PulseAudio to be replaced by Pipewire

After all the shit design, shit implementations, and broken audio involved with PulseAudio, they're just gonna be starting over from scratch anyway, and now we're gonna have this new shit called Pipewire
Not even fucking kidding.
pipewire.org/

Other urls found in this thread:

en.wikipedia.org/wiki/Wim_Taymans
github.com/i-rinat/apulse
ltsp.org
wiki.x2go.org/
jwz.org/doc/cadt.html)
rust-lang.org/en-US/
cloveros.ga/s/CloverOS-x86_64-20180118.iso
debian-interix.net/
twitter.com/SFWRedditGifs

Is it Lennartware©?

This is the most cancerous 'fix' for sound on Linux yet. Can't the just port sndiod and call it quits?

Isn't this basically gstreamer?

afaik it's not made by Poettering himself, but it's a Redhat project.

I think it's a weird combination of gstreamer, PulseAudio, and JACK.
To be honest, i'm not entirely against the idea of bringing PulseAudio and JACK together, but why in the fuck does video have to be involved now?

Wim Taymans sounds like a white background character in a Japanese video game.

fuck. this. gay. earth.

Containers, apparently. Forwarding X and audio from containers without compromising security is not entirely trivial, this could make things easier. Imagine using higher-risk stuff like internet browser from secure container with no hassle. Seeing as it's redhat working on it, I'm afraid it's gonna be botched, but since it's not Mr. L. working on it, it might just not end up being shit.

yeah redhat's usually shit, but occasionally has a good idea or two, like libvirt.

Gota get that extra Red Hat vendor lock in.

Why does it keep mentioning "flatpak" when I look this up? wtf are red hat up to?

I'd say if they pick up the difficult parts from jack, there's a chance it will actually be the rare case when it's gonna be widespread thanks to being good as opposed to being widespread thanks to being not-quite-as-bad-or-user-unfriendly-as-everything-else.

Disclaimer: I'm probably wrong on some of this.

Flatpak is a new application distribution format they're making that will allow for the easier management of desktop apps.
For example, you get Chrome as a flatpak, and it will be in a sandboxed container with all the dependencies bundled.
I think it's supposed to be like how proprietary Windows and Mac software comes with all the libraries and shit it needs bundled with it.
I guess their plan is for this new pipewire system to be able to more effectively handle sound and video coming to/from these sandboxes.

You posted this thread on halfchan, too. Are you a fucking retard? This has been common knowledge for literally years.

Yes, MacOS ``apps'' are actually folders, which is easier for the layman to wrap their head around when downloading consumer software.
That is not the reason why pipewire exists.

It's new to me, and it would appear that i'm not the only one.

Check the archive before you make another retarded thread.

no

Yes it literally is. A major motivation for PipeWire was an abstraction layer so that sandboxed applications didn't have to be aware of audio or video drivers... which means interfacing with libGL to avoid multi-vendor hell on the Flatpak side.

Oh great we don't even have pipeware yet and there are already pipeware shills.

All hail pipeware, pulse was good but pipeware's the future right?

This is exactly what I wanted to ask.
also looked at the github issues and it seems like the devs are not clinically retarded like lennart is, they actually seem to take issues into consideration and act upon them instead of EWONTFIXing everything.
Overall, mite b cul/10. A replacement for pulseaudio is long overdue, I've been on alsa for a while now since pulse has shit audio quality and still bugs out to this day. But the lack of per-application sound control is pretty annoying.

Good point. Without Lennart in the picture, this might actually not suck as much.

is he someone's retarded nephew at redhat? they need to get rid of him

this
you can use sndio on linux, but you have to recompile most of stuff to support it but it's not always possible

Why is pulse audio so bad? I've never had any issues with it

...

i wouldn't say white, i think he's spanish.
"Belgian national based in Barcelona, Spain"
en.wikipedia.org/wiki/Wim_Taymans
Belgian national doesn't mean anything since none of the EU countries have borders anymore.

There isn't much on him, he had social media at some point but hasn't updated any of them in 5+ years, he appears to have abandoned them but left them up.

Don't worry. He will integrate it in to GNU/systemd

So is this going to be good, or a monkey paw?

Can't say the same for me, unfortunately

So is the case for most people.
There are legitimate cases to be made against PulseAudio. There are some circumstances were Jack or even ALSA is better; but most of the criticism on *chan boils down to:
For most case scenarios Pulse is good and works good.

Each country still requires visas to stay in. The "free movement" is about not having to wank off at the border with all of it's kinds of arbitrary waits, checks, and regulations.

lennart pls go

Pulseaudio has always worked on my machines I used in on too. But it uses like 1-4% proccessing power when idle, which is literally more then my web browser. So I switched to entirely ALSA to avoid the botnet.

It's a fucking botnet. This is because it is abstracting pulse, alsa, OSS, opengl pre 3 and post 3, DRI2/3, Glamor, KMS, and fucking whatever openbsd/freebsd uses. It's also made by redhat. They would have to design it to be modular so you could compile out what you don't need to abstract per instance. Ontop of the fact that if they are going for wayland usage they are going to have to abstract every wayland compositor's functions which will increase bloat per window manager they support. Also assuming you don't get a won't fix for certain untained by (((them))) wayland compositors.

Pick one, OSS did what pulseaudio did with the performance of ALSA and still performed better then ALSA.

This.

Pulseaudio has no usecase now. Use alsa if you need exotic bluetooth audio devices with multiple outputs. If you want software mixing of audio for things like recording and karoke use JACK with alsa. If you want all the features of pulseaudio without download jack for ALSA then install OSS drivers if your audio card is supported by them. I won't deny that pipewire is a step in the right direction for being rid of pulseaudio, but adding video to the mix is only for (((them))) and for vidya.

How does that work with an open source project that anyone can use?


I'll give one positive case for Pulseaudio: Easy output switching since I have a USB sound card, but some also have HDMI which, frankly, operates like shit with most sound systems.

alsa does support that though rtfm.

...

It'd be Shin Kaf, too... ShK. SheKel.

"Wim" is a very Dutch/Flemish name, short for Willem (William). "Taymans" too, I think, although I haven't seen it before. "mans" is a common suffix.

Can't they just swap to alsa?
It has been working just fine all this time, and pulseaudio itself was just a pile of bloat over it.

There is nothing wrong with putting all software settings into software's directory along with folders instead of evenly smeared configs across /etc, /bin/usr/etc, /usr/etc, /home/user and so on.

There is nothing wrong with supplying software with containerized set of dependencies, on a desktop of course, not an embedded router woth 1mib ROM. Supplying dependencies prevents package management struggles and dependency hell. NixOS is going in the right direction.

However, RedHat are capitalist jews that write shitware to sell their useless (((certificates))) and tech support to end users since selling free software is really pointless with today's broadband speeds, they decided to complicate the software and system administration procedures up to Microsoft's level. See, modern Gnome/systemd distros boot slower than even windows 10 out of the box, not joking, true story.

so systemd is the software equivalent of aids, and they sell the drug?

It makes sense in a single user system (Windows (multi user is a hack that shits itself frequently)), but it's better to have global and local configs in a multi-user system (UNIX).

Yes, they sell both doctor certificates and recovery programs for the sick.

Why would I need PulseAudio® again?
I mean, doesn't Alsa get the job done regarding audio?

If you want Windows, I suggest you fuck off to it.

I thought of Hebrew letters when I saw that image, too. What are (((they))) up to?

because shitware like firefox might drop alsa support

there's a shim now

github.com/i-rinat/apulse

You could always stop using shitware like firefox and use palememe or icecat. Or hell, compile it yourself with only ALSA or OSS support.

When was the last time you've seen a "multi-user" "Unix" system being used as multi-user, not an abstraction for single-user application permission control, or a web server that does all the user-management itself? Multi-user systems are the reason we have "cloud" cancer. Multi-user systems are not needed for most computing appliances.

This. Multi-user UNIX is a 40 year old concept that belongs to the era when computers were too expensive for individuals to own, necessitating time-sharing systems on a single machine.

The multi-user properties of UNIX have been repurposed as a kind of poor-man's privilege separation, which is a pathetic substitute for appropriate kernel-level access controls and capabilities.

Multi-user OSes on single-user machines are a relic, and it's past time for them to die.

sdf.org, 31 years of doing just that

More like 50 years old, not 40.


SDF has a couple dozen active users, tops, and runs NetBSD. It's the past, not the future, and it's an utterly miniscule representation of internet users, operating system users, computer users, and any other metric you can measure. I'm an ARPA member and even I think it's largely a relic.

lol

Proof or get out as a liar without an arguement. Single user is indeed cancer for large organizations and servers. But has its uses in a terminal for plebs. Yet not one person has developed a microkernel OS around such a concept that isn't shit or non-FOSS.

Security is critical though, and multi-user lets you do privilege separation. With old 80's computers it didn't matter so much, even without memory protection. Back then you could just power-cycle the machine and return to factory state. Now you have malware that digs into your firmware and hides itself where the OS can't see.
And security has been a big concern on Internet servers since the mid 90's at least, when the buffer overflow exploits became widespread. Cloud shit came much later.

OpenBSD makes heavy use of user-level privilege separation, with good results. It also enforces system calls with pledge. You can do both, and every little bit helps.

Like what? I can't think of anything that isn't better handled with something like namespaces.

literally not a bug
there's nothing worth its shit on the internet which has to be listened right in the browser anyway

here's the linux community, gents

Only through manual specification. So every time you open an application, instead you should wrap it with an alsa sound-source redirection, but only after manually writing to the alsa.conf to add that sound source. You should rtfm instead of being retarded.

fuck off with this retarded argument.

why don't people just write alsa plugins?

How is it that Microsoft can get away with using the same sound API for 20+ years yet Linux has to have a new one that's total shit every 2-3 years? At this point, why not just admit not having a clue and copy Microsoft's? Putting a real-time buffer in userspace was never going to fucking work and the novelty of watching people try and fail wore off before most of you were born.

Literally today.
ltsp.org
wiki.x2go.org/

I give my friends unprivileged shells on my colocated Debian system, and since I have x2goserver running, they can even use it as a remote desktop. I've also previously run a telnet/ssh BBS on a Linux system. My record was six simultaneous human users plus me.

Since you don't even have a single clue about what you're talking about, you're allowed to not post ever again.

the only good think of PulseAudio is pavucontrol, which is useful when you have multiple audio-cards or outputs

Hans, get the gas.

i think m$ changed their audio shit around vista.

that's the opposite of gnu/systemd philosophy user.

There was never a moment on Windows where there was a question of if a browser would have sound after an OS update.

Worst part is another superior operating system did this already: Solaris! It lives on today as Illumos kernel and OpenIndiana project. Joyent's SmartOS and SDC is the best FOSS hypervisor and pop-up cloud you can find. Once hardware is installed and wired up, it's only a matter of an hour or two with SDC before you have a live cloud. Solaris Zones were invented about a decade ago, yet Linux is STILL trying to ape them and BSD jails still haven't surpassed them despite jails being the project Sun was motivated to duplicate in Solaris.

SPARC-platform second hand servers are going for cheap on eBay right now. SPARC is the native architecture for Solaris. SPARC also relies on OpenFirmware rather than kiked EFI. Not sure if SPARC was immune to Spectre or Meltdown, but chances are good SPARC and POWER+PowerPC were being that they were independent designs unless they leased rights to Intel's IP to use the vulnerable components.

Do you not understand that video is normally correlated to one or more audio streams? They're building inherent support to match video streams to audio streams.

THAT IS BLASPHEMY!
Never intermingle your audio and video streams! How unorthodox, I could faint!

You missed the point. The streams are not intermingled, they are processed in disparate streams and synchronized in the output presentation.

Have fun trying to secure such a system where no one's maintaining a million copies of a million versions of the same library on your system with a vuln.

Wrong.

Done, my system is secure until next time.

Remember that RH makes their money from support.

So what? What's stopping you from not using this technology?

How the fuck is audio still such a big issue on GNU/Linux? Just fucking work to improve what's already around instead of making a new system every couple of years.

This worried me too.
Every distro I tried it on, it was using 1-5% cpu when completely mute.

fuggggg. (captcha beginning: jwh)

=>>850443
just FUCKING REPLACE ALSA

If things don't break Red Hat cannot make money.

fuck that shit. I have no clue how it's possible for a sound system to work but only with specific media formats. Somehow, it's a well-known issue and hasn't been addressed in years though.

Why replace something that works just fine?

Are you retarded? The entire company business model is built around their product breaking.

...

I thought Poottering's PulseAudio was meant to copy Microsoft and MacOS's audio? I think he mentioned this at one point.

Not Microsoft but Apples high-level audio libraries. Part of the reason why most laymen associate macOS with audio and content creation, because the process on MacOS is easier and more streamlined.

Hmmmmmmmm

Not even remotely close, nor related to any of the technology discussed in this thread.

I've also heard systemd being compared to svchost on windows

This the CADT Model in action. (jwz.org/doc/cadt.html)

PulseAudio broke Linux audio for like a decade. After a decade of turd polishing it has finally started to work well. And now they're gonna write another PulseAudio, and break Linux audio for another decade.

What is wrong with these people? Code does not rust, it does not go bad or spoil. Why do that act like it does?

Firefox has an sndio backend.

:^) rust-lang.org/en-US/

lol there's a reason firecucks is disappearing into irrelevance

With what application? VLC works just fine with it. Same for my audio player.
That's a lie, its just how you set it up. I can play sound in webm's just fine you faggot. Did you forget to adjust settings under ossmixer? Are you using crapware like firefox with it instead of something else? Do you even have libsalsa set up properly?

Is this the power of Poettering?

You also have to compile Linux for it to work. Is Linux disappearing into irrelevance?

Linux is so relevant that even the head of the linux foundation doesn't use it.

Go back to whatever asylum you came from. Try not leaving it this time.

I haven't been using Linux for a few years now, but wasn't it only just recently becoming common for distros to come with Pulse instead of ALSA?

Pulse operates on top of--not instead of--ALSA on all of the distros with which I'm familiar.

The answer to your question is no. Fedora started using Pulseaudio as their default sound system in 2007, and Ubuntu did the same in 2008. Fedora and Ubuntu and their derivatives constitute the vast majority of desktop Linux installs.

pulseaudio is very much undesirable by my experience
when trying to get a mic working, the mic was not muted, but it still wouldn't pick up sound
after playing with settings, i got the mic to work, but the speakers wouldn't work, which is necessary for dubbing and monitoring dubs
it took me in all less than an hour to eventually get it to read input from the mic and output data to the speakers,
but really what makes pulseaudio stand out on its design philosophy, there is no quit button or quit menu item to select in pavucontrol, so you are dependent either on the close button of your preferred wm or know by experience the common keyboard combination for closing the program

Broken window fallacy.

No it didn't. Pulseaudio exposed buggy ALSA drivers for like a decade. After a decade of ALSA driver fixing it has finally started to work well for those poor people.

You're replying to a descriptive claim, not a normative one.

And the number one solution for fixing those buggy ALSA drivers? Uninstall pulseaudio.
really jogs the old noggin.

This tbh, stop using alsa and pulseaudio and switch to OSS drivers if your chip is supported.

Is this your solution to running out of gas, just walk away from the car

...

Who are you quoting?

you really must have brain damage

yeah. that's sums up the linux sound situation pretty well.

I just removed Pulse and went back to ALSA. Feels great.

If using false quotes gives me brain damage, I'm filled with brain damage.

Wayland is so restrictive you need something like this just to run software like OBS or scrot. It's not "pulseaudio for video", it's DRM for screenscraping.

I have never had pulseaudio automatically switch to headphones on a new Linux install

You can change it to make the audio go to it without restarting the application. But you can do this same thing in OSS. But I have never had it be fully automatic either.

I found you could make it mostly automatic by enabling some weird proprietary HP setting in ALSA

lul wut nigger? Explain with proof now or get out

There's this weird proprietary HP setting from this third party software that you can use to help you configure ALSA that always comes up when you need automatic headphone detection. That's all I know, I'm not really that good at the internals of linux

realy maeks you think
Stop using shitware and it will just work

That was committed recently.

Why is sound still such a clusterfuck on linux? The *BSD's seem to have put this issue to bed years ago.

I may not be able to get video acceleration to work on Open/NetBSD on some rando hardware but the sound always fucking works.

I'd just like to interject for a moment. What you’re referring to as Cux, is in fact, GNU/Cux, or as I’ve recently taken to calling it, GNU plus Cux. Cux is not an operating system unto itself, but rather another free component of a fully functioning GNU system made useful by the GNU corelibs, shell utilities and vital system components comprising a full OS as defined by POSIX. Many cucks run a modified version of the GNU system every day, without realizing it. Through a peculiar turn of events, the version of GNU which is widely used today is often called “Cux”, and many of its users are not aware that it is basically the GNU cuckoldry, developed by the GNU Project.

There really is a Cux, and these people are using it, but it is just a part of the system they use. Cux is the kernel: the program in the system that allocates the machine’s resources to the cuckoldry processes of our Cucklord, Richard Stallman. The kernel is an essential part of an operating system, but useless by itself; it can only function in the context of a complete operating system. Cux is normally used in combination with the GNU operating system: the whole system is basically GNU with Cux added, or GNU/Cux. All the so-called “Cux” distributions are really distributions of GNU/Cux.

GIT GUD

FreeBSD may actually be worse than Linux. Say Dragonfly, at least.

Second hand sparcs are where it's at. Seriously, 120 bucks will get you a fun hobby rig. Actual usability? Pretty good, just get used to compiling all your stuff. They are loud so make sure you have good fans and a chassis that will fit their proprietary board designs, unless you like the primal screach of 6 40mm fans.

Go back to facebook.

From a security perspective, separating the files itself into different partitions is a drastic security improvement to prevent container escape, rather then some virtualized shit form of dockerpajeet code to do so. Can't access/change files from a specific attack vector if you don't have access. If it's all bundled together it becomes significantly easier to priv esc out of some shitty container.

At least you understand the Redhat (((problem))).

Can't wait. Honestly I'm tired of whatever is in charge of audio on Linux.

Who are you quoting?

Are you running plain ALSA?

PulseAudio. I believe it's the same setup Ubuntu LTS uses.

Linux users are considered lower-caste by Discord's "dudebrah coder" javascript weenies, check out the condescending cringe shit they put in the settings flavor text.

If you are using pulse audio, you can just use pavucontrol to select the correct microphone.

You're right, greentext is for quoting, but it can be used for quoting context/what someone really means as well. This is not new.

Get CloverOS if you want a 100% ALSA-only distro.

cloveros.ga/s/CloverOS-x86_64-20180118.iso

For me it always picked the wrong device or managed to break so hard that shit wouldn't start properly at random, even a fresh reinstall of pulseaudio wouldn't fix it. Not only that I had weird sound crackling and a really shitty bass when using it. This is some years back for those reasons never used it again. Just ALSA works fine for me.

Well I'm sold.

Pulseaudio has a habit of exposing faulty ALSA drivers. When running ALSA directly, there seems to be no problem but when run through PA, all kinds of problems reveal themselves. This is a good thing because it leads the driver writers to look into the driver to improve the drivers making them work better with or without PA.

That could very well be but when PulseAudio shits up like a kid on a minefield I wouldn't touch it. Again, haven't used it in years so the problems I have might have been fixed, I just couldn't care less nowadays.

Yeah, I'm gonna ask that chink fabless manufacturer to improve their drivers. Oh wait, that never happens because reality isn't like your autistic fantasy world.

Kill yourself and take that shitty audio stack with you.

ALSA is a direct part of Linux. Everything that's in the current version of Linux is guaranteed to have a maintainer looking at it. You can find Linux drivers for the most obscure hardware working today simply because there is a maintainer willing to dedicate his time to it.

Teamspeak, while also being proprietary, at least let's you pick capture mode

Mumble has this too. I couldn't take a screenshot while the drop down menu is open (don't know why but this is how Xfce behaves), but it allows me to select ALSA, OSS and PulseAudio. I'll try using ALSA directly from now on and see if discord also has this option. If not I'll try OSS since it also seems to detect voice.

I don't know why everyone doesn't just use OSS.

Upstream it and we'll talk.

Because it isn't sndiod

Question. Why should I use straight ALSA over Pulse, and how would I go about setting it up?

Oh great, yet another attempt at outputting sound.

...

Can you tell me why this is so significant to you? Are you implying that you will be forced to use this?

Because unless you need pulse for something specific you don't need pulse at all.
RTFM

Now you understand one reason why I'm a UNIX hater. When they make something new from scratch, it becomes another thing that has to be kept around forever, like Perl. Perl is supposed to replace sed and awk, but having Perl on your computer doesn't let you get rid of sed and awk because of the UNIX "tool" philosophy. Some program somewhere is going to be using sed or awk. Attempts to consolidate bloat and complexity just make the system more bloated and complex. There's the same situation with audio on Linux.

> What I can't figure out is why there isn't a giant market> for improved unix software. For example, it seems like it> would be straightforward to write a decent C macro> processor or garbage collector, and that you could make a> bundle of money selling them because everyone would want> them. But no one does this. Why not? Maybe it's because> weenies are so used to not fighting city hall that they> can't believe things could ever be better?> You really can't figure this out? It's because everytool depends for its operation on the bugs in every othertool, to exaggerate slightly. Thus anyone promoting animproved version of anything runs smack into insuperablecompatibility problems. You have to work as hard asStallman to make any headway at all.

I don't think there are any plans to port this to the HURD, so it has nothing to do with GNU. Just Linux.

The operating system is GNU.

Like Win32 :^)

That's the propaganda that rms and the GNU project put out because of their failure to actually complete the GNU OS, which was to use HURD.

GNU's not Unix. It's also not Linux. And it's also not an operating system. Sorry. No amount of semantic masturbation and wishful thinking on rms' part will make that true.

Stallman was a highly experienced writer of operating systems long before the birth of GNU. He gets to call GNU an operating system because he was there at the beginning of operating systems doing the work of writing operating systems. GNU OS started in 1983 and it has been an OS since 1983.

underrated post

He aint but it probably can be a mandatory part for a lot of distro's.

Got any pointers on how you'd set up your system for OSS? I don't exactly like to fuck with audio drivers.

GNU was always meant to be a modular operating system that could work under different kernels. HURD itself can use different microkernels. So far GNU has worked using Linux, kFreeBSD, NT, HURD and the Solaris kernel.

What?

Also known as "Linux Subsystem for Windows."
It's literally the GNU userspace running on top of the NT kernel. There is no Linux code running at all, though they do translate some Linux functions to the NT kernel; not unlike how Wine translates some NT functions to the Linux kernel.

If you can write a compliant POSIX layer for NT, then it's feasible to pair GNU with NT. Today there exists (adequately) compliant POSIX systems designed for Windows.

Ironically enough, Windows 7 and 8 had POSIX-compliant layers. Those were, in fact, more POSIX-compliant than GNU itself.
It was deprecated in favor of the Linux Subsystem for Linux in Windows 10, however.

Can you elaborate more on the POSIX compliance of Windows SFU and GNU. I remember seeing talk that while SFU existed and had all the feature lists, I heard it didn't work too well for Unix software. I didn't actually use this on Windows and I don't have experience in this configuration.

Not really. I never used it either.
As far as I know, Windows' POSIX subsystem was 100 % compliant since Windows Vista because some government regulation required the OS they used to be POSIX-compliant.

Reading a little bit more into SfU, Debian was ported to it: debian-interix.net/