Re: [orca-list] Speech-dispatcher won't start after upgrade

[Michael Whapples <mwhapples aim com>]

Not knowing ubuntu well enough, I don't know what directories allready
exist and which ones get created on boot. I would suggest finding out
what group owns the directory which will contain the newly created parts
for the pid file (eg. if /var/run/speech-dispatcher exists after booting
but the pid file doesn't, so find the group
for /var/run/speech-dispatcher) and add the speech-dispatcher user to
the group and make sure the group can write to the directory (eg. chmod
g+rw /var/run/speech-dispatcher).

Does this help? Does the speech-dispatcher package included with ubuntu
have the espeak specific module or the espeak-generic one? Since moving
to speech-dispatcher 0.6.4 (and recently higher) stability of
speech-dispatcher seems to have improved (the crashes I got before
seemed to be related to the espeak-generic module returning incorrect
bytes (from what I could make out in the log files).

From
Michael Whapples
On Sun, 2007-12-02 at 11:49 +0100, Hermann wrote:
> Hi,
> because SD sometimes crashes without any reason, I upgraded from
> 0.6.3-testing to 0.6.4.
> Because this package wasn't in the Ubuntu repository, I had to do this
> manually by downloading the packages related to SD from the Debian pool.
> After installing, /var/run/speech-dispatcher/speech-dispatcher.pid isn't
> created automatically anymore.
> I have to create it manually after every start. Otherwise I get the
> message that it cannot be created because of wrong permission.
> The startup script in /etc/init.d is unchanged.
> There's a user "speech-dispatcher" that belongs also to the group
> "audio".
> So what's going wrong? Why does an upgrade screw up all the SD settings?
> In the meantime I had to switch to Gnome-Speech again, which would be
> sufficient, if that annoying OSS/Alsa thing can be solved in the near
> future.
> Hermann