Various questions

[nolan]

1. I've needed to switch back to pulse because, under ALSA, I didn't
seem to get speech back when resuming from suspend. In my quest to find
the perfect setup, I tried the test version of pulse from Luke's PPA and
these did not work well with SD. I've since restored to the Jaunty pulse
packages, but I seem to have lost the ~/.pulse_a11y_nostart behavior, as
removing the file for a test still gave me a pulseaudio process. Where
is this behavior defined? A quick grep through /etc/X11, /etc/xdg
and /etc/pulseaudio didn't reveal the string pulse_a11y_nostart, so I'm
wondering if it was defined in a config file that got trashed in my
upgrade to 0.15?

2. I have an /etc/pulseaudio/a11y_nostart. Can I remove this? Is it a
remnant from my beta install?

3. While I did downgrade pulse from Luke's PPA, I kept ALSA 1.0.19. With
my user in the pulse-rt group and with this ALSA, if there's a latency
difference then it's miniscule, and I actually wondered if pulse was
running until I confirmed with paman. Hopefully, with the latency
minimized, there'll be research done into the SD crashes, but that's for
another list. Just wanted to report that ALSA 1.0.19+pulse+my user in
pulse-rt was snappier than before, which gives me hope that eventually
we'll just have a single branch of development and focused effort and
a11y via pulse won't be such a black art. :)

4. Is theree any way that this SD bug can be looked at, and its patch
gotten past the freeze?

https://bugs.launchpad.net/ubuntu/+source/speech-dispatcher/+bug/314185

Basically, sd-conf has an incorrectly-configured sound path, so you
can't test the ALSA/Pulse backends. I'd reinstalled the SD packages
after I'd done a bit of hacking and had to manually patch this bug to do
the testing I needed on an inaccessible system. It's literally a
single-character change, so hopefully an exception can be made, but I
don't know Ubuntu policy well enough.

Thanks.