Note that Perseverance does not run on Linux, the drone (Ingenuity) does. I know Perseverance has sent audio recordings. But is Ingenuity doing some audio processing?
> NASA settled on the Linux OS for the rover’s helicopter-like drone, Ingenuity [1]
> Ingenuity doesn't carry science instruments and is a separate experiment from the Mars 2020 Perseverance rover. [2]
The drone doesn't run on Linux either. It has several CPUs (like most spacecraft), most of them run hard realtime systems, and only the COTS board used for the additional hazard avoidance (IIRC) runs Linux; it's not required for the helicopter to function. The whole point was to test non-radhard electronics on Mars without endangering the rover.
What do logicians call this? A null positive? Is that allowed? E.g. obviously it's correct that "no Linux systems on Jupiter do not have working audio", but is "all X do/have Y" a valid syntax for zero X?
Not sure it's related, this set of statements is not about statements containing themselves or the set itself. You need a set of sets for the Russell paradox to have chances to kick in.
I've been using Linux as desktop OS for 14 years this month, and I've never had issues with audio on none of the 6 computers I installed Linux on.
Getting a new printer working (or a fresh install working with an existing printer that used to work flawlessly), has been a recurrent source of frustration, and I've experienced occasional issues with GPU drivers, but audio has never been an issue, so I wonder: are “sound issues on Linux” a meme that survived from the early days of Pulse Audio, or is that still a thing in 2023?
Yes, unfortunately. Since pipewire was introduced (replacing pulseaudio) sound always "judders" when playing or unpausing a video in VLC. It's a slight annoyance so I've not bothered to look into it. Pipewire also crashes after a while (seems to have a memory leak) which causes any programs to lose sound, - that "only" requires you to restart those programs.
Linux Mint 21.1 Vera. I get stutters in VLC playback without pipewire - pulseaudio is to blame, Mint used to work better. Thanks for the heads up, I will not try fix what is broken with something that has known issues.
Little things - on mine, the volume level after resume is being set back to the default level, not the last-set level. I suspect that's because it's a USB audio device though.
> sof-firmware is required for some newer laptop models (mainly since 2019) because they implement their drivers with firmware provided by the Sound Open Firmware project. Checking the journal will provide messages about the missing firmware (see BBS#275577).
This is perfectly legitimate, but I had to scan through `dmesg` and find a recommendation in a hyperlink. Not something you'd ask your grandmother to do.
Only for some people. Lots of people are saying they are having issues with pipewire, but personally I have never had any audio issues with pulseaudio nor with pipewire. And that is on multiple devices across many years.
So no, it's not Linux. Because if it works for at least 1 person, that means there isn't some sort of technical limitation where Linux can't get audio to work properly, since clearly it can.
Maybe the distro they use doesn't package the audio packages properly, or maybe they messed around with too many settings, or maybe their specific hardware has issues. But none of that is Linux's fault.
Does it really matter whose fault it is? When the average user tries to switch from Windows to some flavor of Linux and their audio doesn't work, they don't give two shits whether it's Linux-the-kernel, the package maintainers, or the audio subsystem at fault, they're rightfully going to blame Linux-the-ecosystem.
Strange. What distro are you using? I've used that same stack for a year+ and things have been just peachy. I'm playing Steam games and doing light studio work (Ardour, etc). I'm actually surprised by how Pipewire just works. (Fedora Kinoite)
It's on Arch Linux. I haven't really figured out what's going on yet, the sound dies erratically but usually when eg. a Chrome audio tab is closed or something.
Younger me would debug it more. Older me just made a script to turn the sink off and on quickly until someone else fixes it ...
Only if you an incompetent, bungling fool who refuses to just do things in a sane way, and use a distribution designed for the purpose, such as Ubuntu Studio.
I've had a pro-audio Linux-based DAW in my studio for 15 years and it has easily been the most stable, rock-solid performing DAW - alongside both Mac and Windows systems - by a long shot. The reason is, I chose my hardware and my distribution accordingly - Presonus AD/DA convertors, and Ubuntu Studio - which is such an amazing, powerful, mind-blowingly wonderful resource, I feel inclined to not mention it, lest it get wide adoption and suffer the long-September factor when everyone and their dog realizes just how great it is.
Seriously, choose your distribution wisely and you will have better audio performance than the other OS'es. This is a fact.
I haven’t had problems, but my friend was. The problem was solved by updating his graphics drivers, as he was running everything through his Invidia GPU’s DisplayPort.
That didn’t solve Cinnamon’s refusal to use Steam, however. I have had 0 problems in XFCE, outside of self-made ones like crashing my PulseAudio trying to modify the config.
automatic switching from A2DP (headphones) to HSP/HFP (handsfree) doesn't work most of the time [1]. the drivers for BT or other audio cards are bad as hell, so everything else isn't working as it should.
I’ve had pretty good luck myself, but have on occasion gotten wrapped around the axle with some layer in the middle suddenly having whacked out mixer settings or something.
I’ve always been able to Google a solution so far.
I am Deaf. I never bothered to setup audio (Arch Linux on Framework).
Edit: I would need someone to help me verify that sound works. Of course I could touch the chassis near the loudspeaker and play a 200 Hz sine wave on https://onlinetonegenerator.com/ but why? I don't hear so why? My children leave my laptop alone because they know that there's no sound.
This quote is old. Why would he repost this now? Did he try Linux again? Or it's just for the pleasure of trolling? (which is fine, not saying, just wondering if there's a context to this "reminder")
The quote is also wrong, any planet without a Linux system satisfies the predicate "100% of Linux systems have working audio", and they are many of them.
I'm not speaking about this. I'm wondering what pushed jwz to repost his quote.
I have my own stories of sound that don't work 100% fine on Linux. I'm sure it's true of all OSes, try searching "Windows audio issues" on any mainstream search engine for some evidence of this.
This is one of those things where math doesn’t apply to reality (applied maths vs. theoretical maths). If you have zero of something, you just have zero. You can’t have infinite something just because you have zero of them.
It's not a case of math not applying to reality. It's a case of not using the right tool/model.
There's no meaningful (both theoretically and practically) answer to the question "What proportion '0 out of 0' represents?". You can't answer 0%, you can't answer 100%, nothing in between neither. There can be a proportion only if there are a non-zero amount of things you want to count.
"0/0" is not the right model. It never is. "100% of 0", however, means (100/100) × 0, which is well defined and equals 0. The quote says "100% of [something evaluating to 0 in our case]".
Note that I didn't mention infinity. I mentioned 0/0 being undefined. There's nothing about 0/0 that's infinity. The limit of c/x when x goes to 0 is equal to infinity, c being a constant different from 0 (+infinity if x is positive, -infinity if x is negative). lim 0/x is 0 when x -> 0 because 0 over anything non-zero, however small, is 0. The limit is yet another tool that's not really related to what we are discussing unless I'm missing something.
The reason it is undefined is because 0/0 is infinity or 0 (you can prove it either way). Thus there is not a 'definite' result. However, in practicality, an entity cannot have infinite of something, so the result must be 0 even though it is theoretically 'undefined.'
This is not a question of proof, this is a question of definition. x/y is not defined for y=0.
0/0 is not infinity nor 0. It is not, full stop. At most you can decide by convention that 0/0 should be treated like infinity or 0 in your particular cases.
(though I'm interested if you have pointers about this)
In math, when something is "undefined" it doesn't mean "it is not, full stop." it means that there are multiple answers or no answers with whatever rules we are working with. For example, when you can only work in the field of Real Numbers, the square root of a negative number is undefined (though it is an imaginary number).
In our case, we're saying 0/0 planets (caveat: other than Earth and Mars) have a Linux computer on them; then continuing to say that is 100%. This is clearly not true in any common sense term, since there is not a computer on every planet in the universe. Thus 0/0 == 0% in this case. In real life, the laws of mathematics apply differently than theory. In real life, there is no such thing Zeno's Paradox because eventually you'll reach Planck Scale and then it just depends on how you round, whether you reach the end or not.
Personal anecdote: Windows Update has a habit of resetting microphone frequency settings, causing problems if Windows and connected professional hardware suddenly speak "different languages". This is strictly an OS-level issue, since no driver has been changed and the hw config and behaviour has not been changed either.
It's not just you, you will find most if not all vendors of external audio devices for Windows will have a tech support page saying you have to hit 'reset defaults' after Windows Update if it chooses to change Sample Rate, default device etc.
Windows often changes the settings for internal Realtek devices, too. Windows Update can set unusable settings irrespective of whether vendor drivers are used or class compliant Microsoft supplied ones.
My Dell laptop with Windows didn't recognize any headphones or headsets, until I tried 3 different drivers from 3 different vendor websites. Applications mess with volume settings constantly unless you know which checkbox to uncheck, at work we have a speakerphone where with some laptops you can't understand anything at all, and so on. Windows is not better here in any way, in my experience during normal day-to-day use it is actually worse than Linux audio.
> I've yet to meet someone that encountered Window audio not working, barring some hardware/firmware issue
Hit that with an All-in-one-PC thing of a friend two years ago, had to reinstall Windows after some issue with malware. Sound didn't work because the driver from Windows Update was generic, you'd need the years-old version from the manufacturer installed first (as it contained the pin mapping somewhere in the bloatware that came with the driver). Fun fact, Ubuntu also didn't make sounds, I thought the speakers were busted and was already thinking about bringing a 'scope with me next time. Why pin mapping of audio chips can't be stored in an SPI chip or ACPI table is beyond me.
I've yet to encounter a case where Linux audio is not working. Thats because just like windows, I buy hardware where the manufacturer helps write Linux drivers. I don't financially support hardware makers who don't care about me and buy from those who do, simple as.
Also hardware companies should be happy if people are trying to reverse engineer their devices to get a working driver: It menas people want to use it. Instead of trying to further lock down or obfuscate their devices they should be happy and help these folks. They are doing free work for your company.
> Instead of trying to further lock down or obfuscate their devices they should be happy and help these folks. They are doing free work for your company.
They are, but legally the situation is way less clear for companies:
- for anything involving RF communication, allowing unauthorized/unverified selection of stuff like wifi country codes can lead to illegal RF emissions
- low level assistance may touch stuff that's nominally under NDAs and expose the company to legal risk
- anything involving "jailbreaking" may lead to issues with DRM
As a result, the official policy often is "no one touches code or helps people from outside, corporate legal doesn't want to get Oracle'd".
As I said all that does is further push me to only buy from companies that support my os. I have mixed feelings about people reverse engineering hostile hardware. It helps those who mistakenly bought that hardware esp as an onramp for Windows users who just bought whatever laptop. And it's always useful to keep reverse engineering skills shrap. But on the other hand you are just helping someone who hates you, the best way to play the game is to not play it.
Any setup with a bit more complexity than "playing audio from a single app without regards to latency" can be terribly janky and unpredictable in Windows, especially if it's multithreaded or running along some other CPU/IO-heavy app. Audio engineers and streamers know well what I'm talking about.
Oh, I have so much issues with audio in Teams on windows. Spotify works fine and I can hear the sound when connecting a call or meeting but then the audio is out. Or the mic. You never really know which one will cause problems. Replaced the BT speaker that caused problems for years but usb mic+audio seem just vaguely better.
There's a small but loud crowd which insists on using Linux as their daily driver on modern laptops despite this makes very little sense. https://xkcd.com/619/ is just as true today as it was near fifteen years ago, it's just today it would be streaming from various services (especially in high resolution). I have also struggled with Bluetooth, multifunctional devices and weird enterprise networks. For people who do not enjoy fighting their computers, Windows + WSL is just a smoother experience. Leave the Linux kernel for headless devices: routers, servers etc. It excels there.
I use Linux not only on laptops, but also on servers.
Also, I am more likely to have 2048 core laptop than playing full screen Flash videos.
https://xkcd.com/619/ complains about some Linux development focusing on support on server/supercomputer features and ignoring UX issues and support for some then popular media format.
The ironic part is that support for absurdly high core devices becomes more relevant, while Flash died.
I have been an open source developer for near twenty years now but that doesn't mean I subscribe to such notions. I am just way too old to feed another Tamagotchi for an OS.
Bluetooth, audio, etc works fine in the machines I use, because the manufacturer helps the development
of Linux drivers. The real question is why are you buying hardware from companies that dislike you and your os?
Tough luck, that doesn't exist. Linux doesn't work and the others, by your definition, work for some corporation. Although this is blind zealotry, O&O shutup takes care of those problems but when did facts stop cults -- because that's what Linux people became. I have been using Linux since 1993 and once I thought similarly and have grown thoroughly disillusioned. It never worked well on the desktop and extremely likely never will. There's a long, long list of forum posts on Arch of me trying to use the damn thing over too many years and then this breaks or that breaks.
> https://xkcd.com/619/ is just as true today as it was near fifteen years ago, it's just today it would be streaming from various services (especially in high resolution)
Speak for yourself; IME playing video is sufficiently boring that I don't even think about it.
Note that Perseverance does not run on Linux, the drone (Ingenuity) does. I know Perseverance has sent audio recordings. But is Ingenuity doing some audio processing?
> NASA settled on the Linux OS for the rover’s helicopter-like drone, Ingenuity [1]
> Ingenuity doesn't carry science instruments and is a separate experiment from the Mars 2020 Perseverance rover. [2]
[1] https://www.pcmag.com/news/linux-is-now-on-mars-thanks-to-na...
[2] https://www.nasa.gov/feature/jpl/6-things-to-know-about-nasa...