Re: [orca-list] Getting Environments to Orca
- From: Michael Whapples <mwhapples aim com>
- To: Orca-list <Orca-list gnome org>
- Subject: Re: [orca-list] Getting Environments to Orca
- Date: Wed, 13 Oct 2010 07:26:12 +0100
Have you tried using the .profile or .bashrc files in the GDM home
directory (normally /var/lib/gdm/)? I can't precisely remember which of
these files is sourced when but I would have thought one of these would
be the correct place to set environment variables.
I am a little surprised you can't configure this setting in
speech-dispatcher's configuration files.
On -10/01/37 20:59, Steve Holmes wrote:
I'm looking for the most proper way to get environment variables to
pass through gdm to Orca so Orca can refer to them. In greater detail
what I'm trying to do is set up speech dispatcher to run in system
wide service mode. To do this, I have to set in the environment a
SPEECH_ADDRESS and any of the clients wishing to use speech dispatcher
would then use the contents of this environment variable to
communicate with it. So Orca needs this info to work in this mode.
At present, if one starts gnome from a normal user session with the
startx command, this all works fine because Orca seems to inherit the
environment from that normal user. However, if gdm does this, the
info does not make it to Orca.
With my recent studies of this, I find that the gdm startup script in
/usr/sbin/gdm does inherit the environment and even pulls such
settings from the /etc/profile. But I have no idea what the
gdm-binary is doing but when the Orca launch script in /usr/bin/orca
gets going, that information is not present. In a local hack on my
machine, I inserted this variable assignment into the script and now
Orca can work with speech dispatcher in system mode. The advantage of
system mode is you then only have a single instance of speech
dispatcher running at any given time.
Some distros use that dreaded pulse audio system which requires
individual instances of speech dispatcher to run in each user session;
in those cases, this system service approach would not be taken.
What I'm asking is then what are the best steps to take to enable Orca
to see this environment variable and of course, when not present, then
launch speech dispatcher like it nows and it will autospawn itself.
Any good ideas out there?
] [Thread Prev