1 |
On 05/20/2010 10:25 PM, Canek Peláez Valdés wrote: |
2 |
> On Thu, May 20, 2010 at 2:04 PM, Nikos Chantziaras<realnc@×××××.de> wrote: |
3 |
> [snip] |
4 |
>> What doesn't work is PulseAudio, actually. Too many problems with it. Pulse |
5 |
>> is simply broken by design; it's too far from the kernel to be any good. |
6 |
> |
7 |
> If I may use (most of) your words: "Well, it works here. It's been |
8 |
> rock-solid through months." And with various use-cases, if I may add. |
9 |
> |
10 |
> Can you elaborate why the audio architecture has to be close to the |
11 |
> kernel? The part that talks to the hardware obviously has to, but why |
12 |
> the part that handles the features, the mixes, the virtual devices? |
13 |
|
14 |
Because as soon as you disable ALSA dmix and/or Pulse, suddenly you get |
15 |
acceptable sound latency. |
16 |
|
17 |
With OSS4, which has in-kernel mixing, it doesn't matter if you enable |
18 |
the mixer or disable it; sound always has acceptable latency. |
19 |
|
20 |
Thus, I can only conclude that mixing has to happen in-kernel. But I |
21 |
base this only on the ALSA/Pulse vs OSS4 comparison. It could also be |
22 |
that the user-space implementation of ALSA just sucks. But that's hard |
23 |
to believe, since if that were the case they would have fixed it several |
24 |
years ago already. |
25 |
|
26 |
|
27 |
> I'm under the impression (correct me if I'm wrong) that it was one of |
28 |
> the major reasons to leave OSS4 outside the upstream kernel; too many |
29 |
> stuff in there that belongs in user space. It sounds reasonable to me. |
30 |
|
31 |
It sounds reasonable from a designer's point of view. But a system is |
32 |
useless if it's only designed good but doesn't actually work in a |
33 |
satisfactory manner. |
34 |
|
35 |
|
36 |
> Specially when PulseAudio just works, for me and many more. |
37 |
|
38 |
Sorry, that just pretentious of you here. PulseAudio is the most flamed |
39 |
at, hated, sound-related software around. And this is because it does |
40 |
*not* work for many, many users, and the first thing they try to do is |
41 |
find out how to disable the thing. |
42 |
|
43 |
|
44 |
>> ALSA can't switch to Bluetooth either. You could use PulseAudio with OSS4 |
45 |
>> instead of with ALSA though, but this is not officially supported. |
46 |
> |
47 |
> Indeed it's not supported, because it's (using your words again) |
48 |
> "broken by design" by trying to do too many things inside the kernel |
49 |
> that belong in user space. That's my understanding at least; please |
50 |
> correct me if you believe I'm mistaken. |
51 |
|
52 |
You're mistaken in that a mixer should be in the same boat as network |
53 |
streaming, bluetooth, etc, etc. I believe the *mixer* should be |
54 |
in-kernel. Everything else doesn't need to be. PulseAudio's extreme |
55 |
latency problems (which even upstream admits can't be fixed easily) stem |
56 |
from that. |