speechd-discuss
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: speech-dispatcher and userspace chroot environment on Android


From: Nolan Darilek
Subject: Re: speech-dispatcher and userspace chroot environment on Android
Date: Wed, 27 May 2020 09:44:38 -0500
User-agent: Mozilla/5.0 (X11; Linux x86_64; rv:68.0) Gecko/20100101 Thunderbird/68.8.0

Do you have any setup instructions for this?


To be more specific, I'm sure I can figure out the Linux-specific bits, but presumably there's some point where the Linux environment isn't accessible because Android, so you're installing the accessibility stack. How do you use Termux then, or is it finally accessible under TalkBack?


And, to be clear, you're successfully running Orca and a graphical desktop?


My phone actually has more RAM and CPU than the 10-year-old laptop I haul around on occasion, so I'd be very interested in getting this working. Thanks!

On 5/27/20 6:30 AM, Mgr. Janusz Chmiel wrote:
I would like to inform every one of us about very positive findings related
to Speech-dispatcher and its compatibility with various speech engines. When
running it on userspace chrooted environment on Android devices.
Thanks to Pulseaudio for Termux, which provide robust and stable sound
server controllable by using Opensless protocol it is possible to perfectly
use speech-dispatcher with Espeak and Espeak-ng driver. Mbrola is also fully
compatible.
I will now test Speech-dispatcher with Festival. But I have one problem.
Festival must be run as A server. Is it possible to run it as A normal app
or is it really necessary to run it as A Linux service? Because systemd do
not function in chroot mode on ArchLinux at all.
I Am wondering, if would be possible to run Festival as A server but not
like a Linux service.

If it would be possible, It would be possible to work with Espeak, Mbrola
and even with Festival from Termux and Linux distribution executed thanks to
Proot Termux package.

Because it is even possible to use Evolution mail, Thunderbird, even
Seamonkey with ArchLinux AArch64, it brings a new possibilityes for all
visually impaired users. So  they can have benefit from using  little and
light veight Android devices. You can combine ANdroid with Linux. Because
Proot allow folder sharing, you can define shared folder to exchange data
between Linux distro and Android OS and even with The computer if you will
assign folder in The accessible file storage area for computers.

I have more than positive experience while using Speech-dispatcher and
Mbrola. No speech output quality issues, perfect stability, reliable char by
char browsing support not only on several editable fields. But really on all
situations even while browsing on Firefox or Seamonkey.

You are not being forced to use shared memory support native .so C library
for Linux distribution to use Speech-dispatcher And Espeak or Mbrola.
In fact shared memory support on Android have big senses only for some apps,
because complex shared memory allocations such as X11vnc is not compatible
with Androidshmem.so.
So if somebody is being afraid about security, no problem. Shared memory is
not necessary even for Mate desktop, because Mate developers have coded
their environment even for situations, where shared memory will not be used
at all.
Because security is big problem today thanks to powerful hackers all around
The world, I will be intensively debate with some professionals about
encrypting installed Linux distro on Android. It will be necessary to
develop compromise solution related to performance, internal chip
overburning problems. Because some Android devices have allready enabled
their internal data encryption so encrypting some data is even impossible.
But because Termux is Android app, Android build in security mechanishms are
not affected. Every one can encrypt his device, protect it by using password
or A fingerprint. So I think, that there is really not dangerous issue which
would prevent users to use Arch Linux on Android.
This dream is only possible thanks to Orca and Speech-dispatcher.
Because none of those apps enforces shared memory. Fortunately not only for
Me.





reply via email to

[Prev in Thread] Current Thread [Next in Thread]