I had a case where fingerprint sensor was working out of the box fortunately. Although I had a problem where cryptfs would stop authenticating successfully with fingerprint sensor after distro update
Absolutely not outdated. I had a horrible time getting my hands on a working driver for the WiFi card in my brand new laptop last year. Horrible enough to resort to Ubuntu and even that gave me the finger. When I finally had it working I had to manually rebuild the damned thing each kernel update because I couldn’t convince DKMS to do it automatically. Had to wait two or three kernel releases for the card to be supported ‘out of the box’.
So no, fuck WiFI drivers in Linux. If it is not in the kernel and the manufacturer doesn’t provide one, don’t expect fun times.
Outdated for Linux Intel, still valid for Broadcom, probably not so bad for somewhat recent Realtek and AMD/Mediatek (last I’ve read is that Mediatek WiFi hardware sucks in general and disconnects happen on Windows, so the same happening on Linux would be the fault of the Linux driver).
I installed linux on a new pc 2 days ago, had no problem with the wifi drivers. I don’t know if it’s the fact that the wifi is integrated on the motherboard, but it was up and running without any tweeking from me (unlike windows)
But was the cause the Linux driver or the hardware? If the fault is the hardware and the experience on Linux is the same as on Windows, it’s feature parity.
If in doubt, get an Intel WiFi card. Even in otherwise not upgradeable notebooks those are usually not soldered on. Also whatever is in a Steam Deck OLED looks like a good pick.
Does Intel sell wifi cards that use USB rather than PCI slots? My motherboard doesn’t have the slot for a wifi PCIe card, and I’ve only seen Intel sell those :/
I can absolutely confirm it’s still valid for Realtek. I had one using the RTL8812AU chipset that basically no kernel version nor distro provided out of the box, so I constantly had to download a third-party driver from Github and manually patch it via dkms, or use a third-party repository containing the driver package… and then the driver broke so badly that it wouldn’t let me update at all unless I uninstalled it, which left me without the internet I needed to actually update, effectively leaving me unable to update until I could buy another one from Mediatek that’s compatible.
And said Mediatek wifi is really slow, so I just went from the frying pan into the fire…
I can absolutely confirm it’s still valid for Realtek. I had one using the RTL8812AU chipset
Yeah, and I was explicitly writing about recent chips. RTL8812AU isn’t recent. The very latest Windows driver is from 2018, so the chip itself was released a good while before that.
I know exactly what you had to go through because I had to do the same with mine a couple of years ago but since then for newer chips Realtek started contributing to Linux itself:
Ahh I see, thanks for clarifying. It seems that where I live mostly only has the older Realtek chips for sale, so I likely mostly had bad luck.
I tried USB tethering, but it wouldn’t work for some reason… I don’t remember exactly what happened, but I think either the phone or my computer couldn’t detect each other.
I do occasionally fall for just buying shtuff without a quick google search to see if my kernel would be cool with it, but I have an even greater number of stories about good experiences with Windows shtuff driving me bonkers.
For example, the Brother ADS-1200 under WIndows beats anything SANE supported scanners can do hands down. Scan to PDF with excellent compression and top of the line OCR. The spousal unit needed a scanner and I found a good deal on an ADS-2100. Under Linux, scan results are totally comparable to the ADS-1200, so the hardware is fine. But the Windows software for this scanner is crap. JPEG and TIFF are identical to the Linux scans, but OCR and PDF compression are atrocious. I’m 100% sure that if I were to edit a table in the ADS-1200 software, it would happily apply the same excellent results to the ADS-2100. But I’ve had it with hacking Windows goop, been there, done that, got the t-shirt, so onto Craig’s list the 2100 goes… Built in obsolescence, welcome to the Windows world.
With Linux, once the kernel accepts it, it’s smooth sailign without too many vendor introduced hickups.
And even on Windows, if you need to use third party scan software like VueScan because your scanner happens to be older than your Windows. it’ll work but it won’t outperform SANE supported scanners.
Situations like that aren’t very common these days. It usually happens when your hardware is very much new and drivers aren’t yet in the Linux kernel, or they are in the newest mainline, but your distro wont ship it for some more time. For that matter, it’s always bad when the kernel doesn’t have the drivers built in and it always requires dealing with DKMS or akmod whether it’s wifi, webcam, bluetooth or GPU (that’s why NVIDIA tends to be problematic on some systems).
That being said, the meme only works for anecdotal cases.
If it is not in the kernel and the manufacturer doesn’t provide one, don’t expect fun times.
This could be shorted to if your device has no driver it wont work which is obviously true.
If you have very recent hardware and you find it doesn’t work out of the box on stable options the easiest thing to do is install a more recent kernel. Even current Ubuntu non-LTS is 2-4 releases behind.
It’s even easier in arch/void where the latest kernel is already available.
Respectfully if DKMS wasn’t automatically kicking in then you configured it incorrectly. It’s a lot easier to just rely on a package that sets this up for you properly. If for some reason this can’t be done the logical thing to do is script the process so that all operations are completed in the appropriate order that way you needn’t remember to do one then the other.
This could be shorted to if your device has no driver it wont work which is obviously true.
What I tried to tell is that if you have to rely on community driver projects, don’t expect fun times, at least not when it comes to Realtek in my recent experience.
If you have very recent hardware and you find it doesn’t work out of the box on stable options the easiest thing to do is install a more recent kernel.
I already had the latest available kernel at the time, as in: the very latest officially released kernel by kernel.org. Ubuntu was just a last-ditch effort as it will sometimes have drivers included that other distros might not have, normally I wouldn’t touch it with a ten-feet pole and go either Arch or Manjaro. The driver simply wasn’t included in the kernel. How do I know? Because I stumbled upon some discussions that mentioned the lack of support and 3 kernel releases later support for my card was specifically mentioned in the changelog.
Respectfully if DKMS wasn’t automatically kicking in then you configured it incorrectly. It’s a lot easier to just rely on a package that sets this up for you properly.
Yes, like a Realtek-XXXX-dkms package, which simply didn’t work. I’ve configured stuff for DKMS before, scripting stuff for Linux is part of my daily workload, so yeah, you don’t need to tell me scripting beats doing stuff manually.
The fact that getting an f*cking wifi card to work takes this much effort is what I meant with ‘not fun times’ and for me validates the meme, anecdotal as it might be.
Resorting to other distros, configuring additional repos so you can install a different kernel version, having to try different community projects to see which gives you a working driver, having to deal with getting DKMS to work, this is all stuff which hampers Linux adoptment. And without more adoptment we won’t have to expect more support from manufacturers for desktop related consumer hardware. So yeah, that does make me cry a bit. It’s a catch-22 unfortunately.
The difference between paranoia and fear is the difference between not wanting to buy a Google Home because it listens to you and not wanting to buy a Google Home because you’re afraid you’ll break it.
That is actually a great metaphore. I always just used:
It’s like me not wanting to use google photos because they scan your photos to train algorithms vs my mom not wanting to use google photos because she is afraid all of her photos will get deleted.
You know how I know I've gotten better at using linux?
I saw the command and read it and figured out what it was although I've never been exposed to a fork bomb before in my life.
I was like okay, this is an empty function that calls itself and then pipes itself back into itself? What the hell is going on?
I will say that whoever invented this is definitely getting fucked by roko's basilisk, though. The minute they thought of this it was too late for them.
99.999% of that function’s effectiveness is that unix shell, being the ancient dinosaur it is, not just allows : as a function name but also uses the exact same declaration syntax for symbol and alphanumeric functions:
Well, openSUSE did it long before everyone else. So, Debian, Fedora, Arch?
I would kind of be surprised by Fedora, too, as I thought, they shipped out-of-the-box automatic snapshotting, but the comment from @bruhduh sounds like that is still a problem…
OpenSUSE does this as default, which is laudable. Mint will only use Btrfs if you manually tell it to, it just handles it gracefully once you do choose to use it.
Yeah i was surprised as well) thought automatic btrfs partitioning by fedora gui installer would suffice, but it’s not, it did not had subvolumes set after installation, so timeshift btrfs didn’t worked, after i set subvolumes timeshift started working, but after update from 38 to 39 everything broke and locked up my ssd
I doubt it. It’s the halting problem. There are perfectly legitimate uses for similar things that you can’t detect if it’ll halt or not prior to running it. Maybe they’d patch it to avoid this specific string, but you’d just have to make something that looks like it could do something but never halts.
They could always do what Android does and give you a prompt to force close an app that hangs for too long, or have a default subprocess limit and an optional whitelist of programs that can have as many subprocesses as they want.
The thing about fork bombs that it’s not particular process which takes up all the resources, they’re all doing nothing in a minimal amount of space. You could say “ok this group of processes is using a lot of resources” and kill it but then you’re probably going to take down the whole user session as the starting point is not trivial to establish. Though I guess you could just kill all shells connected to the fork morass, won’t fix the general case but it’s a start. OTOH I don’t think kernel devs are keen on special-case solutions.
You don’t really have to kill every process, limiting spawning of new usermode processes after a limit has been reached should be enough, combine that with a warning and always reserving enouh resources for the kernel and critically important processes to remain working and the user should have all the tools needed to find what is causing the issue and kill the responsible processes
While nobody really cares enough to fix these kinds of problems for your basic home computer, I think this problem is mostly solved for cloud/virtualization providers
That’s why I run all my terminal commands through ChatGPT to verify they aren’t some sort of fork bomb. My system is unusably slow, but it’s AI protected, futuristic, and super practical.
i’m gonna get crucified for giving apple a single benefit of a doubt but i think there are just as many windows users who “fear technology” as mac ones. think of all the grandparents running shitty dollar store pcs. mac is only a walled sandbox until you turn off the safeguards, then you can see exactly as much dumb back-end shit as you can on windows
I think this is sdvice on what you should do, not what people actually do. This would be why there is such a big industry for windows tech support. Tldr: Windows: Be afraid, very afraid.
Well, the other side would be operating systems you can’t really screw up too badly because they are locked down harder, so perhaps it’s fear of the unknown?
Or in the office, the hardware-software relations between the laptop and Windows and in some parts Linux are strained at best, where drivers, power management, and so on get crappy. E.g. after a year or two of updates, it gets out of control and nice things like hibernations don’t work. It’s usually a driver for some small thing you don’t care about that forgot to read the Windows specification change and now it can’t do that power handling in a good way. Oops the computer refuses to sleep and your bag is burning, your battery is 1% when picking the computer up again.
I completely understand that with windows, especially with hibernation like what the fuck is “windows modern standby”
but with Linux, it depends on the distro you use.
if you’re using something such as Pop_OS, I can pretty much guarantee you you’re never going to run into a power management issue or even a driver issue for that matter since its based off of Ubuntu and is very well supported.
That’s a lot of money, but same sentiment in the opposite. I would avoid any dev job requiring me to use Windows. Chances are they’re also using some crap tech stack too.
I’ve seen plenty shit stacks on macos tbf. Windows has better window management which saves a lot of time when you’re juggling between seperate windows.
I’m not sure, many developers use mac to get working unix tools and working “enterprise” tools at work like Teams and other crap that the company uses for “everyone”. Sadly many of these tools work like crap on Linux and maybe in best case the web-version is workable.
You’re confusing developers with power users here. At my company, the developers can do one thing well, but are far, far from power users with any technology. The amount of times I’ve seen them get stuck at a simple error message without doing more than throwing their hands up thinking they don’t have permissions or something is actually broken, without doing the least bit of troubleshooting is both baffling and frustrating.
linuxmemes
Oldest
This magazine is from a federated server and may be incomplete. Browse more on the original instance.