Hey, that's not a bad idea Lawrence. E-speak up could probably still
work with Dos emu under Arch. It doesn't work on the newer version of
Slint for some reason. It would be ideal to have a boot disk that would
come up talking like Linux does. Maybe your thought of using Linux could
work. Someone who knows programming could set it up like you mentioned
and if possible have e-speak up be the screen reader under dos emu if
that's possible. I don't know a thing about programming on that level so
I don't know how much work it would take. There may even be a newer dos
emu out there I'm unaware of that would work well in the newer Linux.
Thanks for your idea.
On 10/15/2024 10:13 AM, Lawrence Perez via Freedos-user wrote:
Hello,
I'm not a developer, but I believe there may be some complications with getting
a software speech synthesizer to work. Most DOS systems are designed to be
16-bit, whereas modern software speech synthesizers typically run on 32-bit or
64-bit, or even, in the case of Raspberry Pi computers, an ARM CPU. I am
unfamiliar with the intricacies of compiling software, but I imagine it would
be challenging due to how ESpeak was designed to be used with more modern
hardware. The screen reader wouldn't necessarily be a challenge, but the
challenge lies in getting a working speech synthesizer to function due to the
older architecture DOS was designed for.
I wonder if it would be possible to create some type of Arch Linux/Debian
minimal install image and design it in such a way that it only works with
FreeDOS and nothing else? My idea is this: when Debian starts, immediately load
FreeDOS in a virtual environment, and then use virtual serial cables to send
data to a program that then sends the data to ESpeak to make it talk? This
setup may still use virtualization, but this way we could use existing DOS
open-source screen readers and make them talk through ESpeak. Espeakup could be
used as a backup screen reader in case the user needs to troubleshoot.
One way users could get files onto the virtual DOS would be to use a special
DOS directory, and mounted drives could appear under `/dos/mnt`. This would
make it easier to import files from other drives to the new DOS environment.
What are your thoughts on this? For those seeking to get DOS running out of the
box with accessibility, a Debian/Arch minimal install would be a great
solution, in my opinion.
Hope this helps!
Sincerely,
Lawrence
---
Let me know if you need any more assistance!
Sent from my iPhone
On Oct 15, 2024, at 9:16 AM, Karen Lewellen via Freedos-user
<freedos-user@lists.sourceforge.net> wrote:
Other factor speaking personally with espeak is the largely poor speech
quality.
There is an effort to get copyright clearance for one of the dectalk software
speech sources.
I have no idea what the pi would provide speech wise, nor the cost factor.
Still, there are dectalk USB units available as well.
Lastly, Joseph got clearance for aSAP to be built into freedos. Once
installed, you can change the synthesizer output to a far wider variety of
options, including a generic voice intended for any speech not given a
specific driver.
At the end of the day, choice is the key element here for such a distribution.
On Tue, 15 Oct 2024, Mike Coulombe via Freedos-user wrote:
Thanks! I'll have a look at this.
On 10/15/2024 6:36 AM, Mateusz Viste via Freedos-user wrote:
On 14/10/2024 17:37, Mike Coulombe via Freedos-user wrote:
Hi, That distro sounds interesting, but I need to run free dos on > different computers.
If it had speech accessibility free dos would be > very useful to blind people world wide
considering all the free dos > programs you can get and the amount of old computers many
of can be had > for free or next to nothing. Do you think it would be worth asking about
> this on the developer list or is that only for developers?
Running espeak as a TSR might be technically possible, but it would
certainly be bound to many limitations. I doubt anyone will want to invest
time into this, given that the result is likely to be of a limited
usefulness for any practical purposes.
Having a DOS distribution with a preinstalled screen reader that outputs
speech commands to the computer's serial port is a much more realistic
goal. I implemented this in SvarDOS today, so now the SvarDOS system is
available also as a version that "talks". It does require a Braille 'n
Speak synth to be connected, though:
http://svardos.org
I tested it on VirtualBox with an emulated BNS (emubns), and it seems to
work, assuming one knows how to use provox hotkeys to "explore" the
screen.
Hardware PCs require a hardware BNS, obviously, but a cheap solution could
be to run emubns on a Raspberry Pi connected to the PC with a
USB-to-serial cable.
Mateusz
_______________________________________________
Freedos-user mailing list
Freedos-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/freedos-user
_______________________________________________
Freedos-user mailing list
Freedos-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/freedos-user
_______________________________________________
Freedos-user mailing list
Freedos-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/freedos-user
_______________________________________________
Freedos-user mailing list
Freedos-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/freedos-user
_______________________________________________
Freedos-user mailing list
Freedos-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/freedos-user
