Hacker News new | past | comments | ask | show | jobs | submit login

Never had any problem with systemd, but I did have problems with PulseAudio — the high latency audio server. (These days it seems to work good enough, still looking forward to PipeWire.)



Back in the day I got into the habit of removing avahi and pulseaudio when I could because all they gave me were issues. Only later did I learn they were both by Poettering. When systemd was pushed, I started getting a whole new class of issues (due to renaming of drive and interface names - a problem systemd was designed to fix) which I never had in my entire Linux life. Nowadays it's not like that. Though I still get some systemd task running on a timer or indefinetely keeping the system from booting or shutting down. When the timer ends you realize it wasn't an essential thing after all.

And from a few posts of his I had read when systemd was being introduced one thing struck me. All of the posts of his that I read where he was defending something of his, the post would be full of blatant straw man arguments. It was very odd.


Pipewire is ready. Solved all my bluetooth headphone troubles too, I haven't had a single audio frustration since I switched.

I believe pipewire is already the default in Debian, Ubuntu and Fedora


Yeah. It's actually funny how Pipewire and Wayland turned out. Wayland was supposed to be ready for general use "any day now" since 2011. It's 2023 and the situation looks thus: <https://community.kde.org/Plasma/Wayland_Showstoppers>. Whereas Pipewire seemed to have scarcely any marketing, instead it just happened. I remember reading about it for the first time at some point and started wondering whether Debian, a conservative distro, turned it on or whether I should change a setting to opt into it. I looked at htop and lo and behold: Pipewire was there. So it happened before I even heard about it, without a single hiccup, so smoothly that I didn't even notice the sound system changed.


I think large part because pipewire was basically just going in with the attitude of fixing the underlying system while also supporting all the existing interfaces, which made it very much a drop-in replacement which could support pretty much all the common usecases. Wheras wayland is a big 'lets change the interface' deal (probably a good call actually) which didn't really seem to understand that it couldn't just punt on all the existing usecases for the old interface (I think there was some belief it could just be offloaded to compositor-specific APIs, which is a terrible idea guaranteed to make app developers hate you and irritate users when every combination of app and compositor acts differently), so you get a lot of breakage while people beg them to actually add standard interfaces for things.




Consider applying for YC's Spring batch! Applications are open till Feb 11.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: