running speech-dispatcher-0.7.1 in system mode

[Steve Holmes]

I just tried from a cold boot and inserted the proper environment
variable into a profile that gets read by all programs that startup on
my Arch Linux system.  This gets inherited nicely by normal users at
login and again, startx works fine but gdm for some reason does not
inherit this variable when it runs.  I think gdm is the user but it
must run something else that does not inherit these environment
variables.

Has anyone else gotten speech dispatcher to work under gdm but have it
running as a service with inet_sockets? I really don't wanna use
gnome-speech and I think it is gnome-speech that may have been causing
some of the instabilities when it came to starting Orca Preferences
and then getting out of them and losing speech.  It seems to be much
more stable when using speech dispatcher.

On Sat, Oct 09, 2010 at 12:04:36AM -0700, Steve Holmes wrote:
> On Sat, Oct 09, 2010 at 08:25:09AM +0200, Halim Sahin wrote:
> > Hi,
> > Well under gentoo I have a /etc/init.d/xdm
> > init-skript.
> > Exporting speechd_adress needs to be placed there to get gdm work with
> > speechd.
> 
> I've been trying that in my startup script for gdm but no go so far.
> I stuck the following, 
> export SPEECHD_ADDRESS=inet_socket:localhost:6560
> just before the line that starts gdm 
> /usr/sbin/gdm &
> and when I run it from a normal root shell, speechd doesn't get
> started.  Orca brings up gnome-speech instead.  I can do this from a
> normal user login and use the startx command with no problem.  About
> the only thing I got left would be to add the export line back into
> the gdm startup script and try it from a cold and have the system
> launch it before I login; maybe that might make a difference.  But it
> sure doesn't work when trying it interactively.