Greetings everyone,
I do respect the nature of this list.
However, none of you are medical professionals...nor is the below correct.
I experience an aquired brain injury. My dectalk set up is not from 1984,
what an odd idea.
The dectalk voices have been updated long after then...ask Stephen
Hawking
Further, I do not feel I have to justify information documented from
direct exposure, years of medical testing, and the advice of practicing
doctors responsible for my care...so I am sharing the below with the list.
On Wed, 21 Oct 2026, Charles wrote:
Dear Karen,
I work at Canonical on the Linux Desktop. I was very interested to read about
your computing setup.
I quote below you saying Linux speech would put you in the Hospital. Can you
elaborate more on that? Were you even able to try a Linux speech system?
I read elsewhere you run a DECtalk system. It amazes me that the speech
synthesis from 1984 sounds better than what Linux distros in 2026 ship by
default, if at all.
I also work on a speech-to-text system for Ubuntu, which I hope will be a
useful accessibility feature going forward. However, it sounds like you will
never be able to try since your primary interface is a non-graphical DOS
system, connected to a DECtalk? Is that right?
Kind regards,
Charles.
On Wednesday, 21 October 2026 at 14:25, Karen Lewellen
<[email protected]> wrote:
Again, there will be no Linux speech of any kind on this unit.
such would put me in Hospital.
Instead, I am using ssh, from DOS, into this box.
On Wed, 21 Oct 2026, didier gaumet wrote:
Le 21/10/2026 à 14:32, Charles Curley a écrit :
On Tue, 20 Oct 2026 23:51:56 -0400 (EDT)
Karen Lewellen <[email protected]> wrote:
HP Compaq dc7900 Small Form Factor Desktop
Intel Core2 Duo CPU E8400 @ 3.0GHz
4MB DDR2 RAM
[..]
Given that you won't be running a graphical user interface (GUI), you
should be fine with 4 GB. If the hardware will let you add more RAM,
you can do so at any time.
Hello,
I am not knowledgeable about all that but 4GB is not much and Karen's
computer processor, while 64 bits so Debian 13 compatible, is old (18 years).
From what I have heard, Text-To-Speech (TTS) can use AI (either a local LLM
or a cloud based solution) that improves quality, but potentially at the cost
of a greater need for physical resources (memory, computing power...).
There is a little explanation here:
https://linuxvox.com/blog/tts-linux/
I don't know what Karen uses, is currently satisfied with, and so on, but I
would personally be cautious about giving Karen advices to migrate fully to
Linux, because if something turns bad, that is she, not I, who risks being
stuck with a temporary lack of a working PC :-)
If she can be assisted, on site, by a linux-competent friend or Linux User
Group, that is different :-)
