Probably due to a power outage, I suddenly lost access to a connected USB HDD yesterday. According to parted, I get the message “unknown partition table,” and gdisk says the GPT is corrupted. Using testdisk, I was able to copy the files to another drive and restore the partition table and mount the HDD.

Is it possible that the partition table was damaged by the power outage, or does this point to a different problem? Can I safely store data on such an HDD again, or should I replace it?

I have an oddly-specific networking question that I'm hoping to get solved (and would like to have Lemmy be the home of the solution, in case I stumble upon this myself.)

I have a device that needs to "phone home" on a weekly basis. This device needs to have a static IP in order to be "valid". On windows, I usually can get it to work by starting openvpn, plugging in the external device in question and "routing that" through Openvpn. You do this by simply connecting the external nic and then right click on the nic in the control panel and set the "Allow other network users to connect through this computer's internet connection" along with assigning the external device to the OpenVPN connection. This routes all internet packets from X device to the internet via the OpenVPN connection without a major problem. I should note that it's extremely fickle, and it hurts my ability to remote in the machine on occasion, so I'm looking to change it.

What I'd like on my linux system is to simulate the same setup, but with 1) a permanent "connection" to the openvpn end point that is always available but not used by most networking tasks on the computer (for example, browser will use standard ethernet connection) and 2) The external nic will automatically route itself through the OpenVPN server, acting as a "router" for a (singular) external device that needs to access the internet.

Is this possible? Are there any guides for setting up this kind of complex network setup? Is it possible to do without going crazy and making my computer a full-blown router (like, doing DHCP work and the like)?

I'm building a desktop that I want to use both as a LAN server and a media player by the TV.

I've been looking into distros and or solutions that would best solve it, and not gotten too far. Aiming for Debian currently, and testing if openmediavault can be also ran with kodi. I want me and my two roomates to be able to access files on it, and to use qbittorrent on the pc directly. acessing it outside of our house is not a big priority but would be a nice plus

Question is: is there some distro or solution that is for me, and I just don't know it - and if not, do you have any suggestions for my setup? :)

(Pc is gonna be run with Ryzen 5 3600x, GTX 1080ti, 16gb Ram, 128SSD for OS and, 5 3TB HDD in raid.)

When clicking using the bottom right side of the touchpad on my NexDock nothing happens when using EndeavourOS with XFCE4. When connected to Windows I am able to right click using the touchpad without issues.

Does anyone know how I can remedy this? I feel like it's probably something obvious.

I tried looking under the Mouse and Touchpad settings application but don't see anything that looks relevant. The NexDock does appear as multiple devices. SINO WEALTH NexDock Touchpad Keyboard Consumer Control, SINO WEALTH NexDock Touchpad Keyboard Mouse, and SINO WEALTH NexDock Touchpad Keyboard Touchpad are all listed but only give me the option to switch hands or adjust pointer speed.

I recently had a... spicy game crash for some reason, and now that game's name shows up in DrKonqi. Can i delete it? I tried removing the systemd core dumps, running balooctl purge, deleting the cache folder but it won't go away. I don't want it to pop up if i ever have to show crash reports to someone.

i want to place this cava module in my bumblebee-status, is this feasable? it seemed like you could via the documetaton.

I've been researching information about /etc/machine-id, a file that contains an ever-persistent machine ID that identifies the install across hardware or netwotk changes, and that is world-readable.

Most documentation I've seen says it is mostly safe to change this file and generate a new ID on shutdown, and there are example scripts to do it via eg.: systemd or rc.shutdown . That's nice, but... we're on Linux, we don't "shutdown" our machines, what do they think we are, Windows users? We don't shutdown at least intentionally.

So, I was wondering, is it feasible to regenerate this ID on hibernate? It's another instance where the machine powers down, there are ACPI hooks to run scripts on hibernate/wakeup, and I feel at least for a laptop it's a more common use case than a shutdown.

Basically $TITLE.

It happens sometimes that I need to install both the native and the flatpak version of a program, for compatibility or feature reasons. Say for example, native Gnucash has a crash in a specific dialog that I can skip with the flatpak version if that dialog crosses my workflow.

But this causes an issue. Many flatpak applications I have installed, like Norka, result in duplicating the icon / .desktop entry in the menu, which is annoying but survivable, But a few others, such as Luanti, seem to hide or replace the menu icon of the native program, leaving only one entry in the menu that always leads to the flatpak version. And this is quite breaking.

Is there any to prevent Flatpak from hiding or overwriting the native program icons, or at least prevent it from camouflaging itself? Basically I'm looking for how or where to achieve any of the following:

• Make flatpak menu entries install separately, with names or descriptions guaranteed to not conflict with native program icons (say, "App - Flatpak version").
• Make flatpak menu icons install into their own subdirectory in the menu.
• Make flatpak not install desktop menu entries

Thanks in advance.

So I've got a RasPi 3B+ and self hosting a bunch of bare Git repos (git init --bare). And I currently store my dots there. BUT I also have a public mirror on Codeberg. For now I just have two remotes set up and do `git push´ individually for each one. And I don't really envision a case where I start developing it on 2 different machines well. I don't know whether those 2 repos are in sync when I'm developing.

Now, is there a more robust way to keep those 2 mirrors in sync?

I was thinking of adding some kind of hook on the Pi, so that I push my changes to the Pi, and then the Pi pushes the same changes to Codeberg. so instead of my current workflow:

home computer -> Pi

home computer -> Codeberg

this happens:
home computer -> Pi -> Codeberg

Is that the right way of doing this?

Please let me know if you have any suggestions!

Hi, I'm relaying a problem my gf is having. She has an acer laptop with Intel + nvidia hardware and a ThinkPad Thunderbolt dock. Running arch with kde. Whenever she starts bevy engine, including its demos, the dock disconnects and reconnects, causing everything connected to it to also do the same.

The journal has some GL_FRAMEBUFFER_INCOMPLETE_ATTACHMENT errors but I also have some in my journal so I think they aren't specific to bevy.

I don't understand how a user program can cause such an issue. Even if bevy has a bug I don't think it should affect peripherals? Has anyone seen something like this?

cross-posted from: https://piefed.social/c/linux4noobs/p/1953356/my-pc-is-making-a-high-pitched-noise-like-coil-whine-only-under-linux

So I built a new PC a couple weeks ago. My old one broke one day (I think I helped it, but that's another story). I was happy to move from a stock Prism Wraith cooler that was LOUD and ANNOYING. I put a new AMD Ryzen 9 7900X in this box with a couple of new nvme drives (Samsung 9100 PRO and 990 PRO). The thing is silent under Windows 11. Silence at last!

Then I installed openSUSE Tumbleweed (from which I'm typing this) and as the post title says, something is making a high-pitched noise, like coil whine, when the system is mostly idle. I searched the web and the first suggestion is that Linux handles CPU power states (C-state) differently than Windows. Or, it could be the new disk(s), too. It's a fly in the ointment, I am very happy with the new PC, with how powerful and fast it is, I'm so far happy with openSUSE, but there HAD TO be SOMETHING to spoil the experience.

Has anybody had a similar problem? Any tips on how to troubleshoot it and not BREAK my computer?

EDIT: more info from the comments

I just use the Ryzen iGPU, don't have a dedicated GPU. I set the fan curves in BIOS, so it's the same across all OSes. I'm pretty sure it's not the fans. My main suspect is the CPU because the noise is there in openSUSE's installer, so even before anything touched the disks (they were straight from factory with no partitions). As soon as I launched the Tumbleweed installer I heard it. Not hearing it in Windows 11. I can hear it when the CPU is idle, if I start some program, run a compiler or even scroll fast in the web browser, there is no noise.

I had the same monitor for several years. I hear the noise from the PC case and I'm 100% sure about that. I used the same monitor with my previous PC and there was no noise, including in Fedora Workstation. This is a new PC.

The noise is audible in openSUSE's installer, that's the first time I heard it. So even before there was anything on either of those nvme disks, at that time they were straight from factory with no partitions.

I dual boot on this PC and the whine is not there in Windows 11, neither in BIOS. I have fan curves set in BIOS, so it's the same across OSes.

✅ So I managed to troubleshoot and find the root cause. As suspected, it was the CPU, however I thought it was only the difference between how Windows and Linux manage its power profile. While troubleshooting, I ran hwinfo on my Windows install and also double checked the active power profile. It was set to the AMD Ryzen Balanced profile that was added automatically. If we look closer, it was set to always run in high performance mode so it never went into deep sleep and my Ryzen makes the whining sound only when in deep sleep (that would be the C6 state).

I think that Linux sets power efficient modes, even under performance*. The fix for me was not in the OS, but rather in UEFI BIOS. The magic setting for my motherboard (ASUS TUF GAMING B650E Plus WiFi) is at Advanced\AMD CBS\CPU Common Options\Power Supply Idle Control. It had to be changed to "Typical Current Idle". Now I can keep running the balanced profile in openSUSE, so the CPU doesn't operate at top frequency all the time and it is now SILENT, same as under Windows. :)

* at first I tried flipping between powerprofiles set balanced and powerprofiles set performance and for a minute I thought that was it, but not quite. Maybe balanced makes less sound, but still does it.

Hi all, first time posting here.

Installed Fedora KDE Plasma half a year ago on a Thinkpad mostly used in the living room - for streaming / playing videos and minimal gaming. Having zero regrets. It is amazing how far we've come.

I have had very little issues with it, all being only slight annoyances. Nothing compared to win10 or needless to say the current abomination from Microslop... which doesn't want to run on this laptop anyway.

So, here is the first annoyance: when watching a video with VLC, fullscreen, sometimes Discover (the app store / updater thingy) will pop up full screen, says "refreshing..." or whatever, clearly wanting to inform me of some brand new updates.

Question is, how to make it not bother me when a video is being watched?

It is not a big deal, I just immediately close it, rewind a bit and that's it. But if there is some setting to tweak or some magic words to type, I am interested...

PS. I am so happy about this move to Linux. Thank you Microslop for pushing me over.

Heya,

I hope that title isn't too confusing. I recently set up a custom Arch install (Niri, Noctalia, SystemD) and noticed that a lot of WMs/DEs offer to control display brightness with ddcutil. Display control protocols (HDMI-CEC, DPMS, DDC/CI) are somewhat of a topic to me, though my current monitor only speaks ddcci.

I was now wondering why there isn't a daemon, or other centralized way to turn on/off the display when appropriate? Such as boot/shutdown/sleep/resume/power-saving. I can use the D6 feature in ddc to control the power-state. I can even rig a systemd-unit to somewhat do this automatically.

But there doesn't seem to be a comprehensive framework, where I can go "this command is off, this command is on" and it places the calls in the right places. Which is strange. I get that standby exists, but I don't want to leave it in standby for long periods.

Am I missing something?

Thanks.

Hi, I’m trying to get the hang of xinput. For a start I tried setting an option of a device to a value it already has.

I want to go for my Touchpad. From the following output I identified ELAN050A:00 04F3:31B1 Touchpad as the device name.

$ xinput --list
⎡ Virtual core pointer                    	id=2	[master pointer  (3)]
⎜   ↳ Virtual core XTEST pointer              	id=4	[slave  pointer  (2)]
⎜   ↳ ELAN050A:00 04F3:31B1 Mouse             	id=12	[slave  pointer  (2)]
⎜   ↳ ELAN050A:00 04F3:31B1 Touchpad          	id=13	[slave  pointer  (2)]
# more lines

Then I listed the available options:

$ xinput list-props "ELAN050A:00 04F3:31B1 Touchpad"
Device 'ELAN050A:00 04F3:31B1 Touchpad':
	Device Enabled (180):	1
	Coordinate Transformation Matrix (182):	1.000000, 0.000000, 0.000000, 0.000000, 1.000000, 0.000000, 0.000000, 0.000000, 1.000000
# Lots more …
	Synaptics Scrolling Distance (350):	-79, 79
# Lots more…

Then I tried setting the Scrolling Distance:

$ xinput set-prop\
"ELAN050A:00 04F3:31B1 Touchpad"\
"Synaptics Scrolling Distance"\
"-79, 79"
X Error of failed request:  BadMatch (invalid parameter attributes)
 Major opcode of failed request:  131 (XInputExtension)
 Minor opcode of failed request:  57 ()
 Serial number of failed request:  21
 Current serial number in output stream:  22

I also tried without the separating comma in the options list:

$ xinput set-prop\
"ELAN050A:00 04F3:31B1 Touchpad"\
"Synaptics Scrolling Distance"\
"-79 79"
X Error of failed request:  BadMatch (invalid parameter attributes)
  Major opcode of failed request:  131 (XInputExtension)
  Minor opcode of failed request:  57 ()
  Serial number of failed request:  21
  Current serial number in output stream:  22

What’s wrong with the parameter attributes? I’m on Linux Mint 21.3 and Kernel 5.15.0-173-generic.

Edit: Thanks for the help, Yazi seems to be what I'm looking for.

Basically the title. I am looking to use a TUI file manager, but most seem to only support opening files in one external application based on mime-type.

What I need:
Select an image -> file manager shows a preview (unpixelated)
Open image -> it opens in my default image viewer
Open image with... -> file manager lets me select a program to open it

I need this cause I often want to open images in gimp, but I don't want it to be my default image viewer.

if i have libinput-gestures autostart normally then it conflicts with the compositor and then the compositer wont start. how can i make libinput wait 10 seconds?

using i3wm on fedora 43, decided to give beterlockscreen a try and it works great BUT, on inital login it shows the desktop then immediately locks the screen. effectively making me login twice. any suggestions?

Ever since SDDM was dropped (though it could just be coincidence) in favor of plasma-login-manager on CachyOS when using KDE, Steam scaling breaks whenever my display turns off and on again. Not sure why.

It's fine starting Steam and it'll display correctly 200% as set in KDE settings for my 4K TV.

When we turn off the TV we often leave the PC on. When we turn the TV back on everything looks fine except Steam. Even if it's closed to the system tray. It'll be the only application wrongly using 100% scaling for some reason. edit: forgot to mention restarting Steam is annoying but restores 200% scaling until display turn off again.

Radeon 9070XT since display driver may matter?

Power Management too idk: https://media.piefed.social/posts/yb/bM/ybbM1Ks63dLQj2l.png

All of my important data is on btrfs drives. I intend to install my system on ZFS. Why, you may ask? Because I can. That's the fun of Linux after all. I intend to mount btrfs drives as well. I hear that ZFS can break fairly easily? Is this a bad idea?

Edit: I understand ZFS is out of tree but CachyOS maintains their own package and dkms so it shouldn't matter I would think?

I have plank start on lxqt desktop but i don't wan't it to start in my i3 desktop. For some reason it pushes the window up to the top half of the screen. So when i start i3, first thing i do is close plank. Is there a way to make it show on one DE vs the other?

I need to mount ISO files on launch of some really old games. I see I can create a script and then have lutris launch the script to mount and launch another script to unmount when I run the game however it requires root privileges to mount. I’ve tried adding the specific ISO files to fstab And changing the mount location to something in the users’s home folder, but it’s just not working. What am I missing?

The icon in the app selector and taskbar thing are both the proper chrome logo, but when I open the window it shows like this.

This is on KDE, on fedora.

(It's ungoogled chrome, don't suggest installing firefox to me.)

Hey,
So after hardening a system with brace and rebooting, the wifi interface is down and I struggle to get it back up.

Tried ip link set [wifi interface] up
But no success

I'm also not very good at wifi issues.

Maybe a module got deacticated? I dont know.
The script also says its for debian 12 not 13 and only really fedora is supported, but heyy.

Would be grateful if anybody could help.

Thank you

Often when downloading a game on steam my nvme drive gets pinned at 100% IO usage and my whole DE (gnome) can lock up at times. Is there a way to limit it so that this never happens?