really no idea? Re: sd_generic problem please help!

[Halim Sahin]

Hello,
On So, Nov 22, 2009 at 07:20:45 +0100, Bohdan R. Rau wrote:
> On Sat, 21 Nov 2009 15:08:01 +0100, Halim Sahin <halim.sahin at t-online.de>
> wrote:
> > On Do, Nov 19, 2009 at 06:06:40 +0100, Halim Sahin wrote:
> >> Hello,
> >> 
> >> sd_generic starts after every ". " a new player proccess.
> >> Can I easyly disable this?
> >> The problem is that the players take a while to startup.
> 
> I think this is problem with player, not with sd_generic module.

ACK maybe.
The question should be:
Why does speech-dispatcher split the output after every dot?
I wouldn't expect this behaviour.
SD should profide the whole data through the generic module.

> Is it possible to run your synthesizer in server mode (like Ivona)?

Afaik espeak doesn't support this.

> If not - is it possible to create wrapper (daemoniser)?

Maybe but this whould be synth specific.
It would be better to fix this in sd.

> What syntesizer you use?
Espeak.

The reason not to use alsa and the native espeak module is the
following:
the native alsa driver crashs very often on my new computer.
This machine doesn't have a soundcard with hardwaremixing.
I don't want to restart speech-dispatcher every 15 minutes :-(.
The alsa code seems to have some critical problems:
sd_espeak segfaults.
I tried to debug this but wasn't able to produce a core file for
sd_espeak.
The second problem is that all distros are now using pulseaudio.
Unfortunately pa is a dependency for gnome if the user wants system
sounds.
So the only usable working solution (currently) is to use sd_generic and
pipe the output through paplay.
This works fine on my machine (except the dotproblem).

Regards
Halim