From this user’s perspective (fairly typical Ubuntu user, not developer), Audio in Linux is a mess and this creates an opportunity for Ubuntu to add value to the stack.
I ran across this discussion while struggling how to understand / resolve my libvirt/qemu audio emulation problem (Looking Glass passthru-GPU support doesn’t and will never support spice audio).
As I said: what a mess. Apparently QEMU runs as a different user, has various esoteric settings for things like AppArmor, etc and I need to figure out somehow to get a pulse audio server running as that user (?) and route audio to my user id? This on top of all the ALSA/PulseAudio/JACK layers of complexity - it gets hugely confusing as obsolete & other distros are involved and probably often leads to screwed up machine configurations as us mere mortal users like me tinker with their systems in an effort to get basic things like audio support in virtual machines working.
At this point I’m mulling, “Hey, if I’m going to fix this, I should future-proof it a little.” I.e. switch to PipeWire. But at this point this seems to mean installing yet another ppa, drifting further from the typical supported configuration, probably causing problems at some point in the future. As someone who doesn’t mind tinkering but has to prioritize getting things done, I’m trying to keep the hacks to a minimum but it’s difficult because that always seems to conflict with the need for functionality that’s not provided by Ubuntu out of the box. And these hacks always seem to end badly, with support being withdrawn, breaking, conflicting with other hacks, etc.
For Linux and especially Ubuntu to become a mature operating system, more of these things become turnkey, meaning tinkering / hacking is minimized in order to get functionality that works for the mainstream user. Audio that ‘just works’ is a mainstream need.
From what I understand (very little), since it appears to be the latest, presumably most supportable and probably the long-term solution to Linux audio support, I’d suggest Ubuntu prioritize getting PipeWire installed into the stack ASAP with PulseAudio, etc configured to run on top of that if/when needed, but with the intention of depreciating alsa/pulse/jack/etc in the near future to simplify the system for those who just want a simple, functional system with minimal hacking & time-waste.
I hope this semi-relevant post is viewed constructively and helps resolve what I see as a sore point that needs to be deftly handled in order to get Linux & Ubuntu to a better place.