5 ms·

As a side question, who is kicking and screaming about PipeWire? PulseAudio had contention because of compromises and sacrifices in various places, but PipeWir

by Taywee 5y ago

As a side question, who is kicking and screaming about PipeWire? PulseAudio had contention because of compromises and sacrifices in various places, but PipeWire is really quite great for basically every use case, professional and casual alike, and effectively unifies the long PA/JACK split. I couldn't imagine anybody who has had to fight with Linux audio being anything less than ecstatic for PipeWire.

gjsman-1000 5y ago
It's bad wording on my part. PipeWire has been well-received, but that's partly because the kicking and screaming over PulseAudio (which PipeWire implements) already happened. Though if PipeWire had been introduced before PulseAudio occurred, it would have no doubt gone through the hate phase.
- rvense 5y ago
  I don't use Pulseaudio, so I only know it from "I've heard..." and that's all been negative, that it doesn't work or is annoying. For Pipewire, I've only actually heard good things. Maybe I'm talking to a specific set of people, but I still might try it.
  - ncmncm 5y ago
    PA mostly works, nowadays. It still spontaneously gets muted, once in a long while, but if I pop up enough different volume controls, I have always found a button to unmute it.
- yjftsjthsd-h 5y ago
  > Though if PipeWire had been introduced before PulseAudio occurred, it would have no doubt gone through the hate phase. PulseAudio's hate phase was caused by it breaking everything. If PipeWire had been introduced and worked, out of the box, correctly, then I doubt that it would have been poorly received.
  - ncmncm 5y ago
    The claim is that stuff was broken, and PA just exposed the bugs; and PW benefits from all the fixes done for PA. I have no way to evaluate the claim, and don't really care what the real story is. PA mostly works, now, and PW mostly works, now. But PW does have bugs, and is, inexplicably, coded in C, for which there can be no excuse at this late date. It could be switched over to build in C++ in probably under a day, and then begin to be modernized, anytime, as was done with Gcc and Gdb with unalloyed success. PW is written by the main guy from Gstreamer, which used to crash all the damn time back when PA went silent a lot. But he is older now and has lots of help.
- michaelmrose 5y ago
  I have continually heard the assertion that Linux desktop folks just hate change much at odds with the actual community of users who by and large willingly changed from what comes on 99% of computers learning not just one new thing but in fact many over time. Anyone who has stuck with the Linux desktop for any substantial portion of the 20+ years its been a thing has had to change plenty. Indeed the Linux community has adopted many things eagerly over time. Most of the biggest communities are around distributions that have releases every 6 months - 1 year and proudly proclaim all the things that they have improved and changed. It's an entirely bad faith argument. Audio prior to pulseaudio on the Linux desktop was hot garbage primarily for lack of hardware support for various built into the motherboard sound chips. The only viable solution was to buy supported sound cards which of course was only easy on desktops. With the launch of pulse even those machines experienced curious problems with a crashy, confusing, poorly engineered face to the already bad audio situation on desktop Linux. It ended up with both the credit for its own bad engineering and for the underlying shoddiness of the substrate it operated on by being the thing that visibly failed to produce sound when a user clicked on a video. The cases wherein audio was magically fixed by disabling it using alsa alone created loud and persistent critics whose criticism persisted even when both pulse and the underlying audio problems decreased. This is indeed how things work. If I put syrup and lots of salt in a glass and float a radish in it and proudly plaster my brand across the cup I shall never sell you another cup no matter how much I improve the formula for my "soda". Labeling people who came by their criticism honestly thoughtless Luddites doesn't help. You might notice that most products don't go through a "hate phase" and new software often does. It's not just because people don't like change its because as a species we are remarkably bad at making software and new software is often garbage unfit for purpose. If much software were a toaster you should have girded your foot with a steel toed boot and kicked a field goal out your back door with the disgraceful junk you were tricked into buying.
  - smoldesu 5y ago
    Precisely, thank you for saying this. I don't hate change; in the case of PipeWire, I absolutely welcomed it. The alternative was obviously broken and couldn't be iterated on in any meaningful way, so replacing it with a higher-level interface made all the sense in the world. I loved a lot of the iterations GTK3 made over GTK2 so the interfaces would be equally at home on a tablet as they are on a desktop. I appreciated the work that went into writing GUI package managers even though I'd never really use them, and while I don't agree with them, I'm happy that people are experimenting with containerized packaging. Linux users don't hate change, we just hate regression. A lot of people's workflows rely on exploiting extensibility and niche features, much as others do on Windows and MacOS. When people pull the plug on old systems, or limit the interoperability of their application, they shouldn't wonder why their reception is negative; people want to take your project to the next level, all you have to do is let them. PipeWire was beloved because it did let people extend pre-existing features. It was a powerful tool for managing audio cart-blanche, and it released in a fairly feature-complete state. It's an example of the perfect Linux software overhaul in my opinion. There's a lot of generalizations in the grandparent comment about how distributions 'hate' the lack of theming and whatnot, and all I can say is that those kinds of opinions are almost entirely founded on a premise of not understanding the majority of Linux users. Extensibility is king, and it makes a lot more sense for them to build a flexible toolkit that allows for both theming and accessibility options instead of locking it into a one-track design philosophy that just-so-happens to benefit a small handful of users. That's what people are mad about.
pxc 5y ago
I just finally switched over to PipeWire last week. I enabled PulseAudio and ALSA emulation, since that's what my apps expect. I never would have noticed the change except that detecting and setting audio profiles/codecs on my bluetooth devices from my sound control GUIs works much better than before! It seems to be damn good.
- csdvrx 5y ago
  > It seems to be damn good. It is. PipeWire is good. PulseAudio was bad. End of the story for me.

by Taywee 5y ago

gjsman-1000 5y ago

It's bad wording on my part. PipeWire has been well-received, but that's partly because the kicking and screaming over PulseAudio (which PipeWire implements) already happened. Though if PipeWire had been introduced before PulseAudio occurred, it would have no doubt gone through the hate phase.

rvense 5y ago
I don't use Pulseaudio, so I only know it from "I've heard..." and that's all been negative, that it doesn't work or is annoying. For Pipewire, I've only actually heard good things. Maybe I'm talking to a specific set of people, but I still might try it.
- ncmncm 5y ago
  PA mostly works, nowadays. It still spontaneously gets muted, once in a long while, but if I pop up enough different volume controls, I have always found a button to unmute it.
yjftsjthsd-h 5y ago
> Though if PipeWire had been introduced before PulseAudio occurred, it would have no doubt gone through the hate phase. PulseAudio's hate phase was caused by it breaking everything. If PipeWire had been introduced and worked, out of the box, correctly, then I doubt that it would have been poorly received.
- ncmncm 5y ago
  The claim is that stuff was broken, and PA just exposed the bugs; and PW benefits from all the fixes done for PA. I have no way to evaluate the claim, and don't really care what the real story is. PA mostly works, now, and PW mostly works, now. But PW does have bugs, and is, inexplicably, coded in C, for which there can be no excuse at this late date. It could be switched over to build in C++ in probably under a day, and then begin to be modernized, anytime, as was done with Gcc and Gdb with unalloyed success. PW is written by the main guy from Gstreamer, which used to crash all the damn time back when PA went silent a lot. But he is older now and has lots of help.
michaelmrose 5y ago
I have continually heard the assertion that Linux desktop folks just hate change much at odds with the actual community of users who by and large willingly changed from what comes on 99% of computers learning not just one new thing but in fact many over time. Anyone who has stuck with the Linux desktop for any substantial portion of the 20+ years its been a thing has had to change plenty. Indeed the Linux community has adopted many things eagerly over time. Most of the biggest communities are around distributions that have releases every 6 months - 1 year and proudly proclaim all the things that they have improved and changed. It's an entirely bad faith argument. Audio prior to pulseaudio on the Linux desktop was hot garbage primarily for lack of hardware support for various built into the motherboard sound chips. The only viable solution was to buy supported sound cards which of course was only easy on desktops. With the launch of pulse even those machines experienced curious problems with a crashy, confusing, poorly engineered face to the already bad audio situation on desktop Linux. It ended up with both the credit for its own bad engineering and for the underlying shoddiness of the substrate it operated on by being the thing that visibly failed to produce sound when a user clicked on a video. The cases wherein audio was magically fixed by disabling it using alsa alone created loud and persistent critics whose criticism persisted even when both pulse and the underlying audio problems decreased. This is indeed how things work. If I put syrup and lots of salt in a glass and float a radish in it and proudly plaster my brand across the cup I shall never sell you another cup no matter how much I improve the formula for my "soda". Labeling people who came by their criticism honestly thoughtless Luddites doesn't help. You might notice that most products don't go through a "hate phase" and new software often does. It's not just because people don't like change its because as a species we are remarkably bad at making software and new software is often garbage unfit for purpose. If much software were a toaster you should have girded your foot with a steel toed boot and kicked a field goal out your back door with the disgraceful junk you were tricked into buying.
- smoldesu 5y ago
  Precisely, thank you for saying this. I don't hate change; in the case of PipeWire, I absolutely welcomed it. The alternative was obviously broken and couldn't be iterated on in any meaningful way, so replacing it with a higher-level interface made all the sense in the world. I loved a lot of the iterations GTK3 made over GTK2 so the interfaces would be equally at home on a tablet as they are on a desktop. I appreciated the work that went into writing GUI package managers even though I'd never really use them, and while I don't agree with them, I'm happy that people are experimenting with containerized packaging. Linux users don't hate change, we just hate regression. A lot of people's workflows rely on exploiting extensibility and niche features, much as others do on Windows and MacOS. When people pull the plug on old systems, or limit the interoperability of their application, they shouldn't wonder why their reception is negative; people want to take your project to the next level, all you have to do is let them. PipeWire was beloved because it did let people extend pre-existing features. It was a powerful tool for managing audio cart-blanche, and it released in a fairly feature-complete state. It's an example of the perfect Linux software overhaul in my opinion. There's a lot of generalizations in the grandparent comment about how distributions 'hate' the lack of theming and whatnot, and all I can say is that those kinds of opinions are almost entirely founded on a premise of not understanding the majority of Linux users. Extensibility is king, and it makes a lot more sense for them to build a flexible toolkit that allows for both theming and accessibility options instead of locking it into a one-track design philosophy that just-so-happens to benefit a small handful of users. That's what people are mad about.

pxc 5y ago

I just finally switched over to PipeWire last week. I enabled PulseAudio and ALSA emulation, since that's what my apps expect. I never would have noticed the change except that detecting and setting audio profiles/codecs on my bluetooth devices from my sound control GUIs works much better than before! It seems to be damn good.

csdvrx 5y ago
> It seems to be damn good. It is. PipeWire is good. PulseAudio was bad. End of the story for me.