words a minute because i dont type that much. Most time is used to read sourcecode, chassing references through the codebase and reading api references in the browser. If i have to do more hardware related stuff i would never want to use a keyboard to scroll through datasheets.
When working with Linux I want a CLI and GUI for some applications. No need. To be fair, I primarily use windows because VS22 with resharpen is pretty nice (with graphical debugging).
Yes but thats sorta the point, WSL users like you are Windows users. Not really Linux users, you run a glorified VM. It makes perfect sense for devs to get annoyed when WSL users complain about WSL bugs to package or distro maintainers. Theres nothing wrong with that obviously but its still misleading whenever a WSL user calls themselves a Linux user (not to say that applies to you)
This pleb thinks we should listen to them. Frankly, this neighborhood is done for!
I knew, as soon as they installed that damned GUI, that we’d have Windowers coming around with their “Windows key this and WSL that”. I’m going to have to move to BSD at this rate, I hear they have a more permissive license. I was telling my friend Margaret just the other day that I was meaning to move to BSD. That and that I wanted to get a shrubbery, for the garden.
A shrubbery‽
The same
On another side it stops people from switching cause as better it’s working less reason to switch
When someone ask for help with a Linux issue: 🤩
It’s a wsl bug: 🤭
Open source now tho
What’s the benefit for that? Can you fork it and run a copy in windows (or is it too integrated in wondows) or is it just Microsoft wanting your free labour (bugfixes etc)?
Mom! Windows is doing EEE again.
At least you can look at how it works under the hood I guess
Oh yeah, that is true !
Me looking disgusted at myself in the mirror, for I am doomed towards eternal Microsoft-dependency at work.
Still better than cygwin
God I remember suffering through cygwin at a VB shop back in '17… literal hell
And msys
Small blessings. Seeing a WSL user means that some dev out there didn’t have to implement Windows support.
Yea… But I have 3 Linux VMs running in Hyper-V. That counts for something right?
It is honestly better
fuck it
run windows in a virtual machine in Linux, then run wsl within it.
Most WSL users I know all run Linux at home; WSL is the best they can get at work.
True
it’s a unix system on top of a windows system; i know this!
Yo dawg…
We don’t talk about
Brunothe Microsoft POSIX subsystem
I tried to get the *arr stack running on it at one point, using Docker.
Do not do this. Just install the Windows apps. Yes, it’s a mess. Yes, they work.
Just curious, what were the issues?
It was a while ago now, bit I think it was trying to get all the individual bits to talk to each other (radarr to prowlarr, etc). I was following some guide and that’s where it all fell apart.
Sounds like a network configuration issue of the containers - you either have to use the host network (probably not recommended) or to map the necessary ports of each app. But trying to do that in WSL sounds like an extra layer of fuckery that you don’t necessarily have to deal with. Running Docker directly on Windows sounds like the more sane thing to do in that case.
Jokes on you, docker on windows uses wsl too :')
I know, but it’s managed by Docker, i.e. you don’t have to do anything special.
Yeah, it was blocking the networking between them, and after Google failed me for an hour, I realised they all had Windows installers so there wasn’t really a lot of point persevering with weird half-broken versions of Linux and Docker.
Windows users when they see a wine user???
I’m glad wsl exists so I don’t have to bother with windows and people can still run my programs.
Apt. Wine is pretty disgusting as a beginner linuxer.
Proton however is incredible. An invisible solution that just works with no petformance impact.
And “hardcore tinkering” is just changing the version from a dropdown menu to get old games to work.
I still don’t know what WSL is for.
To keep developers on the platform, because their own ecosystem is shit.
Honestly if it frees me from MinGW I would be happy
I’m happy the Cygwin and Mingw days are over. WSL2 works very well.
Windows Stolen Linux
Serious answer,
Windows subsystem for Linux.
https://en.wikipedia.org/wiki/Windows_Subsystem_for_Linux
Windows Subsystem for Linux (WSL) is a component of Microsoft Windows that allows the use of a GNU/Linux environment from within Windows, foregoing the overhead of a virtual machine and being an alternative to dual booting.
Them tryna suck us back into their dysfunctional os, not gonna work on me Mr Micro
Oddly enoigh, they recognize this and are patching the hundreds of tiny holes. I would argue they began trying (IMO malformed) fixes back since the launch of windows 8 and .NET. It’s backwards compatability means tiptoing around some pretty huge tech dept. (Windows was DOS and had no security model at one point) Each time they try to pull people off of their older SDKs. If and when they dont stick, the pile of stuff to support grows one more.
(Also WTH where they thinking with windows 8 apps!? The oversimplicity of the UI leading to huge patches of unused screen space, the art design or lack thereof, the janky unpolished UI elements. It’s embarrising for how much pride they had for it.)
I do know what it is, I just don’t know why you’d use it instead of proper linux, or a vm.
Bad management
Run Linux stuff on Windows.
A big use case is development with Docker containers.
Thanks - I can kind of see that, as docker on windows is majorly broken. I think I’d just run it in a linux vm, as I do with most of my developing, but I can see some might not want that overhead.
Consider dropping Windows instead
That’s the best bit about WSL (at least, version 2) is that it is a VM running a full version of Linux using Microsoft Hypervisor. There’s a bunch of drivers included that allow Windows and Linux to share filesystems and if you run Wayland/X apps in Linux they run on the Windows desktop.
Sharing filesystems could be useful, I can see that.
I do that with target dev platforms anyway, using things like NFS, samba and sftp, but I do see that it could work well for this.
WSL these days is basically linux running on hyper-v tech natively in Windows.
I only use Windows because I have to work with a corporation’s IT helpdesk staff to get on their VPN if I want to do contract work for them. They are not likely to help me get connected from Linux; they’ll just find another contract dev. Once in, I do everything in Linux because my code will ultimately run in a Linux cloud container of some sort. WSL works well enough for me to do this. I’d rather have Linux on bare metal, but whatever. I’m in; I’m coding; I’m getting paid. I’ll put up with a little bit of suck.
I love having it at work, so I can write and run bash scripts on my Windows work PC.
I have dozens if Linux servers available to me but sometimes it just is easier to run a script locally.
I too do that, working from a windows vm and writing code for linux - but I push it to a linux vm for testing. Never occurred to me to use WSL and have another environment to configure and maintain for dev that’s different to the target one.
But fair play if that suits you! Each to their own, and I’m sure I do things that make no sense to others.
Anecdote: I have an IDE that only works on Windows that can build applications for Linux. I use MinGW as part of the packaging process (AND I FUCKING HATE IT OH MY GOD. All of the pathing is broken!). As of yesterday I learned that WSL is a thing that might replace MinGW and make some processes of packaging for linux targets a little easier.
I’ve used both. What I can tell you is that moving to WSL is like moving to Linux wholesale. Treat it like porting your toolchain.
IIRC, MinGW tools will happily take windows style paths (e.g. “C:\Users~myuser\projects”). If your tooling/scripting depends on being able to use Windows style paths, you’ll have to fix that first or you’re going to have a really bad time. There may be other small differences between MinGW tools and what ships on Ubuntu (or whatever Linux you decide to use in the WSL).
My company only allows us to use the company-provided Windows image, so I do all my work inside a WSL2 tmux session.
JetBrains IDEs and VSCode also have WSL connectors so it works acceptably well.
It also handily dodges all the Windows security policies (like installing software). You can even run Xorg apps from it.
I’m still forced to use MS Teams and Outlook, though…
yup!
its for when the reqs include azure ad and the whole office has a m$ fetish yet you still gotta get your bag without losing your decades-built toolset AND you have a choice at all
I don’t know what WSL is.
