Is high resolution scrolling enabled by default now?
Without doing fuck all (cachyOS) I noticed yesterday after a reboot that my mouse had a different name under the settings, "Logitech G502 X LS", enabled back mouse acceleration and the mouse wheel started behaving weird. Or more like an analog wheel. I found out it was high resolution scrolling and I disabled that shit, but what the hell?

>>108144972
I dunno what you're trying to explain about how your mouse behaves.

Why did no one tell me Debian was so comfy

Linux is unusable because of rust.

>>108144984
The G502X (and a bunch of other mice) have wheels you can unlock to use infinite scrolling.
Normally you expect a step to just be a step on a mouse wheel but these apparently can behave differently. On linux there's this high resolution scrolling method which turns something more digital (step by step) into something smoother which works decent for an unlocked mousewheel, but when you "lock it" so it behaves normal, it's all sorts of erratic.
So what will happen for instance, is that if a program expects one step, it can get interpreted as 4 and it's a mess to use. It's a shitty tradeoff. I don't use infinite scrolling that much, I want a step to be a step. But by default, it never behaved like this. Since yesterday it did, and I found out this is not a new thing somehow, and I had to apply a fix to disable high resolution scrolling, which is present on a few other logitech mice as well.
If you have a normal mouse I'd like to assume this is never an issue. But I didn't even know this mouse could do this gimmick and honestly I don't want it. It's like switching digital to analog, but you're expecting accuracy.

>>108145041
I have a G502X and the switch between infinite scrolling and step by step is on the mouse itself, I've never even heard about high resolution scrolling.

>>108145138
That's because you might not have it enabled, which is what I thought was the default myself.
Again these are two separate things. When you unlock the mousewheel it still works by steps for you. If you enable high resolution scrolling with the wheel unlocked, those steps are much smaller giving you a more analog feel where you can actually make very small scrolling adjustments. It's meant to be "smoother". However, lock the wheel on that mode and suddenly you're not scrolling step by step anymore.

>>108145012
it doesn't last

>>108145012
It was real comfy until I ran into a driver bug that got fixed in Nvidia driver 555 apparently (Debian is on 550). I would use it still otherwise.

How come so many people think distros have a single set DE? I see this misconception every day. Someone literally said the reason they switched from Mint to Xubuntu was because they don't like Cinnamon but the download section literally has Mint Xfce on it lmao. Don't even have to install it afterwards.

>>108145270
There is a misconception there indeed, but you can have a distro where a DE isn't particularly well maintained. And there's always a preference per distro. Mint offers an XFCE flavor, but Cinnamon is the one they want you to use and contribute to. It is their thing after all.
Also the situation wouldn't be so confusing if something that your display is capable to do wasn't actually tied to this DE choice. You cannot have HDR in Cinnamon, for instance. This should not be a thing. Having the choice to make panels look different shouldn't have anything to do with what your display is capable of doing, but here we are.

>>108145297
Having to choose between well supported archaic tech and WIP modern tech is always a headache but wcyd. I think even Xfce now supports Wayland but it's very much in early stages

>>108145324
They're getting there but at the end of the day, this wayland mess changed way too many things. One day, I'd like to enjoy something that is not Plasma without restricting, cucking my monitor.

>>108145297
>HDR
snakeoil

>>108145365
Not everyone bought a poorfag FakeDR monitor like you.

>>108145213
Ok, I think I understand what's your issue is with it now, but still don't know where I'd even enable or disable this.

>>108145401
There's no GUI setting for it, you need to fuck with libinput quirks.
I don't know how it got enabled, but I disabled it by creating a file in /etc/libinput/local-overrides.quirks with the following contents:

[Logitech Hi-Res Fix]
MatchName=*Logitech*
AttrEventCode=-REL_WHEEL_HI_RES;-REL_HWHEEL_HI_RES;

I didn't have any idea about this either, it just randomly fucking enabled itself.

>>108145445
dunno if that works for wayland, probably not.
Anyway, apparently lininput has a new option HighResolutionWheelScrolling which is enabled by default. You might just need to disable that.

File: 1747788164484124.jpg (283 KB, 1125x1476)

283 KB JPG

the only book you'll ever need

>>108145516
It does, I only use wayland.

>>108145557
So, if I update my system my mouse might shit itself, interesting.

>>108145574
Hey maybe for some weird random reason it got enabled and it was meant to be by default, fuck if I know. Thing is I got that issue.
Maybe there's a simpler way through that new option but I don't know how I would do this, and I already applied the fix anyways.

>>108145592
Dealing with your mouse on Linux is a good example of why the desktop experience is still not ready for the average person. A lot of stuff can only be changed by modifying libinput files in a text editor and sometimes it's hard to get the info on how to do what you want.
Got a new mouse and it's detected THREE times in the Mouse settings of KDE and one of the three entries allows me to set both speed and acceleration in the GUI but changing the values does nothing while another one of the three only allows me to set speed and that setting is actually reflected. Still haven't bothered to start properly dealing with it. I already did try only enabling the entry with the best available settings but that just killed my mouse.

>>108145640
mouse vendors just expect you to run their stupid software at all times eating up like 200MB of RAM. People actually do that.

>>108145640
Mouse and mousewheel is one of the most annoying aspects of Linux. It's incredible that in 2026 - there isn't a simple way to customize the scroll wheel sensitivity. Jesus Christ it has been 30+ years already.

>Debian stable released more borked kernel updates than any rolling release distro in the last 3 years, by the way.
Debian sisters...

>>108145640
You can blame the manufacturers for that. Why is there no Logitech or Razer, etc, GUI for Linux? In Razer's case there's plenty of excellent third-party drivers and configuration utilities but this all had to be reverse engineered. The manufacturers contribute nothing despite the fact that Linux is now more popular than ever, especially for gaming and it's only growing.

>>108145676
>>108145666
You guys are correct, although in most cases on Windows you don't need the shitty proprietary mouse setting software that comes with 30 000 node packages to be running except when you change the settings, the driver will remember them after you close the GUI shit.

>>108145640
>>108145667
>buy broken hardware
>it doesn't work for shit
AMAZING

File: ynqlwv[1].png (219 KB, 1390x900)

219 KB PNG

>>108145698
>>108145698
Okay it was not 30k but I was not exaggerating by that much. All this just for the GUI in the image.

>>108145669
How?

>>108145820
Not him but Debian does an awful lot of backporting, their kernels don't even resemble an upstream kernel from kernel.org since they backport so much fixes to it and stick with the same kernel version for the entire lifetime of the OS.

They probably just broke a bunch of stuff countless times whilst doing that where as upstream regressions happen on the stable kernel series but are rare and fixed quickly.

Basically, in their pursuit of a stable kernel API/ABI they probably broke a bunch of other stuff and made things unstable because Debian stability does not mean bug free / problem free, it's simply talking about not breaking the interfaces by sticking with the same version of software for the entire lifetime of the OS.

>>108145666
Logitech in this case has this thing on windows called the onboard memory manager. You open that shit up, change settings, done for fucking ever. Delete the application, never to see it again, and the mouse works the same today as it does 3 years from now if onboard memory doesn't somehow malfunction. Because as much as microshit stagnates, they don't often change how input works either when it comes to something as simple as mice. It's predictable.

>>108145935
Debian just uses LTS branches of the kernel, the fuck are you even talking about?

>>108145988
Yes, that's where the problems come from because they have to cherry pick fixes from stable that were never even tested on this kernel.

>>108145999
The cherry-picks are done by upstream, not debian.

>>108146017
No, Debian does them. You can check the source of their kernel package yourself:
https://salsa.debian.org/kernel-team/linux/-/tree/debian/6.12/trixie/debian/patches/bugfix?ref_type=heads

Then there's all these "Debian" patches for doing "Debian" things because they didn't like what upstream did for whatever reason. This is very typical of Debian, lots of software gets "Debianised" like this, applying patches that upstream may not even be aware of. Some developers flat out won't even support Debian because of this.
https://salsa.debian.org/kernel-team/linux/-/tree/debian/latest/debian/patches/debian

>install Debian
>random peripherals like usb wifi dongle, m-audio air, webcam just work
>install GPU drivers
>games work, significantly better than Fedora, and Mint
>install all my normie apps
>just works
>install my dev shit
>just works

What's the point in Mint and Ubuntu in 2026?

>>108146031
Debian doesn't customize anything and mostly keeps upstream defaults so if the defaults look and act shit then it'll be shit on debian while ubungo and mint will customize stuff somewhat.

>>108145669
Yes there was a Debian 11 kernel update recently that finally allowed me to use the ttys other than 1 again. Lots of regressions with updating old stuff. Ridiculous.

>>108145935
Which distro am i supposed to use for a server then? Nothing redhat related.
Truenas scale, proxmox, some other server distros all use debian as a base.

>>108146074
Debian is still stable enough despite the retarded kernel so Debian.

>>108146074
>but all this gay shit from reddit shit uses debian
Exactly. RedHat.

>>108146074
Don't know, I use Debian but I barely update the kernel.
>>108146079
Yep, Debian will be tarded but it's still the best thing if you want a stable system (and that doesn't mean bug-free, just that it will have predictable well-known bugs and issues that are easy to workaround if you have to)

>>108146030
When the kernel gets updated from 6.12.48 to 6.12.49 that's a change done by upstream. Not a backport done by debian.
Debian having their own patches on top of upstream is a different beast altogether, as those should also not be backports.
It's also not a debian thing, most distros do this shit.

>>108146041
Idk looks pretty clean to me. Maybe that's just how Debian used to be? I read that the only semi-recently switched to two year release cycles, so prior to that it could be 2-6 years or something crazy, so it could just be fighting against its own reputation. Or maybe it's placebo because it's more "pure" than downstream distros. I had been avoiding Debian while hopping because "old" and the website is so dogshit ugly it made me think of early late 00s Linux - ugly and jank.

>>108146164
They wouldn't have any "bugfix" patches at all if upstream took care of all of that. The only thing upstream does is backporting some trivial patches that merge cleanly if (and only IF) one of the kernel developers tagged their commit with "For stable".

>>108146203
>I had been avoiding Debian while hopping because "old"
It all depends on things like hardware support.
If in the last 2 years you haven't been looking at new features, support for new devices or whatever, it's not such a bad idea. Problem is the misconception about "stable" around Debian and how it's not really all that amazingly secure because your shit is outdated to hell and back.
Also a lot of people kinda miss the point by using unstable or just filling the system with flatpaks and distrobox to make up for the lack of up to date software. You can do that practically anywhere at that point.

>>108146211
Also sometimes they backport patches that should never have been backported in the first place because they depend on something else that's not in a stable / longterm kernel.
Yes, even the kernel developers often have no idea if something should be backported or not.

>>108146250
I do hate how everything I read about it is just
>ROCK SOLID STABILITY
Just like every other meme catch phrase in tech I assumed it was BS. From the little reading I did it just means the versioning doesn't change between point releases.
>Also a lot of people kinda miss the point
Yeah but if you're using "unstable" (up to date) do you even need the flatpaks and distroboxes?

I'm kind of doing literally what you just said because it sounded good on paper to have less dependencies intermixing, and using containers for dev just so I can have versioning control over the entire stack more simply. Though I'm doing that more so because I found out about that practice from atomic distros and liked the idea, but hated the idea of atomics because flatpaks + boxes seemed like it solved the problem atomics were trying to solve anyway.

>>108146085
Not interested in having to deal with the eventual third episode of redhat trying to kill rhel-based distros.

>>108146348
flatpaks and distroboxes are useful even on arch if you want "stability" (forma de debian)
not everything has to update on the same schedule, but as soon as you want to mix you need containers, it's all glibc's fault

>>108146396
The wonderful thing about RHEL knockoffs being cross-compatible is it doesn't actually matter. The only people who think this is an issue don't use RedHat.

>>108146074
Ubuntu LTS with the GA kernel?

File: 1767860411399690.jpg (904 KB, 1000x1010)

904 KB JPG

How to deal with constant python version upgrade on rolling distro like arch/manjaro/endeavor?
Because I use pip to install some applications like gdown, etc.
Should I remove the old packages in ~/.local/lib/?
>inb4 use pipx
I know I should but I've written some scripts that automate the upgrade and re-installing of those when a new version is installed, and I didn't realize how bad this would get so now I have 3 GB of old packages that goes all the way to python 3.9

File: 1763931309303735.jpg (177 KB, 1024x1024)

177 KB JPG

Can someone explain why I'm having this weird issue??
When I'm connected to wifi, sometimes my connection drop because I either go out of range, or the AP get restarted.
My system doesn't re-connect back in about 33% of the time and demand to re-insert the password or think the password is incorrect.
How to solve this?
Because simply closing the dialog and re-connecting manually works just fine.

>>108146909
You can just rm -rf the entire python3.x folder for the previous version.

>>108146939
Although there may also be some symlinks in ~/.local/bin, etc.

I had to install firefox-esr after having some problems with firefox. The problem is that AAC codec seems to be unsupported on firefox-esr and I can't watch some videos or listen to some radios I want to. Is there any way to enable AAC support?

>>108146742
What's a GA kernel?
>>108146348
Stability is a Linux jargon, it means any package versioned 1.0 won't ever get upgraded to 1.1.
>>108146164
>>108146030
You can upgrade the kernel all you like, it's not like the userland cared.

>>108146993
>What's a GA kernel?
Don't you have the Information Superhighway at your disposal? It's the Ubuntu LTS kernel that maintains ABI compatibility, unlike the default HWE kernel that breaks it.

>>108147092
What's a HWE kernel?

Why do people say troubleshooting in distros like opensuse and fedora is simpler than the meme distros?
I've been using cachyOS and nearly every single problem I had I could quickly find either documented in the arch wiki or by quickly googling it. Not finding so much stuff specially for openSUSE, it even looks fucking barren over there.

>>108147251
Because meme Arch distros as basically just Arch + some configs. Like isn't Cachy just Arch plus drivers and a DE preconfiged? At that point you're just using Arch, which has an amazing wiki, and any problem you have on it is just an Arch problem... Now if you're using Bazzite or something, you're dealing with a meme within a meme within a meme within Fedora (Fedora -> Fedora Atomic -> Universal Blue -> Bazzite) and you're more on the fucked in the ass end of the spectrum.

Suggestions for XMPP clients? I'm using Pidgin but I wonder if I'm missing out on anything.
This is unrelated to Discord's bullshit, I use XMPP for my autistic internet spaceship guild

>>108147329
it's Arch with a custom kernel and package repo supposedly optimized for your CPU. I've been using it for the last two weeks and struggle to notice any difference though, to be honest. it actually already broke shit in the first few days so I'd honestly just stick with Arch anyways.

honestly very much enjoying niri

>>108147795
i am just running the cachy kernel and with 6.19 they managed to nuke my LAN driver. Now i have to remember to always pick the default arch kernel on boot until they bother to put a new release.

Which file explorer is better at handling folders with thousands of large images?

I don't know about large images but dolphin is the fastest at loading my shitposting folder

>>108148360
meant for >>108148204

>>108145012
Removing Evolution removes Bluetooth support completely.

>>108144945
I got this solved btw
>>108142404

You can run AppImage GUIs with limited filesystem access using AppArmor on Ubuntu.

I don't have to install any libs or packages, or fuck around with users and groups and their permissions. Just whitelist what needs access and execute the program. Any program. GUI or other.

AppArmor also comes with a nice way to find DENIED permissions.

I have more testing with firejail to do. Maybe I can get it to work.

The program is Cursor btw. Obviously it's a black box that should be restricted as much as possible.

>>108145935
>Debian does an awful lot of backporting
You're making shit up, it is against Debian policy to backport anything.
They absolutely don't backport anything kernel related, ever.