running speech-dispatcher-0.7.1 in system mode

[Steve Holmes]

OK I found the answer to one of my questions.  I apparently forgot to
uncomment the last line in speechd.conf to disable autospawning.  OK,
I did that and now I get the following error whenever I run spd-say.
Script started on Fri 08 Oct 2010 09:46:50 PM MST
steve at lnx3:~$ spd-say "hello party"
Failed to connect to Speech Dispatcher:
Error: Can't connect to unix socket 
/home/steve/.speech-dispatcher/speechd.sock: No such file or directory. 
Autospawn: Autospawn failed. Speech Dispatcher refused to start with error 
code, stating this as a reason: Autospawn requested but disabled in 
configuration

steve at lnx3:~$ exit
exit

Script done on Fri 08 Oct 2010 09:47:15 PM MST

So my question is now, why do the clients keep trying to spawn a new
instance instead of calling the established server? I looked at
speech-dispatcher.log and it shows that it is definitely using
inet_socket and is listening on the default port.

Any more ideas?

On Fri, Oct 08, 2010 at 09:23:20PM -0700, Steve Holmes wrote:
> I thought I would give system mode a try with the latest git pull of
> speech dispatcher.  I'm pushing for this now because this afternoon, I
> ran into a conflict situation where my ALSA device was locked out and
> I had to kill all instances of speechd before I could play any
> sounds.  So as I said before, Ithought I'd start up speechd as a
> system service using inet_socket for communication method and alsa for
> output method because I don't do pulse!
> 
> My problem is every time I run spd-say or start up a gnome session, a
> new instance of speech-dispatcher gets spawned and when I do a 'ps ax
> |grep speech-dispatcher' to see what is going on, I see additional
> parameters showing up in the spawned version including such things as
> --spawn and so on.  In my /etc/speech-dispatcher/speechd.conf, I
> specifically uncommentted the DisableAutospawn and the original
> instance which I started as a daemon remains running but none of the
> client requests make it to this instance; they keep spawning new ones.
> 
> Also when I started speech-dispatcher centrally, I tried both
> specifying inet_socket from inside the .conf file and from the new
> command line optionss (thanks) but the logs keep getting written to my
> home (/root/.speech-dispatcher) instead of what I thought to be
> /var/log/speech-dispatcher.  I do have that directory here but it is
> not being used.
> 
> Any ideas why I can't get the clients to use the central version
> instead of spawning new instances?  Oh, I do not have any config files
> stored in the local .speech-dispatcher directories either.  In fact, I
> delete the whole dir before running a test just to be sure.