[manjaro-dev] catch 22, should I run pulse audio in system mode?

[kendell clark]

hi all
I seem to have run into a catch 22 that I'm not sure how to solve. As
all of you might or might not know, I've been trying to replace speakup,
the in kernel screen reader for text console, with brltty, which is the
software that drives braille displays, since it can also speak. I've
been having a lot of trouble with it, and I managed to track down the
problem after a lot of help on the brltty mailing list. The problem
seems to be pulse audio. When it attempts to speak, pulse audio isn't
letting the sound through, just like speakup. The brltty people want me
to run pulse audio in system mode, which will allow the sound to come
through. This will work, but I've heard all sorts of dire security
warnings about doing so. I don't think there are any serious risks
involved, but I'm not positive. I do know that running it in system mode
means there is just one copy of pulse audio running for all users, and
applications connect to that. This means that different users logged
onto the same system won't have sounds tripping all over each other. I'm
asking you guys if I should try to configure pulse audio to do this. The
brltty people seem to think it's the only way I'm going to be able to
fix the problem with their speech, since the pulse audio people are
dodging me. It's not our bug, have the brltty people fix it, etc. Same
old shit that's kept me from being able to make speakup work as seemless
as it should. I'm going to forward this to the Sonar list as well to see
what they think. What do you guys think I should do? There's a third
option, that of jupider. It's a userspace screen reader for the text
console developed by tyler spivey and some other person. I don't know if
it's still actively developed, and even if so, if it's packaged for
arch. If not, I'm going to need help getting it packaged up. What do you
guys think? Should I run pulse audio in system mode?
Phil, do you have any ideas? I need help from some expert linux guys who
have been around longer than I have before I change something like this.
I have to tell you guys. I'm about fed up with trying to get pulse audio
devs to actually understand. When I can get ahold of them, which is not
often, they dodge me. It's not our issue, go ask "insert software hear".
I've had this with both speakup and now brltty. They won't just fix the
problem, they have to put me off. Behavior like this from an open source
project drives me insane. Especially since you guys go above and beyond
what I'd ask of anyone to help out our sonar distribution. Then I get
the opposite of that from a major part of modern linux. The worst part
of all of this is that pulse audio isn't going away. If anything, more
and more things are using it, and accessibility software has to be able
to keep up. But when the devs of that software have one hand tied behind
their backs because the pulse people won't cooperate .... ok, end of
rant. What should I do?
 Thanks
Kendell Clark