Observation:

OpenAL kind of sucks.

>>40
Listen here, jerkface.

>>38
In some cases the distributions even try to prevent users from removing ALSA and installing OSS by keeping ALSA’s mixer interface busy (the Gnome/GTK mixer appled is immediately relaunched if it gets killed).
Wow.

>>37
No. I don't care about his software. I don't want to use it, and bitching to him has the precedent of going unappreciated.

>>38
It's not just distributions that ship it, it's projects that use it as a backend. You'll probably find it in just about every distribution's repo for that reason.

Ubuntu is not reasonable.
Correct.

OSS
Not ported to Windows. See, this really isn't about backend sound systems, it's about that necessary evil in Linux: the intermediary sound system. Most of these do double-duty so it's easy to be confused about what you want and what you need. OpenAL here is acting as the intermediary, the API. The scenario is easy enough to describe:

1. Assume the user has working sound.
2. Talk to that using a suitable intermediary.

I don't care what drives #1. If the user chooses a terrible backend that's their problem. OSS is unsuitable for use in #2, which is my concern: abstraction. The conveniences you list are what I try to avoid.

>>40
Hahaha. Go argue with the PulseAudio users.

>>43
"Try to avoid"? I could understand that it may be of little consequence to you that an API is simple and well-documented, but to avoid it for that is nonsensical. Much less that you would want to avoid low latency or good quality mixing different streams with different sample rates.

Oh and did I mention that each application gets a separate volume control? Kickass feature.

And OpenAL works fine with OSS, I'm not sure how you say it's unsuitable. Regardless, I brought up OSS in regards to all the silly comments about PulseAudio, not necessarily because it had anything to do with >>1.

>>44
I was fairly explicit about not caring about the backend. As I said, I don't care what backend the user supplies: it can be OSS or PulseAudio or ALSA or anything else that will talk to the hardware. Using OSS in my application as my intermediary is what is unsuitable.

By the way: I get latencies less than 2ms using ALSA as a backend, with interpolation. ALSA is pretty gross as far as interfaces go, however as an abstraction it is much better than a Unix character device.

>>45
2ms is pretty horrendous if you're trying to play a softsynth.

>>46
It's fantastic for a softsynth.

>>47
If you like the Haas effect.

>>48
I'm going to bump this everyone will see this wonderfully relevant post you just made.

CoreAudio fuck yeah!