>>100375860
Very much ready for primetime

>udisksctl power-off -b /dev/sda1

Is this even useful in current year?

Can I somewhere in Linux see in what order all systems and configs are loaded? Am I supposed to insert my own log file scripts everywhere to debug the system? If a config has all permissions, will it run anything I insert into it or are config files bound to some predefined rules?

What kind of jobs should I be applying for at junior level that let me work with GNU/Linux?

I have no professional experience with it, so I realize I'm gonna have shitty options, but I have to start somewhere. I've been using it for many years on my personal computers, so I wonder what kind of things will be different at an enterprise level.

>>100375958
consult your init documentation

I am the true master of my system when I am running Linux.

>>100375828
Best WSL distro to run? In normal cases I run RHEL distro's. Should I go Leap, Tumbleweed, Oracle Linux or just plain old Debian.

>>100375860
>>100375704
you understand why someone may not want to have to run filesystem operations on 18TB of data to fix stupid fucking defaults?

and no, 5% is NOT A good fucking default for multi-terabyte drives

you should never use passwords that contain non-ascii in them you put yourself at massive risk on multiple fronts
there's no value in using it for local passwords either

File: 1574893433727.jpg (219 KB, 1667x1250)

219 KB JPG

>>100375828

anony@fedora-pc:~$ sudo systemctl start php-fpm
Job for php-fpm.service failed because the control process exited with error code.
See "systemctl status php-fpm.service" and "journalctl -xeu php-fpm.service" for details.
anony@fedora-pc:~$ sudo systemctl status php-fpm
× php-fpm.service - The PHP FastCGI Process Manager
     Loaded: loaded (/usr/lib/systemd/system/php-fpm.service; disabled; preset: disabled)
    Drop-In: /usr/lib/systemd/system/service.d
             ─10-timeout-abort.conf
     Active: failed (Result: exit-code) since Wed 2024-05-08 12:48:29 -03; 3s ago
    Process: 14947 ExecStart=/usr/sbin/php-fpm --nodaemonize (code=exited, status=78)
   Main PID: 14947 (code=exited, status=78)
        CPU: 15ms

may 08 12:48:29 fedora-pc systemd[1]: Starting php-fpm.service - The PHP FastCGI Process Manager...
may 08 12:48:29 fedora-pc php-fpm[14947]: [08-May-2024 12:48:29] ERROR: failed to open configuration file '/etc/php-fpm.conf': Permission denied (13)
may 08 12:48:29 fedora-pc php-fpm[14947]: [08-May-2024 12:48:29] ERROR: failed to load configuration file '/etc/php-fpm.conf'
may 08 12:48:29 fedora-pc php-fpm[14947]: [08-May-2024 12:48:29] ERROR: FPM initialization failed
may 08 12:48:29 fedora-pc systemd[1]: php-fpm.service: Main process exited, code=exited, status=78/CONFIG
may 08 12:48:29 fedora-pc systemd[1]: php-fpm.service: Failed with result 'exit-code'.
may 08 12:48:29 fedora-pc systemd[1]: Failed to start php-fpm.service - The PHP FastCGI Process Manager.
anony@fedora-pc:~$ 

hitler dood
wat nou

Yakuake is gimmicky but I love it

>>100376311
so change it
getting angry over defaults while using linux of all things is just silly

>>100376353
Read, nigger. It tells you exactly what needs to be fixed.

>>100376438
you should never have to perform filesystem operations after formatting unless something bad or dramatic happens, defaults should be sane, not lurking retard-patterns that only show up after 18tb are written to disk

reserving 1TB for root "as an emergency" is inane

>>100376484
i'm annoyed because it's a rare miss from the debian team, they usually do things right

>>100375860
Which bug report is this from?

>>100376480
It doesn't give me enough information. Who is requesting permissions here? And what permissions? Everyone has read permission to that file already, at least as I'm understanding it.

>>100376548
>It doesn't give me enough information. Who is requesting permissions here?
The user your service is running as, which is specified in the systemd service file.
>And what permissions? Everyone has read permission to that file already, at least as I'm understanding it.
PHP is a terrible programming language that sometimes needs runtime write access to its own config file.

File: 1684489835672043.png (69 KB, 240x240)

69 KB PNG

Do you always use "set -e" in Bash scripts?

>>100375828
Quick question about DPI/using HiDPI monitors: I'm getting my 27", 3840x2160 monitor set up with i3wm. The arch wiki says "for Xft.dpi, using integer multiples of 96 usually works best, e.g. 192 for 200% scaling." Does anyone know why this is?
I ask as my monitor's dpi should be ~163.18 I believe, so I'm debating just setting it to 163 and calling it a day. Will doing so result in screen tearing/other issues? It's a 60hz monitor if that affects anything.

>>100376581
set -euo pipefail
if I'm being paranoid

>>100376581
Nope. Often use grep to figure out what I'm passing to it, which involves a lot of fails.

File: screen.png (52 KB, 585x462)

52 KB PNG

>>100376580
>The user your service is running as, which is specified in the systemd service file.
I'm not seeing it, unless you mean a different service file.
Man, all I wanted was to mess around with PHP for an afternoon.

>>100376043
It literally just says systemd is doing everything in parallel. I have no idea how to take control of my system.

Which distro if I want something that actually just works and I have real work to do?

>>100377461
Kubuntu, Ubuntu, or thereabouts.

>>100377461
Mint.

>>100376353

ausearch -m AVC

>>100373489
What's so great about mageia?

>>100375958

systemd-analyze plot > x.svg
xdg-open x.svg

>>100377114
php-fpm likely daemonizes itself and switches user.
it's more likely this is a selinux issue. see: >>100377547

if you're using fedora, I'll install php-fpm and see if I can reproduce the issue.

>>100377547
>>100377611
That command gives me

<no matches>

>someone decided that putting themes, fonts and icons in /usr/share/ was a bad idea so now there's ~/.themes, ~/.icons/ and ~/.fonts shitting up my home directory as well as in /usr/share/

>>100377461
The question is not which distro but which hardware.

>>100377689
Huh, that's a little annoying, especially because .local also exists.

I am wondering if there is a 2FA app that I can use between my desktop and my phone. Authy has closed their doors.

My current plan is to run an android emulator for google authenticator. This isn't a great plan.

>>100377461
Fedora

>>100377461
Doom emacs.

>>100375828
what are the default CFLAGS gentoo uses for its non-hardened x86-64 profiles?

Where do you put your personal scripts and custom binaries?

$HOME/bin ?
$HOME/.local/bin ?
Somewhere else ?

>>100378157
second one

>>100376670
Because bitmaps look like shit with fractional scaling.

updooting arch lets see if gcc 14.1 breaks anything

>>100377809
Aegis is my go to on Android. I just type in the TOTP code manually on my laptop if I need it there.

>>100378157
$HOME/bin
>why
It's easy to access.

>>100378157
$HOME/Software/My Scripts

>>100378770
>adding whitespace to things you're gonna be using from the CLI
wack

>>100378265
Why bury it one directory deeper into the dotfiles? To keep them out of sight? Wouldn't $HOME/.bin accomplish the same thing if so?

>>100377809
>>100378408
Seconding aegis. I use kdeconnect to sync my phone's clipboard with my PC when I need to type in a code.

>>100378823
I mean it's added to the path, not like I need to type it out every time

>>100378408
>>100378875
Thanks. I will give it a shot.

>>100378887
I mean if you need to edit something for example
vim ~/Software/My\ Scripts/script.sh
or
vim "~/Software/My Scripts/script.sh"
versus
vim ~/software/my_scripts/script.sh

I guess it's a matter of preference, I try to avoid spaces as much as possible

>>100378901
The killer feature of Aegis is the ability to export backups as encrypted files. Aegis + KeePassDX + Syncthing-Fork on Android and Syncthing + KeePassXC/kpcli on GNU/Linux gives you the ability to lose a phone or a laptop (or both, if you have servers you sync to!) and seamlessly restore all your passwords and 2FA to new devices, all without a third party cloud service.

>>100376581
&& everything

>>100378157
>scripts
~/Scripts/${hostname}/script.sh
>binaries
~/.local/bin
The scripts aren't in my path but it doesn't really bother me.

>>100378938
I use emacs and I don't open scripts directly from the terminal, I use it's file browser, so I've never had that issue. But yeah if you do it that way it would be annoying

>>100378157
>$HOME/.local/bin
Here, since it is in $PATH by default on Ubuntu.

Flatpak on Debian stable. Basically necessary for continually updating applications (e.g. Steam, Discord)?

reposting here because /tpg/ is too slow.
is there any way to switch currently charging and discharging battery? as far as i can tell, here's no such function in tpacpi-bat, but maybe another tool or some script can do that? i have a slice bat which i often would like to charge it before main battery and being able to manually start discharging main batt instead of slice without removing it would be nice too

>>100380168
yeah definitely

>>100380168
Yeah. There's also the backports repo for kernels+browsers.

File: 345343434532.jpg (129 KB, 675x899)

129 KB JPG

>>100375828
I've been getting MCE errors on Linux for a while, I hada b450m asrock steel legend mobo and with two different cpus (r3 1200 and r5 1600) I keep getting the same errors, I bought a new mobo and I still keep getting the same errors, they happen about once a day, my computer crashes and restarts and I get the MCE error thing. Ive done memtests and my ram is all good. I really don't know what else i can do. These issues dont come up under windows. But i really dont want to use fucking winjews full time

>>100380248
update your bios

https://www.wheresyoured.at/the-men-who-killed-google/

>>100380300
done with two seperate mobos, still happens. I guess theres a chance it might be the gpu, im at my wits ends here, the issue happens randomly, not under load

NTFS formatted stick keeps getting corrupted when putting data on it from Linux. Works with Fat32. I have confirmed transfers getting fully completed with sync. Any idea what could cause this?

>>100380900
Are you unmounting properly before eject? NTFS is very fussy about this.

>>100380900
>NTFS
Found the problem

>>100378843
it's standard

File: 1697371579983247.jpg (45 KB, 525x700)

45 KB JPG

If I run a Bash script from my file manager by clicking it, doesn't it just stay in the background indefinitely if it's expecting input from a terminal? How can I solve this?

>>100381335
either use a terminal or make it open a text box with something like zenity to get the input

how do i isolate obsidian using the linux and windows firewall? im on debian

I am in Ubuntu 23.10 and 24.04 released a a few weeks ago. I wanna know if anyone has tried it? Is it worth it to upgrade or not yet? I am okay with 23.10 so far.

File: spil.jpg (20 KB, 489x424)

20 KB JPG

hi!

>transferring 100gb of data from an external HDD to my new linux install
>do large cp operation
>"Segmentation fault"
>hmmm
>seems only half the data got copied... okay I'll delete it all and try again
>do `rm -r` on the destination directory, i leave it going for two hours but it seems to have hanged, can't even ctrl+c to stop
>investigate further
>everything i wanted to delete is indeed gone, except for one folder
>`~/phone-archive/05jan2023/Screenshots` is the only thing that remains
>try `tree`. it hangs
>try `ls`. it hangs
>try tab completion beyond `Screenshots` directory. it hangs
>try to `rm` the Screenshots directory. it works! but now any time i try to access beyond the `05jan2023` folder, it hangs. ls, tab completion, anything that tries to access that location hangs the terminal and makes it completely unresponsive

I have learned that `rsync` is superior for large file transfers.
I have backed up my current file system. (excluding the problematic corrupted area ofc).
I have checked the original data on the HDD, it is not corrupted.

questions:
q1) how do I diagnose what is wrong?
q2) how do I fix this?

solution ideas:
s1) pretend like nothing happened and simply never touch the corrupted folder again. (surely nothing bad will happen).
s2) use a metaphorical scalpel to precision cut the corruption out.
s3) reinstall system and restore from the backup i just made (which does not include the corrupted folder area)
s4) something else?

File: analyzing.jpg (60 KB, 750x675)

60 KB JPG

is there a single distro that still ships kde 5.27.11 that is not made by canonical

>>100381781
Check dmesg, it sounds like it could be a broken/unseated cable, dmesg would show something like "I/O error" if the cable/connection was bad.
Also, the is external HDD SMR (shingled magnetic recording)? If you don't know, post the enclosure name. SMR is notoriously slow and I've had very long hangs due to it before.

>>100381866
Kubuntu is made by a loving community :-)

File: hep.jpg (210 KB, 1045x774)

210 KB JPG

>>100381884
I don't care much to diagnose the link between HDD and internals. Seems to be fine, freak accident. I'll investigate it more if it happens again. Seagate OneTouch 2TB if you care.

I *am* concerned about the corrupted folder on my SSD now. How do I get rid of it?

>>100381866
gentoo

>>100377809
2FAS

How do I transition from XFCE to Trinity Desktop? It seems that I'm missing a lot of dependencies.

>>100381866
>>100382091
to add onto that, you can mask versions higher than 5.27.11 in /etc/portage/package.mask so that whenever they decide plasma 6 is stable enough, you won't update to it
>>100382134
install the dependencies?

Someone, god or whatever shows up and points a gun to your head and tells you to chose one distro and stick to it till the end no matter what. Which one and why? Would you leave your current distro?

I need a light weight browser that is somewhere between Dillo and Firefox. Would Epiphany would fall in this category?.

>>100382171
>install the dependencies?

I'm getting the following error with aptitude.

"open: 7853; closed: 10061; defer: 36; conflict: 59"
"No solution found within the allotted time"

>>100382069
Oh, it's an SSD? Then ignore the SMR thing, that's only for hard drives. You can run a SMART test to check for corrupted sectors with

sudo smartctl -t long /dev/{devicepath}

, where devicepath is the path under Filesystem when you run 'df' (dev/sdb for example) but I'm not sure about the corrupt folder, I've never had that occur. If you unmount it and run a fsck it should clear up if there's anything wrong I think.

>>100382214
debian or arch because they will never die
pretty much all other distros are at the whims of corporations or a handful of devs who may call it a day whenever

>>100382214
Linux mint probably. Comfy and stable. I only had issues with it when I have to deal with a poorly coded open source project I am contributing to, and even then the problems stem mainly because of the main dev's autismus about the dependencies needing to be bleeding edge

>>100382214
Gentoo, because I like it and I could setup a binhost for my weaker devices. Nope, currently run Gentoo.

>>100382214
i used to roll my own distro, that was fun but very time consuming. on the other hand, i've been unhappy with every proper distro i used for one reason or another (currently using arch, used gentoo and crux before that).
i think if i had to use one distro i'd rather use one that's fully under my control. right now i'm considering installing guix, which i hope will strike the right balance of tinkering autism and not being too time consuming to maintain.

File: grim-2024-05-08--18-22-29.png (36 KB, 890x288)

36 KB PNG

>>100382267
Thank you. This seems helpful. Will try it.

Hmm. Tried `sudo smartctl -t long /dev/nvme0n1` and it immediately errored out and gave me
`Read Self-test Log failed: Invalid Field in Command (0x002)`

(tried with /dev/nvme0n1p2 as well, same error)

picrel is my lsblk, if it helps

is this something that I'll have to shutdown and boot into my install usb live environment to do? to make sure the filesystem is unmounted?

>>100382514
No you shouldn't have to reboot for a smart test, nvme drives need nvme-cli to do tests for some reason. After installing nvme-cli run

sudo nvme smart-log -H /dev/nvme0

and it should tell you about errors or whatnot.
https://wiki.archlinux.org/title/Solid_state_drive/NVMe#SMART

>>100382284
That anon here. Just found out what a USE flag is and now binhost, very based. Will probably try.

File: grim-2024-05-08--18-47-01.png (125 KB, 942x869)

125 KB PNG

>>100382605
Thank you for the knowledge.

Not seeing anything that jumps out as an obvious error here.

4070 super, last I heard it was totally fucked on linux, especially for most games.
Is there any distro I could run, mostly ootb, that won't make me miserable for even trying?

Is clamscan meant to take hours or am I doing something wrong? I've seen some people say that clamdscan is preferable for speed, is that true? I'm on kubuntu 23.10, if that matters at all.

File: 26_amd_486.jpg (17 KB, 300x274)

17 KB JPG

>>100383048
What drive it has? This is why people get those NVMes.
I remember virus scans taking ages even when operating systems had like 10000 files in total. (my Windows98/Dos6.22 setup had around 10k files)

>>100383135
Yeah i'm using an NVMe, it's a 2TB Western Digital so I guess that's why?
I've never actually initiated a malware scan on Windows or Linux lol, this is all new to me.

>>100383295
>I've never actually initiated a malware scan on Windows or Linux lol, this is all new to me.
That's it. You are scanning and processing all that junk, it takes time.

>>100383349
Damn, alright then thanks.

>>100381718
Same poster here. Seems like Bluetooth works out of the box, something I had to fiddle with before which was annoying as shit.

I will probably migrate to it in a few days.

File: microchip icon.png (32 KB, 981x982)

32 KB PNG

usb 1-2.1.1: device firmware changed

How do I figure what device that is in the "lsusb" list?
Trying to figure devices out from the terminal is generally awful. Do I just have to learn this mess or are there actual tools for conveniently viewing your hardwares? GUI tools are random and provide like zero info on anything and le terminal is like
>program A gives a list of devices
>program B gives a list of devices' capabilities but doesn't use the same naming
>program C gives a list of devices' OTHER capabilities and again using a different naming
Wi-Fi devices and figuring out their capabilities is the worst.

File: 1686838758324577.png (29 KB, 597x316)

29 KB PNG

>>100381866
its off by 0.00.01 but Mageia or OpenMandriva if you're feeling a little french or brazilian respectively

>>100382220
there's a bit of a gap between dillo and firefox, netsurf for example fits there, though it's only marginally more useful than dillo
the lightest browsers that can run most sites that i know of would be pale moon, surf, and luakit. pale moon being the most traditional, being a fork of pre-australis firefox. surf and luakit are made for keyboard navigation

>>100382214
guix

File: 1696655199996398.png (191 KB, 500x381)

191 KB PNG

I have the following minor annoyance with i3. Does anyone know how to fix this?

Let's say as an example that I have two windows open where I can type text, side by side. I can switch between them with my meta key + direction, and the refocus will allow me to type in whichever one is selected. Working as expected

However, if I suddenly throw the mouse into the mix, it wrecks this delicate balance. Because now, if I have window A focused and I click on window B, and then follow that up with meta key + direction to switch back to A, I will occasionally still be typing in the now-unfocused window. To regain normal expected behavior, I would have to use my meta key + direction to "reset" it.

Do any other i3 users have a similar problem? Is it some weird bug with other thing in my setup? It's a minor issue but it's really annoying to start typing in the wrong window

File: tailspinchase.gif (17 KB, 128x114)

17 KB GIF

>>100384017
i used to use i3 but that was several years ago. my memory may be rusty but:

I'm pretty sure whatever window you move your mouse on will have the focus. If your mouse is still, then you can swap focus with hotkeys. However any mouse movement (even 1px) will focus the mouse-hovered window again.

There may be a setting that you can change in the config to disallow hovering changing the focused window, and only allow clicks to change the focused window. I'll do a quick check in the i3 docs.

>>100375828
is plasma 6 stable or are these random desktop refreshes normal? getting windows 10 flashbacks rn

File: cat83legs.jpg (655 KB, 1015x819)

655 KB JPG

>>100381904
well it seems there isn't a trace of snap/snapd on a minimal install of kubuntu. I tried searching for traces of it but other than a greyed out option in discover it looks like snap wasn't installed by default
hopefully it stays this way

File: polyporeon.gif (53 KB, 128x85)

53 KB GIF

>>100384017
yep found this

https://i3wm.org/docs/userguide.html#_focus_follows_mouse

>>100382214
Slackware. Because there is no systemd and is not too rolling release like Artix. Artix would be fine tho.

>>100382214
gentoo, it's the only single distro truly suited for any purpose on any hardware
my current main os is arch, but i've used gentoo before and are considering it again

File: heh---foolish.jpg (38 KB, 382x371)

38 KB JPG

>>100381781
solved!

i did safeguarding (made a full backup of my system excluding the corrupt folder, and was fully mentally prepared to reinstall my system)

was gonna shutdown, then boot into a live iso usb. and do some fsck stuff from there

i shutdown (it took a long time! there were about fifteen processes that were hanging and not responding (all of them `ls` or `tree` or `bash`(tab completion) or `rm` and one from `dolphin`(file gui), not even when systemd attempted to SIGKILL them multiple times. so it tried to attempt that a few times before finally giving up and powering off.

was reaching for my usb drive with the live image on it, then decided "well might as well boot up and see how it is"

so i boot up and.... it's fine!

can cd into the previously corrupt folder with no issues. all that was in there was a single 0 Byte png file.

I assume the system performed some sort of auto-fsck on startup and fixed it. Either that or it was just a transient issue that needed a restart.

Either way, seems fixed! Thank you anon >>100382267
for your help, you taught me things ^-^

>>100384135
Well I'll be. Thanks anon, I somehow missed that I had explicitly disabled this in my options years ago, because it was originally a laptop config.

Hello.
I'm new to Linux and had asked a few questions over the past few days, and, thought I'd contribute something I found informative.

"The Layman's Guide to Linux" https://youtube.com/playlist?list=PL426FzyFBwBphstVtquPX2THCD8ESzyMD&si=K7RoR0qiFPu7drku
This playlist gave me clarification on the "parts" of Linux, how they interact, what they do, and some examples of versions for some "parts" you can install or come with a distro.
For example, bootloaders, window managers, desktop environments, and init systems. I had heard of some but not all of the examples given in the playlist.

I found even the more in depth sections, such as the video on the kernel, easy to follow. I do have a technical background, but I think even for someone with less experience with computers everything should be understandable, with maybe some googling required.

Hopefully this can help out someone else who might be starting out with Linux or trying to learn more about Linux.

>>100384198
problems that fix themselves shouldn't be ignored, though there may not be anything you can do about it now
yes, filesystems with fsck will run fsck on boot (or at least you should, it's something you configure in fstab rather than something the fs driver does on its' own), so it may well be the corruption was fixed by running the boot up fsck
but, what caused cp to segfault in the first place? data-corrupting faults are not fun, since it's an nvme drive you can rule out cable issues, i'm not sure how to go about bad sector checks on nvme, though i imagine doing a secure erase on it will be enough (as it gets rid of any existing writes, bad sectors are found by attempting to write to said sectors)
i would also run memtest86+ through a cycle, just for peace of mind in that regard, minor memory issues are a nightmare and will cause the weirdest shit to happen

>>100384198
>transient issue
I wouldn't bet on it.

I wish SMART was actually a reliable health check on drives. But according to all data I've seen, the majority of drives that fail showed zero signs of trouble on SMART tests

So if they provide you a warning you should heed it, but if the drives pass the tests this doesn't give you any security that they will actually work next boot

File: Whats-an-Engine-Control-M(...).jpg (101 KB, 1140x641)

101 KB JPG

>>100384378
>I wish SMART was actually a reliable health check on drives
I wish reading fault codes was actually a reliable health check on cars lmao.
It's all just software hooked up to (sensor) wires trying to make sense of the world.

>>100375828
Would it be a bad idea to hardlink

/var/lib/gdm/.config/

to

~/.config/monitors.xml

on a single user system?
I just want my login setup to match my logged-in settings for monitor placement/scaling, and I want it to update if I change it.

>>100384759
I wouldn't use $HOME for anything but strictly per-user stuff.
idk shit about GDM but isn't there a /etc/gdm or something?
>/var/lib/gdm/
If such a dir exists, the "local" equivalent would be /usr/local/var/lib/gdm, I guess. (seems like the whole /fglt/ has forgotten the /usr/local tree)

Is it really such a huge statistical anomaly for drives to last long? I have an SSD and HDD I bought in 2011 and have used daily, and the old fucks are still going. Then again I don't put them through the ringer like data hoarders do, so I don't know.

~69k power on hours
~2100 power cycles
No reallocated sectors or retired blocks yet

>>100384826
No, it just means you lucked out with the component quality. I have several drives that have long overflowed their power on hours attribute.

>>100384826
>Then again I don't put them through the ringer like data hoarders do, so I don't know.
What effect does data hoarding cause compared "normal use"? Remember that normal use can involve having a page/swap file on the drive.

>>100384826
I have a stack of about 7 dead hard drives, all of them date from 2012-2016. And yet I have one Seagate Barracuda 1tb still chugging along.
HDD manufacturers were in a race to the bottom at that time and I got unlucky it seems.

>>100384826
They are thirdies dumpster diving old drives.
Had PCs since 1999 and maybe one drive actually broke. Many had the occasional bad sector but at that point they were getting replaced by bigger and faster drives.

>>100384875
Actually that's a good point, data hoarders by definition would write and then leave it there without deleting. What I had in mind was reading and writing a LOT I guess
In my particular setup I used a swap file on the HDD for years before I found out what zram was (that was a game changer)
>>100384859
I guess so
>>100384889
Damn
>>100384924
Funny you should mention "thirdies" because I am one. I've even had ~200 unexpected power losses according to SMART. That was before I finally got a UPS

>>100384924
I (>>100384889) am not a third worlder, all of those drives were purchased new.

Suddenly on Linux Mint, my sound settings aren't saving on restart, like the volume and balance. Always defaults to full volume, any ideas on the cause?

>>100384817
>I wouldn't use $HOME for anything but strictly per-user stuff.
I mean, it is "per user" in the sense that the data that is changing there is for me, the one user.The two files are not sourced by the same program (gdm and gnome-shell), but the format of the files should be the same. I can likely set the system one to readonly just fine since I had to create that file in the first place, so I personally don't see an issue. Was just double checking to make sure this was sane enough.
> /usr/local/var/lib/gdm/
>seems like the whole /fglt/ has forgotten the /usr/local tree
Probably because it only works part of the time. I tried it, doesn't get sourced it seems. Stuck with this solution unless you have a better one.

File: file.png (1.06 MB, 1920x1080)

1.06 MB PNG

>>100382264
update: figured it out

File: 1704847955462957.jpg (522 KB, 2368x1544)

522 KB JPG

alright, the last thing keeping me on wincuck is my monitor setup. switching to mint, so I would probably have to make a custom xorg.conf because they only have experimental waytranny support currently, but I don't want to permanently fuck anything up if possible
any anons with a similar monitor set-up willing to toss a bone my way?

File: 09-03:28:25::064279349_Arandr.png (19 KB, 880x949)

19 KB PNG

>>100385362
You're in luck anon, I have basically the same setup but flipped, but it's really not that complex.
Arandr lets you graphically arrange monitors and even save it as a script to execute whenever.

File: 2024-05-09-03:08:56-timer(...).png (3.01 MB, 5560x1920)

3.01 MB PNG

>>100385473
That arrangement gives this command:

#!/bin/sh
xrandr --output DisplayPort-0 --mode 1920x1080 --pos 0x0 --rotate normal --output DisplayPort-1 --primary --mode 2560x1440 --rate 119.99 --pos 1920x0 --rotate normal --output DisplayPort-2 --off --output HDMI-A-0 --mode 1920x1080 --pos 4480x0 --rotate left

Which is picrel in action

>>100375860

File: 1684405061228164.gif (838 KB, 475x480)

838 KB GIF

>>100385473
I don't care about bit-depth, color profile, etc. like some autists, but does Arandr have refresh rate settings? If not, setting it through xorg.conf afterwards will work together, correct? If so I will switch tonight. not joking. thanks anon for tossing me a bone

>>100385553
It doesn't but refresh rate is literally just the --rate 199.99 or --rate 60 part of the xrandr command. Also this might not even be necessary at all for Mint, desktop environments do their own things, if I were using KDE I wouldn't need to use arandr at all for instance.

>>100385498
how many little girls have you touched inappropriately? hopefully their father finds you in the future.

>>100381866
Fedora 39 KDE
just download the net install of 39 and don't upgrade to 40

t. Debian stable enjoyer on plasma 5.27.5 until 2025

File: 05ba00ad6898cf293fd78fc71(...).png (38 KB, 900x313)

38 KB PNG

>>100385594
Zero :)

>>100384694
EDC's do make sense if you know what they mean. i'm sick of hearing about people throwing parts at a car when they get a evap leak rather than looking at all the evap sensors and using their head.
oh you've got a pre cat o2 sensor code? HMM I WONDER HOW TO APPROACH THIS

File: 1707285015650703.png (1.06 MB, 894x894)

1.06 MB PNG

>>100385585
thank you. I'll see ya around :)

Anyone know the name of the volume control program used in trinity desktop? I removed a bunch of preinstalled programs that came with the de but it removed the volume control as well.

File: _sckeep_oem8eeim2f261.png (77 KB, 243x255)

77 KB PNG

>>100385498
based puppyemonade

Is there good software for burning Dreamcast discs on Linux? I can’t get imgburn to detect my drive

>>100385775
try not to touch any little kids for a few years

>>100385853
Better off installing that GD-ROM replacement mod.

>>100386025
too expensive and i am also lazy

>>100386107
Lies, if you were that lazy you'd be using an emulator, not going through the pain of burning CD-Rs.

>>100386136
This. Flycast is basically at "anything you throw at it is playable" tier

>>100386136
That's why i'm looking for an easy way to burn discs. Is it really that much of a pain?

I didn't realize that a btrfs subvolume would just appear like a normal directory. I've created this one so that it could be mounted with different options than the partition that it's on, so I have two directories. Operating on the subvolume name would use the default options and on the subvolume mount point would use its options? Why don't I see a directory for the subvolume mounted at /home which I created during installation?

>>100386343
>Why don't I see a directory for the subvolume mounted at /home
Answering myself: because @home is parallel to @, not nested in it. But is that a special case for the root partition and I can't do it for a data drive?

>>100386307
There's K3b and Xfburn, which are KDE and Xfce's CD burning applications respectively.

Laptop fell down on the floor. Started getting shitloads of hard disk errors. It remounted itself read only.

What Linux programs do I use to image this shit, mount the image and recover as much data as possible?

i was curious about a possible alternative to allow me to snip parts of my screen on linux, if the distro matters in using opensuse tumbleweed

>>100386620
snip as in screenshot? i used maim for a while, you'd have to write a shell script for it to be useful though. there's also flameshot which i haven't tried, but it's a desktop-agnostic GUI screenshot program

>>100386694 i meant in a way to screenshot a small portion of the screen, cause i mean i could just print screen if i was really desperate... but thats cringe and if i ever got multiple monitors it'd be even worse, also what do you mean by desktop agnostic if you dont mind my asking?>>100386694

>>100386770
Not part of a Desktop Environment.

>>100386770
both maim and flameshot let you select part of the screen, in fact i'd say if you can somehow find a program that doesn't let you do that, it's not worth using

>>100384826
no, that's quite normal

>>100386778
o7 got ya and thanks for the help, ill try those 2 out and see what i like better

>>100386620

grim -g "$(slurp -d)" <filename>

>>100386523
Use ddrescue. It's like dd but made for failing drives.

Any stability risks running dualboot W10LTSC & Debian side by side using dead-stock GRUB2?

Any cyber security hardening tips for GRUB in particular?

>>100384127
Snaps are fine.

File: hello.jpg (161 KB, 995x1198)

161 KB JPG

I already have a kernel that works with the PC in question and now I'm attempting to create my own EFI-bootable USB live system.
My custom kernel already had everything relevant baked in but this time I added the boot argument:

CONFIG_CMDLINE=root=PARTUUID=02e12d9a-cdfc-f140-8d66-2bd04ca41003 rootwait

and placed the kernel image at \\EFI\BOOT\BOOTX64.EFI on the EFI system partition so it wouldn't require a boot entry.
So the firmware recognizes it as EFI bootable device and the kernel gets loaded but hangs at picrel. I swear I got similar EFI setup working before. What I'm missing now? "rootfstype" argument perhaps?
Looks like I'm not even at the part where the root filesystem gets mounted. The actual root filesystem is an EXT4 on the second partition of the pendrive if that makes a difference.

>>100385733
kmix would be my guess.

>>100386427
You can create subvolumes anywhere you want. /@ and /@home are just a convention, you can have /@var_log, /@snap and whatever else you want at the top level.
The convention is to not mount the root of a btrfs volume and mount the @subvolumes instead, but you're free to do that if your use case calls for it.

File: 1689529744440856.jpg (365 KB, 2048x1536)

365 KB JPG

>>100387367
Solved: it was about 'automounter support' - CONFIG_AUTOFS_FS. Had it as a module instead of builtin.

>>100387354
Fine is fine, but some of the alternatives are better than fine.

>>100384127
>well it seems there isn't a trace of snap/snapd on a minimal install of kubuntu
Yeah the minimal Ubuntu doesn't look much different from Arch or Debian or whatever, it only gets bloated when you actually start to install stuff.

>>100387760
There is no alternative to snap, or do you think flatpak does the same thing based on internet lore?

File: lain_mg.png (1.14 MB, 1387x2113)

1.14 MB PNG

Fractional scaling on Cinnamon/X11.
It works well, but some rare types of window are scaled DOWN to less of their actual size.
One example are SDL2 message windows or QEMU.
Is that fixable without Wheyland?

>>100387998
>There is no alternative to snap
yes, there is.

File: 1715244965694.jpg (31 KB, 861x485)

31 KB JPG

>>100388135
Name it

>>100388217
The easy one is to use a better distro that doesn't try to force this snap shit on you.
But it you really have to use Ubuntu, you can tell it to not use things like the firefox snap and use a ppa instead to get a proper version.
Sad you have to go through this, fuck Ubuntu.

>>100388248
What's wrong with the Firefox snap, and when is the last time you used it?

>>100388325
dunno since it was always desnapped.
But I can tell you what's wrong with the chromium snap:
No access to /tmp
No access to webcam since some weeks ago

Had to switch to Chrome to fix the later one (which is also desnapped)

>>100387998
If you want something with the same stupid design decisions you're right, there is no alternative.

If you just want to install desktop software on the other hand then there are multiple alternatives.

does anyone still make 4:3 laptops
if no whats the last good one that will run linux well
something like a dual core yeah? or maybe not, the last time i saw a 4:3 aspect ratio screen on a laptop was in the 2000s
would linux run fine on an 2008 era japanese nec laptop? does region matter at all?

I've recently experienced a lot of timeouts recently with AMD or kernel pageflip timeouts. Is this a known issue?

>>100388706
Show dmesg?

>>100375828
WHELP I'm a little concerned
my external 32'' 4k monitor isn't working with my linux mint xfce laptop now, xfce isn't recognizing it at all
am a collegefag so this is gonna fuck with my productivity if I can't figure out the issue, surely I don't need to totally keep the laptop up to date with updates to maintain functionality with features like external monitor connections and so on right? haven't updated in a little bit to avoid this thing I'm working on from breaking but frick it, will update/restart and hope for the best

>>100388845
If you run xrandr does it show up at all? HDMI or DP? Tried a different cable (they can break)?

>boot up Mint
>sudo apt update
>sudo apt upgrade
>nothing breaks, no libraries suddenly missing, it just works
>fixes to small problems are very simple and don't require scorched earth methods to fix
>shutdown -h now, was a good day
>pretty much guaranteed Mint will boot up again tomorrow
>somehow this is bad and you need unstable distros to feel important
why are archtrannies like this again?

>>100388863
monitor is still working fine, cable is still working fine, works on other devices
yes HDMI, I should have mentioned that
xrandr doesn't seem to show it either

>>100388955
ok, it's showing the monitor seemingly as a 3520x1080 pixel monitor in nvidia settings which is obviously not right

>>100388917
I've experienced apt breaking more than once and refuse to use any distro with it. Never had issues with yum/dnf but, you know, Red Hat.

>>100388955
Have a look at the Xorg log, particularly what happens when you unplug and plug the monitor.

>>100389096
the difference is how you recover when it does break

>>100377461
It has to still be Ubuntu, right? If you're talking about pure "just werks" and "I just want to get shit done." If you just want it to work, assumedly you don't care about snaps or Canonical's other somewhat unfortunate choices.

>>100377700
Presumably he's talking about software just working. The distro matters for that.

>>100377461
Look at the distros the software you're going to use recommends. Most likely that is something Ubuntu based.

>>100388845
>>100388955
>>100389040
>>100389096
alright, the issue seemed to be in the nvidia settings, the monitor, labeled as something like 'y screen 0' was set to a seemingly arbitrary resolution of 3520x1080, which I guess resulted in xfce not knowing what to do and not detecting it at all, so I set the resolution in the nvidia settings back to the monitors default resolution, restarted, and it was detected again
question is how on earth to avoid that happening again considering nothing was updated or changed since the last time the laptop was used
......ok now linux won't boot at all, doesn't even take me to the login screen, just a black screen or it hangs on startup

>>100388917
Personally I have never really needed anything other than mint/ubuntu, but I want to get shit done not baby sit the OS. I really like being able to install my system give it maintenance once in a while and do other stuff I actually wanna do, like watching shit on youtube or working on dev projects.

I'm getting bored of Windows 11 and want to dual boot Ubuntu. But I know I'll just end up on Arch due to its superior documentation and simplicity. And then I'll get tired of constant updates and crave the Debian (Ubuntu) life.

Is anyone else stuck in this loop?

>switch to linux, choose ubuntu 23.4
>ubuntu 23.10 releases
>read about issues
>decide to wait for 24.4 so I can switch to lts
>turns out 24.4 autoupgrade is still weeks away
>try autoupgrade to 23.10
>takes 5 minutes, works flawlessly
I wish I had tried this sooner.

>>100389462
I guess the only reason to choose non-LTS over LTS is if you want slightly more recent packages right?

Although obviously right now LTS is the most recent. But normally it isn't.

>>100375828
What's the verdict on NixOS?
Is it pozzed? The Discord server certainly looks like it

>>100389907
>What's the verdict on NixOS?
inferior to guix.

>>100389907
>>100390095
also
>Is it pozzed?
yes

Has the path been well-tread at all on single-passthrough GPU setups? Saw a vid a few years ago for just running a win11 vm for gaming. Tried on manjaro with an all text guide and got an error that only yielded a single search result of someone else who couldn't get past that step.

>>100390095
>>100390113
I see, I'm on an EndeavourOS right now, I hopped from Manjaro and I like it quite a lot. I prefer using and CL instead of a GUI for the package manager and i'm glad I was forced to adapt to that.
My only gripe is it seems kinda unstable at times, but that might be because of Wayland.
But I think I'll stick to EndeavourOS for now then.

Can someone explain to me why almost every Distro Discord Server NEEDS to have a tranny or LGBT flag on their icon?

>>100390290
>Can someone explain to me why almost every Distro Discord Server NEEDS to have a tranny or LGBT flag on their icon?
it's cute

>>100390288
I'm not touching that shit, and gaming on linux is pretty strong these days anyway.

>>100390343
ok tranny
>>100390456
thanks to steam yeah, but what's your problem with endeavourOS?

>>100390523
bro, I didn't even talk about Endeavour

How to disable the native KDE tablet driver? I want to use OpenTabletDriver instead, but it seems like there's a conflict and I think it's due to KDE supporting it by default.

>>100388134
No.

>>100390290
>Can someone explain to me why almost every Distro Discord Server NEEDS to have a tranny or LGBT flag on their icon?
To make you seethe

>>100390290
US government pressure.

>>100390762
Well darn, I guess no scaling with bigger font works too.

I want to learn RedHat Enterprise Linux on my home server which right now has Fedora server on it, but I do not want to create an account to get it and I want to use it a bit for commercial use.

Is Rocky Linux a good alternative to RHEL if I still want to learn RHEL, or should I stick to Fedora if it is already a good alternative to RHEL?

File: 1689894976623691.jpg (1.15 MB, 1280x1918)

1.15 MB JPG

Yesterday I installed the update and it showed a reboot, so I postponed it to talk over Discord. I forgot to reboot, just closed, so today I see this. How can I fix it? I tried all four options and none of them is booting up.

>>100390980
RHEL is just old Fedora apart from the paid bits.

>>100390903
Really? But it's not among all distros

File: 2024-05-09-172458_1056x10(...).png (34 KB, 1056x1062)

34 KB PNG

can't import mozc dictionary in fcitx5. any tips?

>>100390980
>Is Rocky Linux a good alternative to RHEL
Yes.

>>100391468
Well it's specifically a replacement for old CentOS but it's pretty much the same shit.

>>100391468
But is it worth it to switch from Fedora to Rocky?

how do i isolate obsidian using the linux and windows firewall? im on debian

>>100390980
use alma linux.

>>100391844
Do it and find out, faggot.

is there a utility that will remove everything from a specific directory on a given timeframe?
like every month everything in my screenshot folder gets yeeted

>>100392152
If you're on systemd/Linux you could make a simple systemd timer

>>100392152

$ crontab -e
# m h dom mon dow command
0 0 1 * * rm -rf ~ /Screenshot_dir

>>100391427
Use iBus. It doesn't have this problem

>>100392467
Mind the space. This is going to delete your entire home directory.

>>100392861
whoops, fixed

$ crontab -e
# m h dom mon dow command
0 0 1 * * rm -rf ~/ Screenshot_dir

>>100392935
the absolute madlad did it again

>>100392853
i don't like ibus.

>>100392467
>>100392935
kekaroo

>>100393021
Why not? It just works.

>>100393812
where's the fun in that?

File: debian-2-logo-png-transparent.png (202 KB, 2400x2972)

202 KB PNG

It took 8 years of distrohopping to realize that my first distro was the home I was looking for all along. God, I fucking love Debian.

If you want the latest packages for select packages but not all packages would you use rolling release or stable and build the select ones from source (or snap/flatpak)? I guess I'm trying to ask what does the distro actually do for you such that you would want a more frequent release cycle? Allow integration with package manager, make sure all the dependencies are met, building with flags to target your system?

anyone else running the

init=/bin/bash

linux distribution?

>>100394193
>stable and build the select ones from source (or snap/flatpak)
this for me but it's just personal preference

>>100386620

scrot -s

>>100375828
How do I install arch with some failsafe and FAST way to unfuck myself in case I misconfigure something later down the line

>>100394193
Depends on how many packages you want, if it's more than like 20 you want up-to-date I'd probably pick rolling release. Gentoo solves this issue unironically.

>>100394193
Why would I not want the latest packages for all packages?

>>100395145
Bootable snapper snapshots

Is doing something like running debian and using archlinux in distrobox to run a de/wm like hyprland a good idea?

Give me a single reason not to use Endeavor over vanilla Arch

I'm using Bazzite (yeah yeah I know) and when someone sends me an mp4 in Steam chat it won't load, but if I open it in a browser it plays fine. Trying to diagnose what codec I'm missing - anyone know where I could even look?

How do I do alt+tab in i3?

What does a RandomX hash look like? Does it look like this?:
55072002711798,55072007954678

(MD5 hash looks like a 40-digit hex number.)

>>100395291
Bugs. Linux 6.8.x is buggy as hell, KDE 6 is buggy as hell, etc...

>>100396076
>RandomX hash
IDK, I guess it looks like this:
https://moneroexplorer.org/tx/3a9a3b2cb76153028741864d054878948030934b6baca522f0fe44f93c9c88f6

>MD5, a non-cryptographic hash
128 bits, so 16 bytes represented by 32 text bytes. 40 hex/text bytes = SHA1 IIRC.

>>100394169
Debian still supports 32-bit which is nice.

File: 1715285704876.png (202 KB, 404x507)

202 KB PNG

>>100388217
Just a good old package system?

Newer ZFS doesn't work in a 32-bit Linux computer. Ext4 does work. Does btrfs work?

>>100392152
>like every month everything in my screenshot folder gets yeeted
But why? Why make many screenshots if you aren't going to keep them?

>>100396076
>Number,number
Still don't really know what this is. I could elaborate on it.

>>100396177
I meant "non-cryptographically secure"; that also applies to SHA1.

File: file.png (272 KB, 485x404)

272 KB PNG

Recently got a shitty N100 based mini PC for use as a thin client.
Installed Kubuntu LTS and its been working great.
Been genuinely impressed with the performance for a 6w SOC.
Any one else rocking these cheap things?

>>100396322
>32-bit
Why?

Shot in the dark, but I've been getting nvme errors from the kernel over the last month that eventually crash the xorg server and dump me in a shell with no commands. This look like a lost cause or a driver issue?
errors/warns from journalctl:
>kernel: nvme nvme1: controller is down; will reset: CSTS=0x3, PCI_STATUS=0x10
>kernel: nvme1n1: Read(0x2) @ LBA 657260016, 8 blocks, Host Aborted Command (sct 0x3 / sc 0x71)
>kernel: I/O error, dev nvme1n1, sector 657260016 op 0x0:(READ) flags 0x80700 phys_seg 1 prio class 0
>kernel: nvme nvme1: controller is down; will reset: CSTS=0x3, PCI_STATUS=0x10
>kernel: nvme nvme1: Shutdown timeout set to 8 seconds
>kernel: nvme nvme1: 32/0/0 default/read/poll queues
>kernel: could not locate request for tag 0xfff
>kernel: nvme nvme1: invalid id 65535 completed on queue 11

Extended s.m.a.r.t. test completed without any errors.
smartctl -a /dev/nvme1
>Media and Data Integrity Errors: 0
>Error Information Log Entries: 1,289

I seem to be getting throttled using yt-dlp via mpv. Before I get too deep in possible solutions, am I behind on an arms race by having gotten mpv from the debian stable repo? Is it simple to remove stable yt-dlp and pull testing yt-dlp instead or would I be in for a dependency versioning clusterfuck?

>>100396382
google just hates you

>>100396380
>Error Information Log Entries: 1,289
That's an odd number. No pun intended.
Have a log at the error log with

# nvme error-log /dev/nvme1

.
There are other logs you should have a look at, check

nvme help | grep log

.

>>100396353
I'm not trying to boost my social credit score

>>100396382
It seems to be inconsistent, though. One video buffers like 8 seconds and then falls to practically nothing while another with seemingly all the same parameters buffered over 100 seconds after I skipped forward a few times (both vp9 1080p). And I've gotten shit out of some ancient 480p videos too.

So, according to discussions, this implies that I'm not being *globally* throttled but certain formats are and I need to redress that. I see that I should be able to get more detailed format info with yt-dlp parameters but the --ytdlp-raw-options flag doesn't seem to behave in a way that allows that.

And that still doesn't address whether I'm better trying to fix things by getting yt-dlp up to date vs adding aria2c or fucking with options or whatever.

>>100396507
>He thinks his hardware isn't from China
It all it anon, doesn't matter what you buy

>>100396525
Yes, but which China?

>>100396562
both

File: images.jpg (23 KB, 297x170)

23 KB JPG

hello. i hope this post finds you well. im running into a problem with docker. i have kvm installed and kvm-ok says virtualization is enabled. In docker however im trying to run a container with v-dsm. when i run it it says that dev/kvm isn't present. linux mint on a 8700k. got any ideas anons

>>100396738
VT-x enabled in bios?

>>100396751
yes

File: LookingGlass.png (348 KB, 1010x847)

348 KB PNG

Has anyone tried it yet? Why would I need 300+ UPS?

>>100396506
I didn't post the error-log output bc I think it's mostly noise from non-implemented commands? Only status_field line is
>status_field : 0(Successful Completion: The command completed without error)

The rest of these logs appear to be unsupported on the device
>NVMe status: Invalid Log Page: The log page indicated is invalid(0x2109)

>>100397180
What is Looking Glass?

>>100397180
Oh snap, thats fucking cool
Was just thinking about doing a GPU passthough setup again (since Vmware Workstation is basically abandon ware now), but think this will secure the deal for me.

>>100397229
An extremely low latency KVMFR (KVM FrameRelay) implementation for guests with VGA PCI Passthrough.

What's the bash command that lets you exchange the last command run with a new command, keeping the same arguments? I can't seem to get google to produce it.

eg:
$ ls /foo/bar
changes to
$ rm /foo/bar

>>100397248
Dumb it down a bit. What would someone use it for?

>>100397194
If the drive hasn't logged anything then it's starting to look like motherboard problems. Do you have PCIe AER enabled?
Have you tried swapping to another M.2 slot? If the problem continues to occur with the same drive then you know it's the drive, if it happens on another SSD then you know it's the motherboard.

>>100397345
Windows VM -> GPU Passthrough -> GAEMS
basically.
LG is what makes mouse/sound input lag very low so it doesn't feel like you connect to the VM via dial-up.

>>100397353
>Do you have PCIe AER enabled?
Never heard of it, I'll read up
>Have you tried swapping
That was the next plan, just been putting it off since it means full disassembly. I've got another m2 slot on the back but it's got an another unused drive in it. I'll swap this one to the back first, then try an nvme -> usb adapter. The graphics card being adjacent to the drive always bothered me, so I hope it's not a thermal issue that manifests over years.
Thanks, anon

is it me or wayland feels slower lately?
also what's the snappiest DE for an 8th core i5 + igpu? both KDE/Gnome feel slightly sluggish, or maybe I'm just messed up in the head.

>>100397614
KDE is super snappy for me. But I also reduce the animation times greatly.

https://www.phoronix.com/news/LLVM-Slower-With-AMD-Opts
AMD engineers pretending they're just now hearing about this issue kek. Everyone has known since znver1 to use haswell or skylake instead of the zen options for both llvm and gcc.

At least they're working on some of it... for llvm at least.

It's pretty crazy that a Raspberry Pi can run off a Micro SD card without the card failing within a few days. Considering that Micro SD cards aren't really designed for such write-intensive operations.

What's the deal with NixOS and Void Linux? I keep hearing about them on here. I'm usually an Archfag but just got another Thinkpad for the collection I need to put Linux on, and I'm up for trying something new if there's actually some interesting reason to use them and it's not just the flavor of the month that doesn't stand out in any way I actually care about. What makes you choose either one of them over Arch or whatever other distro?

And this thinkpad is an X61 so I have to keep everything pretty lightweight.

>>100397795
Collecting dust doesn't require many write cycles.

>>100397795
It doesn't write intensively during normal use, and SD cards are designed for intensive writing to begin with. Think tens of gigabytes per day.

>>100397832
my pihole has been up for 6 years. my other 3 raspi3's that run security cams have been up and down every 3 days for charging for 2

>>100397809
>flavor of the month
that's void in a nutshell. nix is mostly a meme and I wouldn't try an immutable os unprepared

>>100397859
>tens of gigabytes
>intensive writing
Huh?

>>100394169
You can't really go wrong with Debian except I prefer Gentoo on the desktop.
If Sid wasn't unstable and was kept up-to-date (When are you going to get Plasma 6? You're way behind other distros) I'd perhaps consider it.

Debian is great on the server though.

>>100397879
Daily. That's more writes than 90% of laptops ever see.

>>100397877
I just did a little Google Fu instead of waiting for you faggots to spoonfeed me and it looks like Void Linux is made to be super lightweight and has a custom package manager made to run fast and handle dependency issues well. Is that real or just a meme? It sounds like a good fit for the x61

>>100398045
>super lightweight
if you're stupid enough to fall for those ancient buzz words, you should definitely use it. make sure you use a musl spin and not glibc

>>100382214
>>100382284
>>100384196
This. I'm using a bunch of different distros right now and its the only one where i can everything to work correctly.

>>100398129
except my own grammar apparently.

>>100398078
>super lightweight
>ancient buzz word
It's an ancient computer. Besides, what the fuck are you saying, that the whole idea of some software using less resources to do a job than other software is a myth?

>>100396738
run dockerd as root

>>100396177
RX hash seen near the end of this video https://www.youtube.com/watch?v=siq4R-3H1sc

>>100396361
I mostly have 64-bit computers, but I do have a 32-bit "Ubuntu 14" computer which I only install security updates on. I also connect it to the Internet and use P2P with it. Hope it doesn't get hacked.

>>100398152
>that the whole idea of some software using less resources to do a job than other software is a myth?
yes. they're not doing anything notably different from anyone else except for offering a musl spin, which is the only place you'll see lower memory usage, although it comes with the cost of compatibility and performance. as for installed packages, you control that yourself.

>>100396738
Are you running your containers with:

--device=/dev/kvm

>>100398198
>they're not doing anything notably different
What about using runit instead of systemd and the custom package manager? Does runit make a significant performance improvement vs. systemd on the old hardware?

>inb4 just use Gentoo instead for runit
No

>>100383922
Thanks for the suggestions, anon.

File: 1699718081896283.png (121 KB, 420x351)

121 KB PNG

How do you decide what should be a script and what should be a function in bashrc?

>>100398275
Functions are for small, modular units of code. If you have something really large then it should probably be its own separate script.

There's no hard and fast rule though.

>>100397809
Low memory footprint and blazing fast boot up time, xbps is faster than pacman too but you don’t get the AUR though.

in fact yes githubcom/vdsm/virtual-dsm this is the project btw

File: 1712951098083416.png (26 KB, 780x670)

26 KB PNG

>>100398346
>>100398200
forgot to tag sorry and thank tou

>>100398325
Which one? NixOS or Void?

>>100397809
LFS with a hyrbid of busybox and gnu coreutils with s6 init.

>>100398171
I can't imagine that 20+ year old 32-bit machine is of much use for anything, especially considering the ratio of power consumed to work done.

>>100397809
Neither will be lighter than Alpine.

>>100398264
>init system
didn't mention it since there's many distros that offer alternative init systems like artix. systemd is understandably heavier than the simpler init systems, yes

>>100398325
>pacman
pacman was one of the fastest fully featured package managers afair. you should be able to disable some checks to speed it up, too

>>100398346
>>100398364
Try also running with --privileged maybe some sandboxing is the issue?

If it still doesn't work then I agree with the others. The problem is most likely your BIOS settings and nothing to do with Docker and your containers.

Do you know if Qemu works on the host at all? That might narrow things down.

>>100398585
>>100398364
Another issue could be you're running this on some sort of VPS or other virtual environment that doesn't allow nested virtualization. In that case the error message is accurate (there is no /dev/kvm) and there is nothing you can do about it.

>>100398585
i i have an asus board if thats important a rog whatever. vtd and vtx are enabled in bios. i did the usual kvm checks and the output xofirms that virtualization is enabled and the folder exists. tomorrow ill try with the privileges. qemu should have been installed thru the repoi think but illcheck that. as a sidenote i have also tried on a win10 machine with wls2and the standard kernel installed but same problem idk

>>100398656
>08
but its not nested its just docker on mint. im not in proxmox etc. now that u make me think about this tho docker on Linux runs always in a vm. idk maybe i can passthrough from the host to the container?

"What would an ENSH*TTIFIED Linux distro look like?" https://www.youtube.com/watch?v=XG7Dy41yJ34 - in this video he said "It's not an advertisement it's a recommendation" haha.

>>100398421
Not sure if it's from <=2004. It does get too hot and use too much CPU sometimes. What's an easy command or something to throttle the CPU? However it worked well for what I was doing:
- did: small but fast storage+networking from ramdisk
- did: large but slow storage+networking from 5TB external HDD
- did not do yet(?): large and fast storage+networking from ~2TB external SSD

>>100397250
Something like
!!s/old/new/

>>100398865
>Not sure if it's from <=2004.
If it's x86, it probably is, since 64-bit x86 has been a thing since the Athlon64 (2003) and late Prescott (2004/5). Of course most people didn't use it because XP64 was a shitshow, Linux was still in hardware support hell, and most machines didn't have more than 4GB of RAM anyway.

>What's an easy command or something to throttle the CPU?
That would depend on how old the kernel is.

>- did not do yet(?): large and fast storage+networking from ~2TB external SSD
Does the motherboard even have USB 2.0? Or PCIe for that matter? Even SATA2 is dubious for the timeframe.

New thread:
>>100399124

(I saw a link to a previous /fglt/ that I was in >>99665304)

>>100399105
I think that 32bit computer is a high-end laptop computer for it's time. x86 "is a family of complex instruction set computer (CISC) instruction set architectures" used in 32-bit and 64-bit computers. i386 and i6## are I guess some other ISAs (instruction set architectures) used in the 2000s. IDK if it is x86. It has been turned off for days because some stupid bullshit happened AFK/IRL.

It is using whatever kernel Ubuntu 14.04 uses.

If USB 2.0 existed back then, then I guess it has it. Would have to check. I am thinking that using SSD instead of HDD would still help in various ways.

>>100399105
>Linux was still in hardware support hell
Linux kernel is a monolithic kernel with support for many things in it now. Nice for working in many system, maybe not great if you want it to be small/specific for some reason.

>>100399407
for its time not for it's time

"The Intel 386, originally released as 80386 and later renamed i386, is a 32-bit microprocessor designed by Intel." how is that different from just a CPU?

"The P6 microarchitecture is the sixth-generation Intel x86 microarchitecture, implemented by the Pentium Pro microprocessor that was introduced in November 1995. It is frequently referred to as i686." How does microarchitecture differ from a CPU?

Reading is 4 nerds.

>>100399407
>It is using whatever kernel Ubuntu 14.04 uses.
I'd recommend upgrading to Ubuntu 16.04 (Xenial) which was also the last one with 32-bit support. You could always restore from a backup if it's worse somehow.

>>100399495
Microprocessor is a specific type of CPU. Definition of microprocessor:
>The entire CPU of a computer on a single integrated circuit (chip).
Single-chip CPU is different from chipset CPU. If multiple CPUs show up in htop you either don't have a microprocessor or it has multiple cores.

>>100399538
Didn't know that. I am using Lubuntu ~19 based on Ubuntu 14 which fits on a CD. I assume it can boot from a DVD, and in that case I think I could use 32bit Debian or Ubuntu 16 or some newer 32bit Lubuntu.

The 32bit laptop has 4 CPUs as seen from kernel messages.

>>100399713
>The 32bit laptop has 4 CPUs as seen from kernel messages.
i'm curious to know what cpu is quad core but only 32bit, i've never heard of that
quad cpu and 32bit, sure, but you're saying laptop

>>100399713
>The 32bit laptop has 4 CPUs
Then it's newer, probably at least a Core 2 Quad or something similar. That should be a 64-bit CPU.
What "model name" is displayed with

cat /proc/cpuinfo | head -n 10

?