Re: [orca-list] is it possible to boot latest Ubuntu by using SDC cart?

[Didier Spaier <didier slint fr>]

Hello,

thanks for the praise, although not deserved.

I just made very small contributions to Speech-Dispatcher, and this reminds me that enhancing or even 
rewriting spd-conf to allow setting directly a voice among those available (as can be done in Orca) is still 
in my TODO list.

About speech responsiveness: so many software are involved that such issues are very hard to investigate, to 
name few:
_the AT* software
_Speech-Dispatcher
_the TTS in use (espeak, RHVoice, flite, Vowin, Voxygen, whatever)
_Orca
_the sound server (PukseAudio, Alsa, others)

Generally speaking though it is true that at least in some contexts software written in Python or other 
interpreted languages are slower than in compiled ones.

But I bet that you won't easily find volunteers to re-write Orca in C++ or any other language <smile>

Best,

Didier

On 05/05/2019 14:07, Mgr. Janusz Chmiel via orca-list wrote:
> Oh thank you very very much.
> Yours Linux knowledge is really on professional level and You have given Me very interesting group of 
> advices.
> Very well done.
>
> I AM very glad that you are The member of this mailing list. And MR Spaier, thank you for yours complex 
> programmers work on Speech-dispatcher and its modules. Latest Ubuntu speech responsiveness is very probably 
> The best made while using Speech-dispatcher speech server.
> To be honest, The best speech responsiveness have been while working with Vinux 2.0, it was Debian 6 and 2. 
> x kernel, and ALSA sound server with no relay on Pulseaudio. But this positive aspect had other 
> disadvantage.
> Speech-dispatcher have not been so stable so it randomly crashed.
> Today, I Am only experiencing very small Espeak sound quality problems while using left or right arrow key 
> on The editable field while working with Firefox.
> Some editable fields while using Orca have better some less better reactivity. But I have found out, that 
> there is other complex problem.
> When using Brltty, result is just now on The display. But when using Speech, result get some Milliseconds 
> later.
> It is because all sound devices will stop to sound sound signal after some Milliseconds of inactivity.
> The solution would be to stream very very lov loudness sound with 6 KH/Z 8 Bits at The same sound channel, 
> which Speech-dispatcher and Espeak C library uses. This background noise would prevent sound chip to go in 
> The sleep. So chip would not be initialized and reinitialized when ever new speech chunks would be send to 
> it.
> Next much more complex problem is may be caused by Python machine. Orca is written in Python, module which 
> communicate with speech-dispatcher is also written in Python itself. I Am wondering, if Python allow some 
> threading support or no. Because speech is very specific programmers think. Chort sound are being streamed 
> by using sound output device.
> _______________________________________________
> orca-list mailing list
> orca-list gnome org
> https://mail.gnome.org/mailman/listinfo/orca-list
> Orca wiki: https://wiki.gnome.org/Projects/Orca
> Orca documentation: https://help.gnome.org/users/orca/stable/
> GNOME Universal Access guide: https://help.gnome.org/users/gnome-help/stable/a11y.html
> Log bugs and feature requests at http://bugzilla.gnome.org