2009/6/16 Jerome Leclanche adys.wh@gmail.com:
On Tue, Jun 16, 2009 at 2:14 PM, Roderick Colenbranderthunderbird2k@gmail.com wrote:
I think a lot of issues can already be fixed without requiring a new driver.
One of the problems is, every time one of these issues is brought up, we hear people complaining "But Pulseaudio should be a drop-in replacement! We shouldn't have to change our code". There is a major (and irritating) circle of blaming that comes up every time the subject of PA is brought up. If anything should be fixed first, it's the attitude people have with it.
I'm not going to argue with designs; I'm no sound developer. I don't like PA, it's way too unstable still; the only reason I use it is because of a pair of USB headphones which don't work properly under ALSA. If issues can be fixed without requiring a new driver, they definitely should be. What is irritating is the "Just don't use Pulse" attitude some people have. Both most recently Scott Richie and Arthur Taylor, and also other earlier authors, have done really great efforts on what is perceived as a "problem with wine" -- *nevermind* whose fault it actually is -- and none of this should be shrugged off with a "What's the point? This won't ever be accepted". Have some respect.
FWIW,there is a simple PulseAudio API that exposes an OSS-like open/read/write/close API to pulseaudio.
The previous attempt, IIRC, was based on the more complicated pulseaudio API and based on the winealsa code. It should (in theory) be possible to create a pulseaudio driver using this simpler pulseaudio API and the wineoss driver.
- Reece