> Bug 1813 is a really old bug for something else (fixed in 2003). Did you
>> forget a digit?
>
That would be http://bugs.winehq.org/show_bug.cgi?id=18133.
>
>- Reece
Hi. Yes, that's the bug.
Here's everything I know, and probably more than you want to know, but I don't
know what to leave out.
My system is on the bleeding edge of everything, which can be a problem when it
comes time to diagnose.
I have been using Ubuntu Jaunty Studio since its first beta, and it's now
fairly mature. I compile every wine git and test it, using Dragon
NaturallySpeaking. Usually everything works as well or better than it has.
In addition to that, I compile my own alsa-driver because I use a Creative
SoundBlaster X-fi soundcard. They are a one-time install and I used the stable
daily build. They had been working well for weeks, and I had not updated them
in at least a day or two.
About 10 days ago, when I did my daily wine build and installed DNS, the
training froze. My first thought was a wine regression, but I was busy and
didn't test for a few days. Then I did a couple of regression tests that seemed
to succeed, but really the just marked the first change I tagged as the
problem. I tested back further than the problem appeared. Still no luck.
Solution? That it's not wine's problem, it is something else that has been
introduced.
Then I tried using a new alsa-driver, and a new alsa-libs with it. Nothing.
Then I tested various configurations of pulseaudio, purged and not-purged. No
changes.
Then I reinstalled the whole Ubuntu-Studio system.
In the meantime, Ubuntu-Studio had switched to using the Real-Time kernel as a
default. The kernel froze before wine had a chance to.
Back to the drawing board. I downloaded the generic kernel. That seemed to work
for USB audio only. I tried compiling alsa-drivers, and installing them, and
then USB stopped receiving sound.
Then someone on my ubuntu-studio users list compiled a generic (non-ubuntu)
realtime kernel.
I installed it in the crudest way possible, by double-clicking on the DEB.
Success, but for the USB soundcard only.
The alsadriver settings that freeze during training are as follows (and they
are the ones that used to work):
wave-out devices: dmix:Generic (there were three of these)
wave-in devices: dsnoop:Generic (ditto)
My working mixer device was HD-Audio-Generic.
(I had three cards, the other mixers were HDA Intel and iMic USB audio system)
Now, to get the program to work at all without freezing, I re-install the
kernel. Then only the USB works properly. The settings I see are as follows.
wave-out devices: ALC660-VD Analog
USB Audio
wave-in devices: (same)
So, what do you make of all this? Are dmix and dsnoop being bypassed?
Susan
