File: 1457989072028.png (159 KB, 409x409)

159 KB PNG

What are the pros of nixos? What do i get out of it compared to arch, opensuse or something more stable like mint? Do i get up to date shit with nixos? Will it get you a reliable system when set up right? Is updating painless? desu

File: Slopuda.jpg (446 KB, 1920x1080)

446 KB JPG

What's verifiably the least Indian distro? The most is obviously Garuda.

I'm considering using Mint again, but I see Mint has a version built without Ubuntu, relying purely on Debian (LMDE). What differences can I expect to see between the standard Mint versus the all Debian Mint? Is there a reason to pick one over the other?

I'm always torn between going full autism and build a system from bottom up, or just install my favorite DE and bloated programs to save time.

>>108118639
Can't really beat the price of 0...
I'm installing Debian right now, maybe I try mint xfce later.

>>108118723
LMDE has less features and follows a different release schedule. Honestly neither Debian nor Ubuntu are good so it really doesn't matter which distro you use as a base for Mint.

>>108118723
>What differences can I expect to see between the standard Mint versus the all Debian Mint?
PPAs dont work on lmde along with other ubuntu stuff.
Officially there's only cinnamon on lmde but if you knew what you're doing you could always install xfce and theme it yourself though at that point you might as well just install regular debian with xfce.

File: UwaaaKuma-.jpg (46 KB, 253x180)

46 KB JPG

>>108118781
>Mint without PPAs

File: screensaver.jpg (250 KB, 1203x900)

250 KB JPG

humiliation ritual
[spoilers]instructions didn't work[/spoilers]

>>108118735
Ok xfce is working, but somehow I missed the part to make a super user. How can I fix this? I don't think I have root access either. Do I need to chroot into the system? Or is there a simpler way?

>>108118790
What do you even need PPAs for? It's not even recommended to use them since they can break your system or prevent updates.

>>108118850
>t's not even recommended to use them
According to fucking who? When I was using Mint I looked up ways to update Firefox and I got PPAs before Flatpaks.

>>108118832
*krash-oh wait

>>108118862
>According to fucking who?
Everyone who has above room-temperature IQ. You're supposed to use Snaps or Flatpaks.

>>108118922
>You're supposed to use Snaps or Flatpaks
No, I don't think I will. I will use native runtimes.

>>108118922
Linux fanboys actually can't take criticism, its pathetic.

>>108118950
Suit yourself. But what you're doing is objectively retarded. You'd be better off using distrobox instead of PPAs.
>native
Meme buzzword. You're literally the embodiment of the meme "Linux users still arguing what native means instead of making the desktop work".

What is the freebsd of linux distros? I like how unified freebsd is as a system, it's pretty neat. But i dislike the aptly named cuck license.

>>108119014
>what you're doing is objectively retarded
Why is optimisation objectively retarded? I left Windows to get away from exactly what Flatpakfags wish to do to Linux.

Are transactional updates + btrfs snapshots overkill?
I want something akin to arch/opensuse TW in terms of updates, but I want to heavily not give a fuck about any updates breaking shit.
I thought about atomic distros but I'm not exactly fond of anything related to fedora and it seems like openSUSE has these little projects around but they don't seem too reliable (like I can find people saying kalpa is pretty good, but it's basically a work in progress with one bearded man at the helm).
I thought of a middle ground of using TW with transactional updates enabled.

I am trying to unfuck the following:
There is an existing incremental backup of a system which leverages rsync --link-dest . The system was nuked and newly setup using a new locale, the XDG directories are different than before. Some of the old files from the backup were migrated onto the new system, more may follow in the future.
The plan is to carry on with the old backup strategy, linking against the old inodes. Else everything that is migrated from backups to new system consumes double the memory.
rsync allegedly supports multiple --link-dest statements.
I've set up a test environment, with mock home directories, like old/home/a/file.txt new/home/b/ , where a and b are the equivalent XDG directories, and cp -p all files so rsync should consider them the same.
So far no luck, rsync -av --link-dest=old/home/ --link-dest=old/home/a ... new/home/ backup
does only create new inodes, speedup is 1,00
same goes for trying --link-dest=old/home/a new/home/b backup
This begs the question: Why does even the explicit case fail?
Does --link-dest work recursively, as in will --link-dest=a reate a link to a/b/c/new.txt in source/b/c/ ?

>>108118848
You should be able to use sudo with your password that you set up. You could login as root with sudo su and change the root password but why? This distro wants you to use sudo.

>>108119204
I am alternatively thinking I could create symlinks in the last backup of the old system that handle the translation once. It seems safe, easy, not too kuch work and should do the job, which is saving hundrets of GB by not backing up ~/Documents and similar twice

>>108119191
Sounds reasonable.

File: Screenshot_20260211_104540.png (44 KB, 3520x44)

44 KB PNG

I'm using Linux Mint and have been trying to find a solution since I switched. Is there any way that I can have an exact identical panel across both screens like on Windows or KDE Plasma? I don't want the second panel with the windows list with the big clickable buttons. I want the start menu, windows icons and everything on the right side of the panel on both screens like picrel

Two questions:
I have a bash script somewhere. I know the name of the command I use to bring it up, it works fine and does what I want it to but I have no clue where I put the damn script file itself. Is there any command to help me find it?
Also, whenever I run a bash script from a terminal and then go to exit the terminal, it hangs complaining it is still running bash, even if the script is finished or if I detach it with & after the command. I can still kill it with the keybind and everything works fine, but can I avoid that somehow?

>>108119054
it depends on what about freebsd you like. if it's just the whole "freebsd is a whole unified system", that's a tricky thing to answer, as linux both doesn't have such a thing because there are many distros, but it also does because each distro is a unified system, so it depends how you frame the question, is there a single unified system around the linux kernel? no. do unified systems built around linux exist? yes (distros).
other than that i've heard gentoo's portage is similar to how freebsd's ports system works. i don't know much else as i've never used freebsd for any useful length of time

>>108119054
Flatpak Runtime

>>108119452
>Is there any command to help me find it?
which <command>

File: 1734540225713648.gif (1.89 MB, 796x1174)

1.89 MB GIF

>>108119663
I knew there was something like that, I was trying "where <command>". Thanks.

>>108118662
Alright does anyone know how to get a pirated DODI repack of Blacks Ops 2 singleplayer working on SteamOS? I've been tinkertrannying for two days now and I have no idea why it won't launch or even get me an error message.
>installed xact
>installed vc dependencies
>using Proton-GE 10-30 in Steam settings

Plutonium works fine, I can play zombies and multiplayer through it, but t6sp.exe does not work AT ALL.

>>108118832
That used to happen on kde when you did an upgrade to specific kde packages related to lockscreen/screensaver while the lockscreen/screensaver was running.

>>108119054
Debian

>>108119684
You were probably thinking of whereis which also kind of does the same thing. There's also

 command -v 

>>108119697
I remember reading something about Black Ops needing NTsync enabled to work properly

Do people with openSUSE really cuck themselves with flatpaks because packman breaks shit all the time (specially codecs) or is this more of an exaggeration?
I've seen this already happen with Fedora, where there were conflicts with mesa packages (mesa-freeworld) breaking stuff a few years back. I was planning to use tumbleweed since I wouldn't be so dependent on flatpaks for everything (given that there's several out of date packages in F43 as I've checked which I essentially end up having to replace with flatpak/appimages or suffer the copr) but it seems like a bit of a similar mess.

>>108120103
>command -v
Wouldn't work here because it's not exactly a command, just a script. whereis works and gives the same result as which, just formatted a little bit differently.
What's the difference between which and whereis?

>>108120187
It doesnt matter as long as its either a builtin function from your shell, an alias, or a file from your $PATH.
>What's the difference between which and whereis?
Dunno.

so i heard a rumor they are killing tiling compositors in wayland is this true? some new shit called a position9ng protocol that will force you to only put certain windows in certain places at certain sizes

>>108118862
>looked up ways to update Firefox
What?? It's updated with your package manager, if you need a newer version than what your package manager provides then firefox provides a .deb package on their website

I was so happy with Fedora after first impressions but no matter what I can't get audio via bluetooth to not be a choppy, garbled mess. Tried every codec and even another pair of headphones. Was fine on Debian (and Windows).

>>108119014
Native literally means the package it works best for the distro you pedantic retard
>Distrobox
Oh, you're an immutablefag. Your opinion? Discarded.

>>108120567
>redefining the word native
I should have known you're just shitposting. My bad.

>>108120567
Native is supposed to just mean the stuff that came with your distro/system

>>108119314
I now doubt everything. I tried with the symbolic links. It made no change but I know --link-dest does follow symlinks

>>108120530
Linux Bluetooth is just always breaking adapters in different kernel versions. Wait or buy a spare adapter with a different chip for $1 / free with any purchase on aliexpress.

File: 1673796634193148.jpg (347 KB, 687x688)

347 KB JPG

Are there any experts in NetworkManager troonspeak here? What's the correct / non-deprecated way to create a network bridge now? Am I supposed to use controller / port-type instead of master / slave-type, or is it something even dumber?

>>108120530
Debian and fedora use different kernel versions. You can try the rescue kernel (it's just an older version that can only be updated manually) and see what happens, you can select it at boot time, or download the same version debian uses and discard the kernel as the root cause.

>>108120829
its still bridge-slave on system

File: bllleeg.png (37 KB, 802x599)

37 KB PNG

Just got CachyOS installed and I'm having internet troubles. When connected to WiFi, my internet speed is super inconsistent, and sometimes goes down entirely now.
All I've done so far is click "Install Gaming Support" and "Install All Packages". What do you think could be causing this instability?

>>108121252
long shot but check your wifi settings for metered connection and set it to no and also for me personally i enabled 4G decoding and resize bar in my motherboard settings and don't have a problem now

File: Screenshot_20260211_140331.png (44 KB, 834x201)

44 KB PNG

>>108121358
I set Metered Connection from "Automatic" to "No" and so far that hasn't helped.
Here is what my Steam download looks like. It goes up and down like a roller coaster.

>>108120273
the positioning would allow the window to tell the compositor where it wants to be placed, the compositor can still say no fuck you.

What's the state of encryption on linux ? I don't care about the NSA I just want that someone yanking my HDD won't be able to read my files. Like if I install CachyOS tomorrow is the drive encryption hard to do ?

>>108121534
Dunno about CachyOS but generally it's relatively easy to do with LUKS

>>108121552
Ty. Haven't tried linux in like 10 years, things seems way better now.

>>108121682
encryption already was there 10 years ago.

>>108120979
I found out that during the live environment booted from the fedora installation media the Bluetooth audio is working without issue so I'm doing a full reinstall and retracing my steps to see if a kernel update broke it or something else I installed after

I'm trying to update my Nvidia drivers on Mint since I just got a 5080, but I keep getting this error. Any ideas?
>Error while installing package: installed linux-headers-6.17.0-14-generic package post-installation script subprocess returned error exit status 11

File: slackwarepng.png (361 KB, 342x359)

361 KB PNG

>>108118708

File: screaming.gif (2.68 MB, 384x216)

2.68 MB GIF

>load torrent in qBittorrent
>accidentally save the contents to an internal drive with an NTFS partition mount point
>oh God no
>entire system grinds to a halt
>desperately try to close qB
>nothing works, not even killing it through the terminal
Linus should really just wipe the ntfs3 module from the kernel. It's just as shitty as the previous 5 or 6 attempts at putting NTFS support on Linux.

>>108121967
ntfs-3g just werks, performance aside

>>108121813
The rescue kernel should be the same version as the one in the live boot iirc

>>108121959
that's not a pipe

>>108121967
What's wrong with it? I save torrents to an internal HDD with an ntfs3 mount. Should I not do that? I'd rather keep my big files storage accessible from any OS

The thing that surprised me about getting into Linux the most is that a lot of obscure niche programs made by randos on github usually either have Linux builds or can be made to work through Wine with relative ease, while it's the big guys like Google and Adobe don't want to play ball and you need to look for alternatives to their programs. Somehow I expected it would be the other way round

>>108122219
how is google shit not running on linux?

>>108122176
NTFS support has always been shite on Linux, and since ntfs3 is a kernel module as opposed to the user-space ntfs-3g it can completely crap up the entire kernel I/O pipeline. When it fails it fails hard. I have no clue what an acceptable alternative for cross-OS filesystems is though.

>>108122031
Yeah in hindsight it was kinda dumb to not try that first but I barely just started using Fedora so it's not a big deal to reinstall stuff at least.
Also I found out the issue. It's fucking KDE (specifically kde-connect which I don't even need) instead of anything kernel related.
https://bugs.kde.org/show_bug.cgi?id=513536
dnf remove kde-connect and excluded it in dnf.conf for good measure.

>>108122240
Honestly asking, what's wrong with it? Shouldn't NTFS be well understood by now? It is truly ancient. I don't remember having issues with ntfs-3g even when I first played around with linux in 2010.
I know it doesn't have good file repair and you need to use chkdsk to fix the fs, but assuming it's in good condition, what's actually the issue? Why is it shit? Obviously NTFS is old and might not compare well with modern FS but that should be true when accessing it on Windows as well.

File: nvidia insanity.png (81 KB, 908x145)

81 KB PNG

Make it make sense

>>108122298
>installed and running
it seemed to be installed but maybe it's not running. look in journalctl -b

>>108122232
There is no official Google Drive client and the KDE's KIO GDrive app straight up doesn't work because Google doesn't give it access to your Drive (despite giving it access to other Google services)

>>108122298
>it's not running
>you have a version mismatch between smi and driver

>mom's laptop is still on Kubuntu 24.10
>I plan to install 26.04 in parallel so she can boot the old version in case I fuck up
>first I need to shrink /home to make room for the new root partition
>then I need to install the new OS manually, without using the installer
>the laptop is at the other end of the continent
Currently exploring my options. The first step is writing an initramfs script that will shrink /home at the next reboot.
Mom has a persistent live environment on a flash drive, which she can boot and that would give me remote access to change partitions or unfuck any mistakes.
The last time I did an in-place upgrade remotely from Kubuntu 20.04 to 22.04 and it went fine, but now I'd have to upgrade 24.10 to 25.04->25.10->26.04 and it would take way too long, and this would remove the ability to boot into the old version if something goes wrong.

File: 1770029366322165.jpg (285 KB, 1024x1024)

285 KB JPG

I updated my bios and it broke secure boot. I can still use my computer if I disable it though so I have two questions.

1.) should I bother enabling it? I don’t use windows

2.) if I do enable it is the process relatively simple? I’m extremely retarded and I struggle with simple tasks.

>>108122330
do you need an official client?

>>108122315
>>108122337
Turns out my computer was in secure boot, forgot to disable it since it's a new mobo

>>108122361
what does that have to do with anything nvidia related?

>>108122268
>Also I found out the issue. It's fucking KDE (specifically kde-connect which I don't even need) instead of anything kernel related.
This is so stupid it made me laugh. Shouldn't be an issue yet here we are.

>>108122369
It's a proprietary driver so Linux gets autistic about it

>>108122343
if mom can boot a live usb, she can install ubuntu on her own as well.

>>108122360
I want my files synced between 3 different computers because I have to move a lot. I can just manually upload and download files through the browser, but it's not convenient. And yes I know there are several third-party command line tools for working with Google Drive, but I'm not enough of a power user to delve into them. I'm switching to Linux because I'm tired of Microsoft's bullshit, not because I yearn to run everything through the terminal.

>>108122268
This is absolutely fucking nuts.
How is kde-connect still this buggy?
I remember having to remove it because it was causing shutdown issues on any distro whatsoever. This shouldn't be included by default.

>>108122449
An excellent example of why people call KDE bloat. All these packages only a fraction of users will need just give more room for bugs.

>>108122477
KDE wallet is the same, a source of confusion that just gets added by default. There's nothing clean about this shit, is there.

>>108119433
>Is there any way
It's software. Of course there is a way

File: Screenshot at 2026-02-11 (...).png (532 KB, 779x437)

532 KB PNG

What are some interesting niche / novelty distros? Stuff that's not worth mentioning from a technical standpoint, but autistically themed, riced, and generally centered around some special interest, like a person, club, worldview ... I'm wanna check some out in a VM for shits and giggles. Give me recs. The more preloaded with on-theme bloat the better, if it's just some custom icons and a wallpaper thats kinda gay.

shit like
> Apartheid Linux
> Hannah Montana Linux
> Red Star OS
> Born Again Christian Linux

>>108122445
Do your 3 computers share the same network often enough that you could sync across it? Syncthing is pretty useful and doesn't need any cloud accounts at all.

>>108121967
I have an NTFS hard drive and I save whatever I want there, and it literally just works and never complains about anything

>>108122506
Nyarch

>>108122532
No, two are desktop PCs in different countries and one is my travel laptop.

>>108122369
He probably broke his system in some way, secure boot does not prevent the Nvidia driver from loading.
Especially considering that he has the -open driver.

>>108122406
Nope, booting from USB is 2-3 very obvious key presses. Not even remotely comparable.

>>108122477
>>108122497
Noob here, should I just remove all the Kshit? I already uninstalled some of the apps I know for sure I'll never use like KMail, but I have this overthinking mindset of "oh, but KDE Partition Manager sounds like something important, and what if I need KFind over the standard Dolphin search one day?"

>>108122569
Still sounds like that could work, since the laptop would be synced with each desktop when you travel there. It's only a problem if you travel without the laptop or it breaks. I know syncthing supports internet sync as well but I always turned that off because I didn't need it.

>>108122631
I really like kde connect, but I don't use bluetooth so I don't know if it breaks that. I like being able to send files or clipboard stuff to my phone, and it can browse the whole phone fs in dolphin which is really nice. I used to have to manually start some wifi ftp app to do that, or get some cucked mtp view.
The only problem I had with KDE connect was getting a scare when porn links I'd copied on my desktop all showed up on my phone clipboard. Maybe turn off auto sharing

>>108122631
The PIM suite is definitely replaceable.
partition manager is ok, it doesn't exactly just break.
KFind is a fucking mess, but so is search in general. I mean they put a search in there you can't even use negated terms on.
Kate is fine though as an editor.

>>108122240
>I have no clue what an acceptable alternative for cross-OS filesystems is though.
Windows has a great btrfs driver.
Use that on Windows. You can even install Windows on a btrfs filesystem and boot that.
If you don't want to use btrfs, just use FAT.

>>108122620
installing ubuntu is clicking next a few times

>>108122754
I'd never trust windows with a linux filesystem

>>108122655
>Still sounds like that could work, since the laptop would be synced with each desktop when you travel there. It's only a problem if you travel without the laptop or it breaks. I know syncthing supports internet sync as well but I always turned that off because I didn't need it.
I've been considering my options, from switching to Mega to trying one of those CLI tools to looking into setting up some kind of self-hosted server at home. But frankly I can't be bothered with any of that right now, I still have other shit to do. I'll stick to the browser version for now and will look into other options later.

>>108122662
Does anyone use KWalletManager?

>>108122902
Fuck if I know, all I know is that for me it just gets in the way. It's a nuisance.

>>108122799
You'd be trusting an open source driver for Windows licensed under the GPL.

>>108122506
TempleOS?

>>108122933
still, nope, no way I'd trust that.

>>108119452

cat bash environment aliases file

>>108122948
TempleOS is not a GNU/Linux distro

>>108122232

you need to apply to google first to run gmail in alpine or mutt

maybe same credentials would work in other google apps

I can't stop shaking my cursor and watching it get big

>>108123150
I disabled that right away.
I don't ever want to see a cursor bigger than my dick.

File: 1770023644529118.jpg (55 KB, 427x452)

55 KB JPG

>>108123167
So do you just not use a mouse then?

>>108123233
motherfucker,

File: 1580320057386.jpg (62 KB, 800x450)

62 KB JPG

>>108123150
Holy shit
Why is this a thing?

>>108123274
It's enabled by default in macOS.
Don't you want macOS?

File: nmcli.png (3 KB, 294x38)

3 KB PNG

>>108118662
Update to this >>108115066
Anyone got any advice on how to make hardware that IS there and that IS detected by other operating systems and the bios itself work on linux? The WWAN card installed is one that already has native linux drivers baked into the kernel, so that shouldn't be an issue.

>>108121249
So it's only deprecated in their documentation?

File: 35e-3941971290.png (442 KB, 492x463)

442 KB PNG

>>108123150
>he likes watching it get bigger

i have an nvidia gpu with 1 display and an amd apu with 2 displays. can an xorg+i3 instance or a sway instance handle both at the same time?

>>108123554
I have a laptop that uses the amd igpu for the main screen and the nvidia dgpu exclusively for the external screen. I've used both and they handle it fine but x11 will pick the highest common refresh rate among your screens (ie if you have 3 screens using 60, 120 and 144, it'll silently use 60 for all of them)
If this is an issue, use wayland.
On wayland I've had performance issues related to Nvidia on this kind of setup, where if the application is displayed through the igpu it works fine but placing them on a screen that uses the dgpu caused issues like high GPU usage to display basically static applications.
Found that forcing those applications to use xwayland solves the issue for some. Use WAYLAND_DISPLAY=no as env variable, ie

WAYLAND_DISPLAY=no myapp

There is something weird going on with Plasma, for a few weeks now I keep failing my root password, it just doesn't accept it, some times in the terminal or even the logging manager, I restart the PC and boom, it works again.
>You are probably mistyping it
Nope, I just reproduced the issue by copying and pasting the password into the terminal and again. "Wrong password."

What semi-modern vidya games are supposed to run noticeably worse on Linux than on Windows 11? So far I've tried CP2077 and Oblivion Remastered and they're running no worse than on Windows

>>108121967
Mount the NTFS drive with the flag that prevents it from using non-NTFS characters

File: 1770812443983396.png (289 KB, 1080x3335)

289 KB PNG

How to destroy all of your data with one simple trick

>>108123817
>WAYLAND_DISPLAY=no
What is this crap and which LLM told you to do that? This means "Connect to the WAYLAND_DISPLAY at the socket named 'no'

The correct way to force XWayland is to either set it to en empty variable

env WAYLAND_DISPLAY= …

or unset it entirely

env -u WAYLAND_DISPLAY …

>>108124132
No, I just thought any invalid value would force xwayland. Using an empty string works on my end, actually.

What do you guys use Docker for?

>>108124169
Home Assistant and Web shit

>>108124166
I mean yeah, it technically works because you don't have a socket named that but if instead of $XDG_RUNTIME_DIR/wayland-0 you had $XDG_RUNTIME_DIR/no (which you obviously don't have but could) then it would try to connect to that.

>>108124205
For example, this works:

$  ln -sv $XDG_RUNTIME_DIR/$WAYLAND_DISPLAY $XDG_RUNTIME_DIR/no
$ env WAYLAND_DISPLAY=no …

>>108124175
Should I only be using it for web application shit or should I use it for any normal program I may think is a security hazard/only runs as a service like my IRC bouncer?

Only asking because I'm finding Docker to be a pain in my ass.

>>108124231
No reason why you can't but you may find Incus (fork of LXD by former developers since Canonical ruined it) easier to deal with if you're tearing your hair out with it. Incus is kind of like a VM but running as a container instead.

Are there any downsides to running Linux Mint via dual boot on a MacBook from like 2022? I really don't want to """(((upgrade)))""" to 11

>>108124251
That sounds exactly what I need to use, thank you anon.

>>108124169
Hax

>>108124254
>MacBook
I'm not sure if that would even work since iirc macbooks since 2020 uses some weird fucking proprietary chip architecture that's incompatible with all linux distros except for Asahi (and even that one is pretty limited as far as features go).

>>108124111
Gemini is leagues better than assGPT

>>108124580
Gemini, Claude, even Grok is better.

Just found out that asus netbooks from the late 2000's had this thing called "expressgate" which was an ultra-light linux installed on the motherboard, made to launch super fast on a netbook and get you on a web browser quickly

Im trying it out on my Asus EEEPc but obviously its mega outdated, i think the last version is from 2009
Is there a way to update this "SplashTop" expressgate? their website seems dead
seems like a neat idea

>>108124580
It's funny though:
"Oh, you're right I did tell you to delete all your data. Thanks for catching that. I was only giving you an example of when you actually want to delete your data"

At least rsync flags mostly make sense like that so you would catch it like I did unless you're blindly copying and pasting which you shouldn't.

>>108124111
man, I don't get mad about people using AI in general, but I gotta say: using a LLM to hold your hand for basic command line utilities is fucking retarded. Type "man $COMMAND" and you have access to all of the documentation you need.

>>108124752
nta but I'm using AI exactly because of snobby elitist attitudes like yours
you ask something online about linux and you instantly get a swarm of pricks on you going "ummmm sweaty just READ THE DOCUMENTATION"
no fuck that, I want a convenient and user-friendly experience and going through 5 encyclopedias' worth of dryass terminal commands ain't it
it always annoyed me and now I finally got an AI slave to do all that shit for me instead

>>108124832
>i can't read so i just get the wrong answer from a glorified spellchecker
open the schools for fuck's sake

>>108124936
literally just proving my point about being a snobby elitist cunt

File: Screenshot_20260212_032840.png (99 KB, 1169x1190)

99 KB PNG

>>108124936
Eh, it's a bit more than a spell checker even if it sometimes does give you the wrong answer once you correct it then it's fine.

I'm using it to build some scripts for managing BTRFS snapshots on my OpenWRT router (can't use snapper or any of the other tools, etc, here, need something lightweight) and it's built me a nice Python script that groups all of my subvolume snapshots by timestamp and optionally shows the size of them using du.

I'm now very carefully going to review the scripts for deleting subvolumes.

The first deletion script it made is using good non-destructive dry-run patterns identical to what I programmed myself. Because I told it to keep this pattern it understood that and made something pretty much identical to what I'd have written myself:

#!/bin/sh

TS="$1"
[ -n "$TS" ] || {
    echo "Usage: $0 <timestamp>" >&2
    exit 1
}

btrfs="echo btrfs"
dry_run="true"

while [ "$#" -gt 0 ]
do
    case "$1" in
        -f)
            btrfs="btrfs"
            dry_run="false"
            ;;
    esac
    shift
done

[ -d @snapshots ] || {
    echo "@snapshots not found" >&2
    exit 1
}

for snap in @snapshots/*-"$TS"
do
    [ -d "$snap" ] || continue

    echo "Deleting snapshot: $snap"
    set -x
    $btrfs subvolume delete "$snap"
    set +x
done

if [ "$dry_run" = "true" ]; then
    echo "!!! NO SNAPSHOTS WERE DELETED !!!" >&2
    echo "re-run with -f to delete" >&2
    echo "!!! NO SNAPSHOTS WERE DELETED !!!" >&2
fi

The Python script it's made for automatically pruning old snapshots will need more review though.

>>108124832
I mean you can do it, what do I care? Your problem is that when given advice, instead of taking it you object. Like all of the instructions to use linux are included with linux. So feel free to use a LLM to learn the wrong commands.

>>108124562
i think it can be done, but apple bootloader cant be swapped for GRUB

>>108125029
I wrote the same thing in pure bash off the top of my head. Bonus: I'm not afraid of it because I understand how it works.

Asking AI to hide bugs in code that looks right and then try to find them later is the most unproductive way to program I can imagine.

>>108125068
I can read the code it's writing. Imagine it as an exercise in doing code review for a junior engineer. Sometimes he makes mistakes but you are a senior engineer so you see that. Sometimes you say to the junior engineer "Why don't you use this pattern" and then the engineer will go "this is a good pattern. I'll reuse that".

Python is also much more maintainable when you have to sort a list of things. Show me your Bash code for doing that. Things will get messy quickly.

>>108125099
Or rather, grouping, not sorting. Sorting in Bash is pretty easy even if the syntax for arrays is funky. Grouping things is more annoying.

>>108125099
I understand the sales pitch. It's just not very useful except as a research assistant once you're past abject neophyte level.

>>108125132
Use awk for grouping.

>>108125182
>Use awk for grouping.
>>Spawning subprocesses for something that should be part of the stdlib
No.
The only thing that requires a subprocess in the Python script it made was shelling out to the "du" command to measure the size of them.

Collections and sorting, grouping, etc, should all be built-in and not some cludgy hack. I'd rather use Lua even which is installed by default on OpenWRT, although I prefer Python and this is an x86_64 router with tons of space so it's not an issue to install.

>>108125207
Oh, and I install Python to use Ansible anyway so it's not like it's only used for this one script.

I understand Nvidia .run packages are not the recommended way of installing the driver, because it is easier to let the package manager handle them along with potential conflicts which kernel updates and such could introduce.
Is there any other reason why I should not just install a .run package? I pay attention to dependencies and updates unlike some normie.

I can't launch anything that's supposed to run on a terminal via the launcher (wofi, in this case). The .desktop file does specify that it should run on a terminal.
I also tried opening a simple bash script with xdg-open and it doesn't work, it thinks it's a text file and eventually tries opening it with w3m. Trying to open some text file with neovim via thunar throws a complain about not being able to find the terminal emulator.
So the issue must be the default terminal not being set correctly. Problem is I don't know how to set it. Editing xdg-terminals.list doesn't fix it (not does it seem to do anything at all) and changing the TERM environment variable doesn't fix this either. The TE I use is kitty. the TERM variable is set to xterm-kitty (changing it to kitty or the binary doesn't fix this either).
What do?

>>108125265
It may break.
As long as you know how to fix it or are happy to keep the broken pieces then there's no reason why not.

>>108125265
You have to reinstall them every time the kernel updates, and they're also a nightmare to get working with secure boot.

>>108125285
>>108125289
I'm willing to work around these. I'll see what happens.
I just want to get rid of the extra layer of bullshit.
For example, I can't find 580.126.09 from anywhere, including negativo17 repo. It offers 590 but I don't want to install that yet...

>>108125322
Just use 590

>>108125265
Nvidia .run files are how you end up with a broken system

File: 1766651708026113.png (1.47 MB, 897x900)

1.47 MB PNG

>>108118662
I have a KVM/QEMU VM with a gluetun+qbit docker setup running. I administrate it over SSH via virt-manager. How can I make sure it shuts down cleanly without sshing into it/opening it in virt-manager? Is the best way to just open it and stop the container and then shutdown?

>>108122761
Can't tell if stupid or malicious...

Anyway, the initramfs hook and script are almost done, just need to test them in a VM and add a check for "new size > used space" before attempting to shrink the filesystem.
And some logging, which is trickier since no filesystems are mounted...

File: 2026-02-12_01-02-00.png (270 KB, 1920x1080)

270 KB PNG

DeepSeek has helped me compile classic Unix software from the 90s on modern Linux distros. You know those ancient motif apps. Then I can credit Mental Outlaw for being the brilliant creator of these super extremely light weight apps

File: Yakub Outlaw and his creation.png (1.19 MB, 1000x722)

1.19 MB PNG

>>108125908
> Then I can credit Mental Outlaw for being the brilliant creator of these super extremely light weight apps
His creations are brilliant

I don't think anyone has ran xneko and xshisen in decades. DeepSeek had to modify the source code and makefile slightly to get them working.

last update was year 2000 on xshisen. Everything else was easy to compil.e

>>108124657
I think the Linux system was on a (very slow) MMC flash, not on the HDD.

>>108125908
Why are you advertising and spamming the same posts all over /g/? Fuck off.

>>108124657
obsolete concept nowadays since we have ssds now

File: Screenshot_2026-02-11_22-41-48.png (55 KB, 108x659)

55 KB PNG

>>108125908
In a similar vein I am reviving old broken GKrellM plugins. I found a tarball of a lost volume control plugin on the Wayback Machine and added PulseAudio support. Once I have all the plugins I want, I'll see if I can upgrade GKrellM itself to work on Wayland.

File: 1770640612984020.jpg (44 KB, 450x488)

44 KB JPG

>>108125053
This. Who the fuck uses LLM before man? First you read the manpage. Then you may or may not ask the AI for a code example

>>108124562
>>108124254
Gemini says yes but i never done it so i can't confirm

>>108124752
>>108124832
If you copy and paste code from LLM into your terminal, you better know what that code does. Or you better have some kinda recovery/virtualization mechanism to revert the inevitable damage you'll do to your own system.

>reads Asahi intro page
Am I understanding this correctly that if i wanna run Linux on my M1 Mac, the hardware drivers for the CPU, GPU, and others have to be reverse engineered and rewritten into an open source implementation? And if this is necessary, why did Asahi devs choose to write the drivers using Rust? Most importantly, why should I, an Aryan white male, run CPU and GPU drivers (among others) written in Rust? Am I really supposed to believe they'll have good performance? Why don't these retards implement it in C? I mean you already reverse engineered the damn hardware, why are you reimplementing it in Rust? Do i need to rewrite this trash in C?

>>108118708
Debian Mint is still kinda white. So is Devuan

>>108118922
*whom
Also fuck flatpak I'll continue manually installing

>>108121967
>>108122240
What? I've been using an ntfs drive with Linux for almost a decade and never had any problems.

>>108122497
>KDE wallet is the same
It's not something you'll ever have to interact with. Unless maybe you have enabled the "automatically log in without requesting a password" option when booting the PC.

>>108123274
To help you find your cursor.

>>108124169
Servers, obviously.

>>108126273
>why did Asahi devs choose to write the drivers using Rust?
Rust is easier/better to work with compared to C and it's what the Asahi devs prefer. It's as simple as that.

>>108126287
>It's not something you'll ever have to interact with. Unless maybe you have enabled the "automatically log in without requesting a password" option when booting the PC.
All you have to do is open a certain popular browser and you already get a stupid prompt for it. It is something I always have to untick after a new install. I don't see why it should be enabled by default at all

>>108126273
Rust is fine, but Hector is a fucking moron dramawhore.

>>108126326
The PAM plugin unlocks it by default if you're not using password less auto-login.
If it's not doing that then your distro is crap. Look up on the Arch Wiki how to properly configure PAM KWallet. When it's working right you're not supposed to see any popups at all

>>108126337
He quit I think. That faggot isn't involved with them anymore.

>>108118662
i have this book and was cracking it again open recently
it covers linux 2.6
can it still be used to understand more modern versions, or have some things changed drastically in the kernel? this book is my only reference other than the internet.

>>108126337
>>108126287
Only retard niggers like Rust. The lowest IQ devs are the ones pushing it. Yes, the devs chose it, and that makes them retarded! Rust is not easier to work with, it's only easier to write trash code with! And there's the chicken egg effect of low iq developers preferring Rust

>Magisk gets rewritten in Rust
>stops working
>Ubuntu's GNU gets rewritten in Rust
>fails miserably

How far has asahi come? Does apple hinder their development?

File: IMG_20260212_100128_595.jpg (265 KB, 1200x1600)

265 KB JPG

So. Updated the kernel to 6.18 from 6.12 and looks like the access point of my fancy router PC works now, for like hours instead of some minutes and suddenly refusing to authenticate.

Feb 11 20:21:06 intel hostapd[278]: wlx00c0cab1ab0a: STA b8:db:38:57:de:51 IEEE 802.11: authenticated
Feb 11 20:21:06 intel hostapd[278]: wlx00c0cab1ab0a: STA b8:db:38:57:de:51 IEEE 802.11: authenticated
Feb 11 20:21:13 intel hostapd[278]: wlx00c0cab1ab0a: STA b8:db:38:57:de:51 IEEE 802.11: associated (aid 1)
Feb 11 20:21:13 intel hostapd[278]: wlx00c0cab1ab0a: STA b8:db:38:57:de:51 IEEE 802.11: associated (aid 1)
Feb 11 20:21:13 intel hostapd[278]: wlx00c0cab1ab0a: AP-STA-CONNECTED b8:db:38:57:de:51
Feb 11 20:21:13 intel hostapd[278]: wlx00c0cab1ab0a: STA b8:db:38:57:de:51 RADIUS: starting accounting session B3C24F51603CA7FC
Feb 11 20:21:13 intel hostapd[278]: wlx00c0cab1ab0a: STA b8:db:38:57:de:51 WPA: pairwise key handshake completed (RSN)
Feb 11 20:21:13 intel hostapd[278]: wlx00c0cab1ab0a: EAPOL-4WAY-HS-COMPLETED b8:db:38:57:de:51
Feb 11 20:21:13 intel hostapd[278]: wlx00c0cab1ab0a: STA b8:db:38:57:de:51 RADIUS: starting accounting session B3C24F51603CA7FC
Feb 11 20:21:13 intel hostapd[278]: wlx00c0cab1ab0a: STA b8:db:38:57:de:51 WPA: pairwise key handshake completed (RSN)

And what's the Radius thing? I don't have anything of the sort set in the config.

$ dmesg | grep -i microcode
[    0.000000] x86/CPU: Running old microcode
[    0.020711] [Firmware Bug]: TSC_DEADLINE disabled due to Errata; please update microcode to version: 0x22 (or later)
[    0.114311] SRBDS: Vulnerable: No microcode
[    0.570010] microcode: Current revision: 0x00000019

And how do I figure which blob to embed into the kernel image? It's an old Intel Haswell CPU.
t. building routers in dusty basement

>>108126356
From a users or a programmers perspective?
For users there is Linux from scratch that has up to date books and they are free in digital form.

What's the standard practice for a new install regarding partition size if I have a 2Tb nvme ssd ?
Should I split it like 512/1488 ?

>>108126326
Doesn't happen at all on my machine. You either have an auto-login like I said or your distro is utter crap.
>I don't see why it should be enabled by default at all
Because it's a way to securely store your passwords. If your distro is set up correctly then you'd never see it. If you have auto-login then you can just disable it or use blowfish encryption without a password.

>>108126379
Why not take the whole disk?

>>108126379
for me it'd be;
[--------------btrfs(max)---------------][--esp(512M)--]
why? you can do subvolumes if you want multiple volumes in the main space, and as for esp at the end, it just makes it easier to resize if needed later

File: 1770650589349546.jpg (33 KB, 563x671)

33 KB JPG

>>108126366
They still don't have USB C display out of the box, nor touch ID or thunderbolt. It probably can't run any actual games. Normal M1 can run Elden Ring on low, but i couldn't find proof that it even runs on Asahi. I suppose this is why dual boot gets used for gaming: i could just switch over to MacOS at any point. With all that said, i haven't bought a MacBook yet (waiting on dat PAYCHECK NIGGUH), so i can't benchmark it. Even though it does run on m1 chip arm64, i highly doubt it runs nearly as fast, or at all, on Asahi. Therefore, Asahi has no purpose whatsoever, especially if a power user just virtualizes Linux and uses hardware passthroughs instead of dual booting, assuming the former is possible.

>does apple hinder their development
Asahi Linux hinders their own development by choosing Rust, the language of Reddit

>>108126384
>>108126396
Interesting, thanks.

File: Jurassic Park Unix.jpg (61 KB, 514x578)

61 KB JPG

>>108126379
What's the end goal? If you want to separate the system from the rest of the drive space do the following:
>0.1 - 1 GB EFI (if an EFI PC)
>20 - 50 GB system partition
>another optional 20 - 50 GB system partition for distrohopping purposes
>rest of the drive for /mnt/whatever which hosts your $HOME, bluetooth pairings, /usr/local, /usr/src, VM images and whatever else you might consider
>>108126396
btrfs would be the smart and advanced move, whatever I proposed can be doned with btrfs subvolumes. But I'm not smart enough.

>>108126408
They have vulkan support

>>108126366
>does apple hinder their development
Decisive yes because the drivers have to be reverse engineered and rewritten. Also they can't just use Mac software directly, which i would assume btw is the workaround to make Asahi usable. Assuming this is possible, as Apple works against any type of activity like this. The Asahi devs are diverse and many are female, so Apple would see them favorably, but we are talking about Apple here. Apple is a company that stopped an ancient jailbreak method for iPhone 6 called checkra1n as recently as last year.

>>108126424
Sorry i deleted the post cuz OpenGL isn't a codec. It's cool that it has Vulkan support. Maybe i should do a benchmark in the name of science?

>>108126404
>whatever I proposed can be doned with btrfs subvolumes. But I'm not smart enough.
btrfs subvolumes are really quite easy, for a basic setup you can do for example;
- mount the root of the btrfs volume (like "mount /dev/<disc> /tmp/mountpoint", just like any filesystem)
- run "btrfs subvolume create /tmp/mountpoint/<subvolume name>" to make subvolumes in it, like "home" or "ubuntu" or something, whatever you want to identify the subvolume as
now you have subvolumes you can mount individually, so like in fstab you can add an entry like; "/dev/<disc> / btrfs subvol=ubuntu 0 0" to have it mount the subvolume located at "/ubuntu" in the btrfs volume, it shows up as if it's its own separate filesystem, only the space is shared between all subvolumes (you don't need to specify how big it is like a partition)
and yes, you can absolutely dual boot this way using only one btrfs volume

I forgot to mention that Asahi cannot legally copy and distribute a single line of actual MacOS code. Apple spends big money on this

>>108126434
According to the Asahi people they themselves are surprised that apple so far has done nothing against people putting Linux on their Macbooks. And it might be controversial but not providing documentation for your hardware is not considered hindering third party development.

>>108126404
putting small volumes at the start of the disc is a hold-over from hdd's. with hdd's they're faster near the start and slower near the end, so it was beneficial to put the OS and swap near the beginning of the disc. these days with ssd's it doesn't matter where you put partitions, so nowadays i prefer to put the big partition first, and small ones at the end, because it means if you need to change the layout later, you don't need to move the big partition, big advantage with no disadvantage, if you're using an ssd

>>108126488
>using an SSD
this is a linux thread rich boy. we can't even afford installing a flatpak here and you think we have money for a 2TB SSD?

>>108126469
The real question here is how relevant Asahi is. If it gets relevant enough, Apple may do something. I doubt this will ever happen, because Apple knows that reverse engineering projects are 1 product cycle behind at essentially all times. Especially when it's a bunch of raging retards doing the project. Why should Apple even take them seriously if they write drivers in Rust? What harm can come of such a dev team? Now, if they were high iq, and wrote the drivers in Aryan C, then those drivers might be better than Apple's own. At that point, Apple could get butthurt.

>>108126488
I just make my ESP 1GB which is more space than I'll ever need and delete old kernel images as and when I feel like it.

File: howmuch.png (42 KB, 178x162)

42 KB PNG

>>108126517
How much for installing a flatpak?

>>108126404
That's Irix, sonny boy.

>>108126396
For me, it's ZFS on the entire disk. ESP goes on a separate USB drive. No messing around with partitioning or swap.

What happens if you use another authenticator and import your codes from say the google authenticator? Does it just work seamlessly? idk how authenticators work but i want to get away from google.

>>108126659
They all work via the same OTP standard.

>>108126659
Your codes are still the same in another 2FA client if you use the same OTP secret code.

>>108126668
>>108126664
So it doesn't make a difference no matter what auth you use?

>>108126671
Yes. OTP is a standard and open algorithm. It doesn't matter which app you use.

>>108126671
As long as they use the same pattern and same secret you will get the same valid token.

>>108126398
Thunderbolt is a bitch to reverse engineer. And IIRC the person working on it before quit 2 years ago.

They recently got USB3 working so things are finally progressing. Hoping we see DP alt in three months.

My main issues for Asahi is Fedora though, too slow in updates. Should have stuck with Arch. And focussing on fucking the microphone and webcam before getting dp alt mode working is beyond me. Like stacies are going to run linux on their macbooks...

>>108126398
>thunderbolt
Linux barely has thunderbolt support in general.

>>108126696
That's not true. On the Intel side it generally works.

There are even things like bolt a daemon created secure the crap because thunderbolt has shit security and can just DMA everything.
https://gitlab.freedesktop.org/bolt/bolt

>>108122240
>I have no clue what an acceptable alternative for cross-OS filesystems is though.
Exfat or fat32 is the only two choices.

>>108126273
>why did Asahi devs choose to write the drivers using Rust?
Because they're trannies.

>>108126370
>And how do I figure which blob to embed into the kernel image? It's an old Intel Haswell CPU.

iucode_tool -S

Source: https://wiki.gentoo.org/wiki/Intel_microcode#New_method_without_initram-fs.2Fdisk_.28efistub_compatible.29

>>108126607
Some bioses will complain during boot and pause if it cant find an ESP on either the sata or nvme connected drive.

>>108126791
why would it ever do that if boot priority is set to USB?

>>108122240
>I have no clue what an acceptable alternative for cross-OS filesystems is though.
ext4, btrfs or f2fs. An OS that doesn't support any of these isn't worth using.

>>108126794
It doesn't matter because the bios is expecting your drive to have an esp and will start complaining if it doesn't even if it still allows booting from a usb.

>>108126755
exfat is only good for disposable data, since the entire volume will corrupt upon a single I/O error.

>>108126814
Doesn't happen. Stop talking shite.

>>108125936
Yes it was even advertised as an "SSD boot" on some places
In other laptops it was just an optional install into the regular hard drive which wouldnt be nearly as fast

>>108126822
I've seen it happen on dell uefi bioses.

>>108126814
I agree with >>108126822 that literally doesn't happen. You do know some people use hard drives and SSD's for bulk storage only, right? In that case there would never be a ESP on that drive and no sanely designed motherboard firmware would complain about that, at least I've never seen it.

>>108126834
Of course it's a Dell. Don't buy their crap. Imagine buying a thin-client from them you decide to use without a disk and it constantly complains about there not being a disk or the disk not having an ESP.
Utter joke of an enterprise company.

>>108126547
Damn, didn't know about them irish unixes.
>>108126785
Thanks. What I did was achtually extremely neckbeardish and autistic
>search internets on "Intel microcode"
>find out that Intel hosts its microcodes under Github instead of it being part of Linux-firmware (why?)
>read info on the front page on how to find out the right blob
>"lscpu", look at the numbers and figure you have to convert it to hex (why doesn't lscpu have an option to hexify the numbers?)
>download corresponding blob from Intel's Github repository
>create directory /lib/firmware/intel-ucode and place the blob there (first I missed the fact there should be an intel-ucode directory (had to recreate the kernel image))
>add "intel-ucode/<blob>" into the list of CONFIG_EXTRA_FIRMWARE
>works
>>108126791
UEFIBIOS wouldn't know shit about what happens after the loading of the executable, ESP can totally be located on an USB drive while the rest of the OS can be located anywhere.

>>108126835
Again, i've seen it happen on dell bioses. It expects a boot drive connected to sata or nvme with an esp otherwise it will complain during post.
>>108126842
Yeah thats pretty much how i found out, was trying to pxe diskless booting with a second hand dell optiplex and discovered their bioses have that retarded "feature"

>>108126434
>The Asahi devs are diverse and many are female
As a matter of fact, there's not a single female developer involved in Asahi specifically or the Linux kernel in general.
Not a single one.

>>108126856
>or the Linux kernel
was Sarah Sharp the only one?

>>108126814
You don't know what you're talking about. Stop embarrassing yourself and go get a spare HDD or SD card to experiment with.

>>108126826
The one I came across was way before SSDs were affordable, and Asus netbooks probably never had SSDs.

>>108126870
>and Asus netbooks probably never had SSDs.
Some of them did have SATA SSDs.

>>108126844
>>108126870
I'm getting tired of repeating myself. Some uefi bios implementations (dell in my experience) expect at least one of the drives connected through sata or m.2 to have an esp on it and if it doesn't the bios complains and interrupts boot until there's manual user intervention.
Now please tell me how i'm embarrassing myself or don't know what i'm talking about when i've experienced it firsthand.

File: 1770890829717.jpg (289 KB, 1340x1226)

289 KB JPG

>>108126865
There was Bhumika Goyal, too, but neither are current contributors.
https://archive.factordaily.com/outliers-46-bhumika-goyal-like-top-open-source-contributor/

>>108126893
your own fault for making it sound like a common issue, when it's a UEFI bug from a particular vendor.

>>108126904
Well shit, *I* have a more meaningful contribution than that in one commit. I fixed my laptop's audio jack by applying a known driver quirk.

>>108126914
Did you miss the part where i first said "some bioses"?

>>108126904
fucks sake lmao

File: 1728102161033.jpg (48 KB, 250x243)

48 KB JPG

>>108126893
The simple fact that this would make it impossible to boot without first partitioning the disk on a separate computer (so impossible to even install an OS) makes me DOUBT IT.
Highly doubt it.

>>108126926
It would be far from the stupidest thing PC OEMs have done. For years they shipped "motherboard RAID" that meant you could only access the drive ports with an OEM-specific Windows driver, and shipped it enabled by default.

>>108126921
not me, but plenty of others did, and it's clear to see why.
>Some bioses will complain during boot and pause if it cant find an ESP on either the sata or nvme connected drive.
>It doesn't matter because the bios is expecting your drive to have an esp and will start complaining if it doesn't even if it still allows booting from a usb.
took you three posts to reveal that it's a Dell specific bug. if you weren't doing it to bait people on purpose, then you're retarded and had it coming.

>>108126926
It will let you boot to usb it just first stays stuck on an error screen waiting for you to press one of the function keys to resume instead of seamlessly booting to usb.
You also gotta make sure the bootloader is installed in the esp to wherever it expects windows to have theirs in efi/boot/bootx64.efi or it will also complain about that.
Dell is retarded.

Docker race conditions with NFS mount. What a nightmare. I have no idea how to fix it. Thank god I do everything in VMs and just zfs rollback when shit explodes.

>>108126966
You're going to have to be more specific than that. What is the actual issue?

>>108126941
I mean, I've personally worked with two Acer Aspire laptops that had the fan controlled by a driver, not by the BIOS, so it was impossible to install Windows without having the laptop in the fridge, because otherwise it would simply shut down during install.
But that was an exceptional occurrence in on year with one vendor, whereas this >>108126791 fag makes it sound like a common issue with UEFI when it was one vendor on one model in one year.

>>108126356
Ik its networking stack has undergone some pretty radical changes in the last 15 years, it's probably not that useful. I wish I knew of any decent educational resources, the full extent of my knowledge comes from bricking linux VMs with kernel modules

>>108126958
Why would i be trying to bait anons when trying to explain how shit some vendors do their uefi implementation?
>It doesn't matter because the bios
In reference to the previous post about "some bioses" in the same chain.

>>108126986
Stop putting words into my mouth i never said it was a common issue and specifically said SOME BIOSES in the post you quoted.

>>108126856
the asahi page says different

>>108126941
On a scale of 1/10 how stupid was the 32bit efi on 64bit system thing that oems used to do?

>>108127034
That was a colossal fuckup when no distro handled it by default and you had to get ia32 grub efi files and manually install them yourself.
I'm glad we are past that and most UEFI that's not named Dell just works now.

>>108126966
You might be able to edit the systemd service file for docker to have it start after the nfs service.

>>108127020
Would you trust a random web page rather than a friendly anon?

>>108127008
Some can mean more than one. You should have said one.

>>108126976
Basically I have a VM with gluetun+qbit containers in it, and the qbit downloads dir is a NFS share. It's actually all working surprisingly well except that when I start the VM docker tries to access the NFS mount when I guess it doesn't exist yet and explodes. Wel, only the qbit container explodes. Gluetun keeps going, maybe because it doesn't need the share to run.

It's this stuff:

systemd[1]: mnt-qbitdownloads.automount: Got automount request for /mnt/qbitdownloads, triggered by 935 (dockerd)
dockerd[697]: time="my_time" level=error msg="failed to start container" container=random_container_hash error="error while creating mount source path '/mnt/qbitdownloads/downloads': mkdir /mnt/qbitdownloads/downloads: no such device"

I tried to mount the systemd way (https://wiki.archlinux.org/title/NFS#Mount_using_/etc/fstab_with_systemd) and the normal way and neither worked. It's weird because the automount works fine if I manually cd into it or whatever, just not during boot with docker

I tried to make docker.service wait for /mnt/qbitdownloads but that didn't work and ended up breaking the mount so bad I had to rollback. I'm going to try something like this (https://stackoverflow.com/questions/45282608/how-to-directly-mount-nfs-share-volume-in-container-using-docker-compose-v3) tomorrow and see what happens. It's not ideal because I'd still like to mount the NFS share outside of docker, but maybe it will work.

File: 032 - jd22VnX.png (8 KB, 450x401)

8 KB PNG

>>108127057
asahi official page vs random g pedo
Hmmmmmm

>>108127039
Was there an easy way to detect if the efi was 32bit or 64bit?

>>108127071
The anon is saying that men who are trans and "transitioned to a woman" don't count as women.

>>108127081
If it booted it's probably 64-bit.

Which distro is the Windows of Linux - unstable, sketchy and prone to having issues?

>>108127123
Ubuntu and anything based on it.

I use arch btw

>>108127131
Oh yeah I have forgotten 'buntu. I think it was pretty popular 10+ years ago.

>>108127170
Arch is cheeks

>>108127170
I converted my Arch system to CachyOS btw

File: 2026-02-12 13-11-59.png (7 KB, 694x240)

7 KB PNG

What are some ways to blast that QR code to all virtual terminals? It's a Wi-Fi access point and has a screen connected but no desktop environments.
Idea being that if and when someone needs to scan it I wouldn't have to log in and all that. Could just turn on the screen instead.
>>108127173
It still is if we count Mint as Ubuntu.
>>108127231
Could've just kept using Arch but installing the funky kernel.

>months after leaving wangblows behind
>firmware update for my monitor finally comes out with a fix I would want
>requires using a tool that only works on windows and updating only works through displayport
God fucking dammit

>>108127254
You can create a live USB version of Windows and just update the firmware there.

>>108127284
Yeah, all anon needs is a Ventoy'd USB with a HirenBootCD ISO to boot off of. Easy.

>>108127236
>Could've just kept using Arch but installing the funky kernel.
I was already using their kernel. I added their repos and mirror list and converted the entire system over to using their x86_64-v3 packages. CachyOS rebuilds a lot of the Arch repos.

File: a.png (14 KB, 650x400)

14 KB PNG

>>108127236
chuck it in your /etc/issue file and it'll be printed before the login prompt

>>108122799

i do not usually use btrfs in linux why would i use it in windows?

>>108126537

i did not pay vuescan myself is complicated

>>108127284
>>108127301
Didn't think of hirens being viable, thanks

>>108127438
>vuescan
I sailed the high seas for that one years back on wangblows. On Linux i just plugged my scanner in and it worked.

>>108118679
trying nixos right now as my daily driver after learning approximately zero about it beforehand and it is a gigantic pain in the ass so far, but im sure if i spent the time to actually learn it it wouldnt be so bad. that said once you figure out your config and a decent workflow it seems like it would be ideal for copying configs across servers or new computers or doing quarterly refreshes. imo it seems like the logical conclusion to ansible

File: linux meme 8.jpg (303 KB, 1600x1074)

303 KB JPG

>>108118679
Better question would be what you get over Gentoo.
>>108118723
>implying Mint is a fork
It's not. Mint Debian edition uses Debian's package repository directly, as opposed to using Ubuntu's repositories. Don't ask why but Linux distributions get away with this.
>>108121252
Wi-Fi support on Linux varies. What chip do you got?
>>108121534
Possibly the easiest way is to only encrypt your $HOME and it's mostly supported on filesystem level so you won't necessarily need another partition.

File: aaaaaaaaa.png (38 KB, 751x154)

38 KB PNG

>>108127823
>What chip do you got?
I think this is it. Its the WiFi card built into my motherboard (MSI b650 Tomahawk WiFi). I remember having to install the AMD WiFi drivers for it for the MSI website back when I was on Windows. It also worked flawlessly back on Windows, so I imagine there must be something going on with Linux to be making it so inconsistent. In fact, it worked fine for the last few hours but is now acting up again.

>>108127236
You can put it as a script in /etc/profile.d/ and it will print it every time an interactive shell is launched.

>>108122754
>You can even install Windows on a btrfs filesystem and boot that.
Has anyone tried to do this and daily drive it?

>>108127284
>a live USB version of Windows
How is this done?

>>108127934
A Windows PE environment like HirenBootCD or installing full fat Windows on a USB device via Rufus's Windows To Go install method.

>>108127951
>full fat Windows on a USB device via Rufus's Windows To Go install method.
So you need like a 40gb usb drive for this? Is there a way to make a custom windows PE environment instead of using someone else's?

>>108127992
You can but I'm not well versed. HirenBootCD is just easier.

File: 1770901948654642.png (458 KB, 1600x1074)

458 KB PNG

>>108127823
>pic
fixed your shitty meme

>>108128076
I've always found memes mocking keyboard driven WMs rather confusing since they're so simple to use

So I used Btrfs Assistant to restore from a 1 day old snapshot and I can't boot anymore, please help

>You are in emergency mode. After logging in type journalctl -xb to view system logs, reboot or exist to continue bootup
>Cannot open access to console, the root account is locked, see sulogin(8) man page for more details
>Press Enter to continue
Pressing enter only gives me
>Reloading system manager configuration
>starting default.target
And then repeats the thing about emergency mode from above

I could REALLY use some help here, I'm on Fedora KDE if that matters

>>108128090
That's not the issue, the problem with standalone WMs is that you have to figure out all the other non-WM related stuff that a regular DE does such as power management, volume control, theme management, file management, network management, ...etc most of this stuff come free with a DE but not a WM. Of course there are applications for all of that, but most people don't want to be spending hours setting all that up.

>>108128099
your snapshot most likely didn't include the /boot partition or the kernels/initramfs were stored somewhere else, and now you have a mismatch. you should probably try to rebuild the initramfs.

>>108128126
>your snapshot most likely didn't include the /boot
Weird, I've booted from snapshots in the past without issues.
>you should probably try to rebuild the initramfs.
How do I do this?

>>108127284
Turns out hirens isn't working for me. The tool isn't detecting the GPU (because it needs to do that to update the fucking firmware somehow) and even after updating the GPU driver manually it won't still work.
I'll try the Winhoes to go method.

>>108128165
chroot into your system via a live environment or rescue shell, then dracut. back up the existing one before overwriting it. the issue could be something else, i'm just guessing, because you haven't really given many details.

>>108128165
you could boot into a fedora liveusb and then chroot into your system, then use something like ugrd to build a initramfs.
that being said, not sure your were even using a initramfs to begin with. I don't use fedora but people are moving away from initramfs and just mounting / unless you have some specific need for a initramfs (encrypted / in my case).

maybe boot into a liveusb and reinstall your bootloader (systemd-boot or grub2 in most cases)

I feel pretty comfy on my FEDora install, but I kind of feel like I'm missing out on snapshots that openSUSE has. What would be the easiest way to setup something like that on Fedora myself? Also how exactly does that work, I never understood snapshots completely, gib qrd please.

>>108128210
>>108128228
Okay first of all is safe just to turn off the pc by pressing the power button or will that fuck with something? As I can't type anything into the terminal so I can't reboot through that.
>chroot
I'm setting up the USB atm but what exactly am I supposed to do when I boot into it?
Shouldn't I try to boot into the snapshot I created before restoring the older one first and see if that works? Not sure how to do that or if it's possible.
>because you haven't really given many details.
What do you want to know? System worked fine I only rebooted into an older snapshot to test an older thing before an update.

Does anyone know an old school dark theme available for both GTK and Qt that doesn't look like shit?

>>108128258
I don't understand how in the year of our lord 2026 fedora still bothers to push btrfs as the default but they don't enable a basic automatic snapshot setup by default. It's true it's not bleeding edge like something like Arch or Tumbleweed, which would probably benefit more from this. But it still boggles my mind. Well, maybe it could benefit Fedora users as well if they use third party repos prone to breaking shit.
I followed this guide at some point, but after going through the trouble, I found that Fedora was giving me issues other distros didn't, so I gave up on it. Anyways:
https://www.youtube.com/watch?v=BCULP0nDPYc

>>108128321
Btrfs is nice even without snapshots because of the compression and deduplication.

>>108128321
I like Fedora since it's not LTS with ancient packages like Debian, it doesn't come with snaps like Ubuntu, and it isn't rolling-based which fucks my adhd and ocd with constant updates that I feel like I have to do. I'd use openSUSE if it was using a point release model like Fedora since I like that it's EU based and I like the gecko logo, but when I tried it I was annoyed with how many options the installer had and I didn't like patterns that felt bloated and small paper cuts like it missing japanese/korean fonts and symbols on website out of the box which seems very lazy. I haven't encountered any instability on Fedora yet, but I feel like it'd be cool to have snapshots, if they work like system restore points in Windows I guess? Thanks for the vid, I'll check it out and hopefully not break my shit.

Are Debian's Australian mirrors being fucking DDOS'd right now? I'm getting 8KB/s downloads from apt.

>>108128353
Is compression there by default?
I remember having to still do that manually too.

>>108128363
all 9 of them?

>>108128370
compression isn't on by default. there's also no in-band deduplication in btrfs, unless he's talking about reflinks or something

wanting to try fedora asahi remix, some packages haven't been updated in 18 months...jfc fedora is a clusterfuck.

>>108128355
I like how "clean" Fedora is, but there's some shit that just busts my balls.
For instance, it seems to be pushing people more to use flatpaks. I'm honestly not fond of downloading too many of those because they just don't want to do something until the next big release. I have to wait til fedora 44 just so I can use an up to date version of easyeffects, for instance, and not one from many ages ago. The third party codecs thing isn't amazing either. But on openSUSE it's slightly worse from what I understand.
Also it feels a bit cockblocky, I tried to get a gamepad to work that requires an extra evdev rule and it wasn't as cooperative as any other distro, it took me 4 hours to try to fix it with no avail and decide to just quit it. Maybe I should try again, but I feel like something's off with either modules or something related to security.

>>108128387
See this is what I mean. They want to push btrfs they should enable at least some of the features that make it stand out over ext4.

>>108128409
Fed 44 is supposedly releasing in April.
Flatpaks? Most github stuff have fedora packages available. More so than some other distros.

File: images.jpg (8 KB, 259x194)

8 KB JPG

>>108122497
>>108122631 (Me)
>>108122930
Don't uninstall KWalletManager, worst mistake of my life
Let it to its thing in the background, I don't want to re-enter my Wi-Fi password every single time

>>108128433
I'm not building shit manually if I can help it, my man. That's for niche stuff, not fucking easyeffects.

>>108128409
>I have to wait til fedora 44 just so I can use an up to date version of easyeffects, for instance, and not one from many ages ago.
So why not use the Flatpak version then? That's the whole point of Flatpaks, to provide you with user end applications.

>>108128434
No shit. That's why other anons were saying it's better to follow the Arch wiki guide. If your distro isn't trash and you didn't tick the "log me in automatically without a password" option when installing it you will never even know kwallet exists and you won't have issues with it. Otherwise you'll have to follow this guide:
https://wiki.archlinux.org/title/KDE_Wallet#Configuration

>>108128409
The beauty of linux is that there's options for everyone, I'm a flathub enjoyer and try to get everything from flathub, I think legit the only packages I have that aren't from flathub are fastfetch and libavcodec compared to a clean vanilla install, I also remove some packages like the weird LGBT wallpapers that are part of gnome-backgrounds package and entirety of libreoffice that I don't use, but try to keep my pc as free of rpm packages as can be as I hate clutter and remnant files after uninstalls, it's one of the main things I despised in windows. My xbox series x controller works fine on Fedora, only shitty part is that I can't update firmware on linux and have to use windows for that with a microsoft account to have access to xbox accessories app. Absolute aids

>>108128447
I wasn't talking about building anything manually.

>>108128280
yes, you can power off with the button. doesn't seem like your disk is mounted so it should be fine.
you can try to replace the snapshot with a known working one, but I cannot advise how to do this as I do not use btrfs on fedora and am unfamiliar with the management of such a system.
chroot is way of changing the root directory for a system, so if you boot into a live environment you can change the unbootable system and repair it.
you may be able to just mount the disk from a live environment and repair it that way, but if you want to use dracut to rebuild your initramfs you have to chroot.
you could boot into a live environment and then investigate what is actually happening, then you can figure out a solution.
maybe search "fedora repair unbootable snapshot." and see if there is a utility available in a live environment to switch back to a working snapshot.

I just discovered btrfs requires defragmentation and rebalancing, it's annoying to discover that btrfs is not self-maintaining like ext4.
Anyway, I'm trying to understand the difference between these two, is rebalancing the same as defrag?

I now have all my btrfs subvolumes mounted with the autodefrag option, is there anything else so I don't have to worry about maintenance?

>>108128557
defragmenting flash memory is mentally retarded

>>108118662
Is the book in the OP worth reading?

>>108128122
I mean, there are flavors that already come with all that stuff done, much like a distro that comes with xfce/kde/whatever so your point is kinda moot.

>>108128473
I honestly don't remember if there was such an option. But I never needed to touch it, I just got bold with blindly removing apps that didn't seem useful like the retard I am, and from the way anons talked about it I decided it must be just another piece of bloat like KMail. Lesson learned.

>>108128523
>yes, you can power off with the button
ctrl alt delete worked, so I did that instead
>doesn't seem like your disk is mounted
So after rebooting and pressing enter again I got the following:

[Failed] Failed to mount bboot-ef i .mount - /boot/ef i
[Depend] Dependency failed for local-fs.target - Local File Systems
[Depend] Dependency failed for selinux-autorelabel-mark.service - Mark the need to relabel after reboot
[Failed] Failed to mount var-lib-nfs-rpc_pipefs.mount - RPC Pipe File System
[Depend] Dependency failed for rpc_pipefs.target
[Depend] Dependency failed for rpc-gssd.service - RPC security service for NFS client and server

So it does seem to me a mounting problem, any idea how to fix it?

File: file.png (216 KB, 772x815)

216 KB PNG

>>108126870
>The one I came across was way before SSDs were affordable, and Asus netbooks probably never had SSDs.
Maybe the tiny 1GB/2GB SSDs were affordable enough? its explicitly called "express gate SSD" in some of the documentation, and its a read-only OS that uses a 360 MB installer in the version i foudn
https://web.archive.org/web/20230521040251/https://dlcdnet.asus.com/pub/ASUS/mb/Utility/Application_Manual/EG_SSD_v14_for_CD.pdf

>>108128447
>not fucking easyeffects.
It has a package in its official repo. What the fuck are you talking about

>>108128616
If you can't follow a string of replies just don't bother

File: .png (45 KB, 886x309)

45 KB PNG

>>108128616
>>108128409
>wait til fedora 44 just so I can use an up to date version of easyeffects
that anon meant up to date

>>108128563
can you elaborate? what about rebalancing?

>>108128654
They could always switch to Rawhide today if they want. What's preventing you from doing that?
>In before it's an unstable rolling release
No, just make sure you get off of it before the branch point for Fedora 45.

>>108128669
>>108128654
Look, whatever. Flatpak it is. But this just adds to the pile.

>>108128099
>>108128126
>>108128210
>>108128228
>>108128523
Thank you to everyone that offered to help, I'm really grateful that you took time out of your day to try to help me.

I have now solved the issue, it was actually very simple. I just had to change to the older kernel version and Fedora would boot into the old snapshot. I can now boot back to my most recent snapshot and it all works. Thanks again.

>>108128557
ext4 doesn't defrag itself either
rebalancing is something you do if you change the disc layout, something that ext4 can't do in the first place

>>108128557
Think of rebalancing like internal bookkeeping. BTRFS shares all of your disk space across all of your subvolumes but this doesn't come for free. Sometimes BTRFS has to rebalance things when you make big changes to them or after a lot of churn. It's usually not a problem though.

>>108128748
glad to hear it

>>108128751
>>108128755
How often would you recommend rebalancing?

>>108128781
When you rejuggle things around a lot. Say you have a large amount of files in one subvolume and then you delete them all it's good practice to then do a rebalance, especially since the files you just deleted may not recover the disk space straight away.
Besides that it's usually not even a problem.

>>108128662
Defragging an SSD is a good way to ensure it dies in 6 months. SSD's aren't mechanical drives and they have garbage collection and TRIM and shit to do all the maintenance they need.

>>108128781
you should balance after changing disc layout, like adding or removing a disc to the volume. this will move things around such that chunks are evenly spread across the available discs again
other than that, there's not much need to unless perhaps you're really low on space, as balancing can sometimes free up some space if things are particularly messy. if you want to automate it perhaps you could do it once a month or something, it really depends on usage. i've only bothered on nearly full volumes

>updated java
>jdownloader2 won't start anymore
At least dnf downgrade makes it work again

>>108129387
>jdownloader
What year is this? Did you just step outside of a time machine?