Hacker News Reader

Comment by nitrogen | Hacker News Reader

For systemd the consensus seems to be that's it's better than the previous init systems as most distributions have adopted it.

I think the parent comment was alluding to the fact that RedHat has the ability to manufacture consensus. All the distributions use systemd (and pulseaudio and ...) because GNOME and other core components were modified to work only with systemd. So everyone agreed to use RedHat's way because there was no other way. No matter how good RH's creations might be, that's the kind of tail-wagging-the-dog power that is bad for an ecosystem.

View on HN

What you are saying is a borderline conspiracy theory and is not true. GNOME still works outside of systemd, and the distributions that want to use it outside of systemd are maintaining the parts that they need, which in reality are pretty small. (I can give some more technical details if you want, but the short answer is: elogind)

Maybe things have improved over the last several years, now that systemd is fait accompli. Earlier on in its life I recall discussions of GNOME features that would only work on Linux+systemd, and the BSDs were complaining.

But the point remains that, even without any ill intent, the lack of competition and compromise can produce outcomes that look like malice -- people gradually get used to making decisions that go unquestioned, and unwittingly walk all over everyone who no longer has a voice in the discussion. This pushes more people out, and the cycle of the tail winning against the dog repeats and continues.

I remember that too and the main issue was that BSD didn't have an answer to logind and it took them a while to port later versions of GNOME 3 because of it. You can't seriously expect Red Hat or some other Linux distribution to drop everything and figure out the best way to get new features to work on all the BSDs before they release something, that's something only the BSD developers can do. If the "dog" here is supposed to be BSD, the wagging is not working on them.

One example is a tree. We are looking at a forest.

Not really, AFAIK this is the only part of systemd that really caused an issue, and it was because this is a capability that really needs to be provided by the OS and can't realistically be maintained as part of a desktop environment.

- Init system responsible for way, way, way more than init, gradually taking over everything to such a degree that distros are forced to abandon Upstart, SysVinit, whatever Gentoo used back in the day, etc.

- Audio system that ignores all the foundations of ALSA and builds Yet Another Layer instead of improving ALSA.

- Display system that leaves way too much as an "exercise for the reader".

- 10+-year-long transition periods on the above, during which nobody's use cases work well.

- Decisions made by fiat rather than by consensus.

Gentoo still has their own init, some distributions still use sysvinit, and both can run GNOME. Upstart is a dead project AFAIK, and the developer dropped it and said it was strictly inferior to systemd. ALSA is just the kernel layer, there is a limit to what you can put there, so some of it has gone into another layer in userspace, that's by design in the kernel. I won't comment on the display system, everything that needs to be said there has already been said in the other 700 comments here. I also wish this stuff went faster but unfortunately it takes time to do real work.

I'm not sure what you mean by consensus, if there was consensus on everything then we wouldn't have so many Linux and BSD distributions. We'd probably have just one or two, and you'd probably be more annoyed about "wagging the dog" if that was the case.

ALSA is just the kernel layer

Just one point about ALSA; most of it actually lives in userspace. The ~/.asoundrc file can specify complex routing and mixing, including the use of LADSPA filters. What's done is done, and pulseaudio is reasonably usable now, but IMO we'd be much better off if the work that went into Pulseaudio and jackd instead went into the ALSA userland.

I don't understand what you're suggesting. To put the work into the ALSA userland, you would make an additional daemon on top of it and put features there, which is exactly what Pulseaudio and jackd are.

ALSA already comes (or at least did in the past) with an optional daemon for dmix and dsnoop. Putting as much of what jack and pulse do as possible into libasound (which already handles a lot IIRC), with the minimal amount necessary into a shared daemon, could have allowed all existing ALSA applications to work with the dynamic switching that is now handled by PulseAudio, or the low latency that is now handled by Jackd. ALSA already had a signal graph that could have been evolved into the independent signal graphs that PulseAudio and Jackd can now build.

elogind from what I could tell started as a hack to get gnome to work outside of systemd. It now provides a way to use gnome without systemd but it's definitely going against the grain. Your comment implies it's easy for gnome to work without systemd. I remember the early days, it was pretty buggy since gnome essentially was built to run with systemd.

Reply on news.ycombinator.com