[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
RE: 0.10.0 release candidate 3
From: |
Mgr. Janusz Chmiel |
Subject: |
RE: 0.10.0 release candidate 3 |
Date: |
Sun, 26 Apr 2020 22:22:25 +0200 |
Very well done Mr Tibault.
Thank you very much that even if you introduced so many professionally
programmed new Speech-dispatcher parameters, that you have not decided to
use shared memory and dynamic shared memory allocations and similar shared
memory based techniques.
So by other words, if I will use original Pulseaudio sound server which will
run by using Proot on Android device, I should set PUlseaudio buffer in
Simple protocol player also to 10 MS to make The best speech responsiveness.
Because you are excellent C programmer. And deeply experienced professional.
How complex would be to release Speech-dispatcher release as a Chrystmass
present only once for Year also for Termux?
Because I think, that if somebody would recode Espeak and portaudio.h so it
would be compatible with Android on Termux with its integrated Pulseaudio
sound server, that speech responsiveness of Espeak would be much more
faster.
Do you think, that ArchLinux packagers will release this new version also
for ARM64 architecture soon? Or I would had to compile it myself?
I Am very curious, why ArchLinux packagers have decided to not compile
Speech-dispatcher with Libao library support. That only Espeak-ng is
currently supported.
As A contrast, Debian ARM64 stable repo contain The Speech-dispatcher which
have been compiled with Libao support and also standard Espeak driver is
supported.
I Am sad, that I do not know, how to get Mbrola-espeak driver and Czech
database to get it work with ArchLinux on Android by using Proot and Termux.
I only want to find The best responsiveness between keypress and Espeak or
other speech engine speech.
I hope, that somebody here will have A good will and that somebody will give
Me some advices.
With warmest regards and with The deepest possible appreciation.
Janusz Chmiel