Re: [orca-list] No speech with Ubuntu Gutsy beta 7.10

[Tomas Cerha <cerha brailcom org>]

Luke Yelavich wrote:
> Only that gnome-speech has so far been adequit for use for speech output under GNOME for most 
> users. There is also the issue of supportability, as in, whether the core Ubuntu developers are 
> willing to support the package, given that its well maintained upstream, and that any security 
> issues are solved quickly.

According to the feedback from users, it seems that Speech Dispatcher
can solve some real-world issues for many of them.  Thus I suppose it
would make sense to think about its inclusion.  Are there any reasons to
doubt that any of given requirements would not be satisfied for Speech
Dispatcher compared to the other components which are included?

> I understand that a client has control over various settings, such as the voice used, rate, 
> pitch, etc. However, so far as I have seen, it is not possible for example, for individual users 
> to set which sound card their synthesizer of choice uses. To do this, one must have 
> administrative rights, and must restart the service to accomplish this. Granted this may not 
> always be desired in multi-user environments, but a user may wish to use a plugged in USB sound 
> card for example, and as things currently stand, this would not be possible.

Well, I'm not quite sure this is a real-world example, but if you have a
use case for this, I'll be glad to discuss it in more detail.  From the
top of my head, I can think of two solutions:

 1. Don't run speech-dispatcher as a system wide service, but run it as
    user within the session.  This is perfectly possible without any
    changes in Speech Dispatcher itself.  You would only need to choose
    a port number and pass this number to clients (for example through
    an environment variable).

 2. Add sound card selection to the SSIP protocol and let the client
    select the prefered audio device.  I'm not quite sure it belongs
    there, but we can discuss this requirement as well as any other
    requirements you might have.

I'm curious how this would be accomplished with Gnome Speech.

Looking forward to your feedback,

Tomas Cerha