Hacker News Reader

Knowledgeable people, please explain. Wouldn't it be better if applications used ALSA for audio output? In this case they could work without Pulseaudio and with (because Pulseaudio adds a sink to ALSA). This way the application doesn't need to know which audio daemon it uses. What are the downsides of this approach?

View on HN

People like to use the pulseaudio API because it has features people want, compares to ALSA or JACK: See here:

https://gitlab.freedesktop.org/pipewire/pipewire/-/wikis/FAQ...

I use pure Alsa on virtual machines. Image is stripped down so much it barely runs XServer.

Another reason is future proofing. PipeWire is yet another audio protocol, and may not be around in 5 or 10 years.

Not everything has PipeWire. Some flavours of linux has support for 10 years. Even some of my desktop boxes still use Pulseaudio.

And finally what PipeWire actually offers? If you need to dump audio into output, ALSA seems sufficient.

From what I've read PW's official stance is to keep using the existing protocols such as PA and JACK. PW comes with a PA-compatible server, replacement JACK client libraries and an ALSA plugin. So it supports clients from all three at the same time. Precisely so software does not need to be rewritten (I imagine).

Reply on news.ycombinator.com