Re: [orca-list] one major question about speach still can't be answered.

[Tomas Cerha <cerha brailcom org>]

Hermann wrote:
> I can confirm the crash, but in my case not Orca causes it, but SD.
> /var/log/speech-dispatcher/espeak.log
> contains lines like
> underrun! xxxxx ms etc.

And does it really crash?  Doesn't it only stop speaking?  Is there
something else in any of the speech dispatcher log files?  Does the
speech-dispatcher process still run after that?  Does the sd_generic
process run?

> The most annoying thing is that I can't restart SD, except with rebooting my 
> machine.
> /etc/init.d/speech-dispatcher resart
> reports that SD has been restarted, but
> spd-say "hello"
> does not say a word, but produces no error message.

Then I suspect there might be some problem in the underlying sound
system.  Do other audio application work at this point?  What do you
have in your /etc/speech-dispatcher/modules/espeak-generic.conf for
GenericExecuteSynth?

> So repeated rebooting of the computer seems not only a Windows typical 
> necessity.

Not really, you just need to know what needs to be restarted...

> The crash in Orc happens often when an E-Mail arrives in Evolution, and a 
> sound is played. I think SD uses Alsa, that handles multiple sound events; 
> does it really?

Yes, it should, unless there is some problem in your alsa setup.  For
example I remember some people had problems because of some old
.asoundrc file in their home directory.

> As Halim reported the fix works, and I don't have the issue of not properly 
> closing the dialog. Which Python errors do appear?

I don't understand the question :-(

> Note: You must configure the speech settings (volume, rate pitch) in the SD 
> config files:
> /etc/speech-dispatcher/speechd.conf
> and
> /etc/speech-dispatcher/modules/espeak-generic.conf

Exactly!

Thanks for your good feedback.  Keep going.

Best regards

Tomas