Eric,
clearly you do not know the difference between software and hardware
speech, if you can even suggest I risk myself ..again, by talking of Linux
below.
The only point I will speak to is one that echos your own.
If Links incorporated additional forms of JavaScript then the one used,
same can be said for e-links, then for me personally those browsers might
be just fine.
I use sshdos to reach a Ubuntu shell, but even in Ubuntu the links and
elinks JavaScript falls short in places, especially if cloudflare is being
used.
In the case of Links for dos, there are two problems, the JavaScript one,
and the lack of automatic speech in links for dos. It is not a problem
for the editions of the software incorporated into the Ubuntu shell, such
as the one dreamhost provides with its shared hosting accounts.
I keep meaning to write the developers about this issue, but have not
found the time.
as for Dos adaptive technology having the ability to manage graphics, it
depends on how the graphical interface is written and the nature of the
tool, there are scores of them after all.
certainly I want a DOS only solution, I use only DOS, and this is a list
called freedos.
If I want to be reminded of all the Linux problems, I need only read the
lists like speakup, where even in the command line Linux cannot find a
dectalk USB. Lastly text to speech is not a screen reader.
Opting out of further discussion, We have a good 12 years of
miscommunication, and I have no vested interest in talking to walls.
On Fri, 25 Jun 2021, Eric Auer wrote:
Karen,
I was not aware that Linux screen reader voices trigger seizures.
Nobody is forcing you to use those, or even to use Linux at all.
Not sure why exactly mentioning Braille is evil. The underlying
software infrastructure task is the same: Take text from an app
and transform it to another output modality. Even back in the days
of CTTY COM1 with DOS, the problem existed that not all apps are
using the proper interfaces to be redirectable and between the
lines, I also wanted to express my doubts that Blue Lion would
be part of the solution, as proposed by Liam, but I have no OS/2
adaptive technology experience myself.
My reference to text oriented browsers was based on the assumption
that in DOS, adaptive technology would NOT have a standardized
interface to communicate with graphical applications, because
DOS does not provide graphical building blocks to the apps.
So if you use Arachne or Dillo in DOS, there might be problems
which you could avoid by using a text based browser in DOS. If
there are NO problems with your DOS drivers, great, please let
us know! Just mentioning that DOS apps with GUI may have issues.
Of course Linux does support connecting texts used by graphical
applications, at least from more widespread GUI frameworks, to
be processed by adaptive technology. As you write, there also
is the problem whether the output side of that can work with
the style, brand or hardware you prefer to use.
My impression was that Liam's suggestion to use OS/2 or Blue
Lion was not solving the task at hand either, so I focused
on which browsers could be suitable in DOS. Knowing that you
had explicitly asked for a DOS based solution.
Links for DOS, for what it is, opens some doors, but not all
Which features are useful in Links, which are missing?
if Linux is such a grand solution, why cannot a graphical
installation be configured so it can communicate with physical
speech hardware?
You mentioned having no Linux driver for your speech hardware,
so I expect that problem to be not limited to the installer.
Support for hardware speech devices in Linux is quite limited:
Knoppix, which explicitly supports the ADRIANE Audio Desktop
Reference Implementation and Networking Environment starting
at the installation itself, uses ELINKS as browser with both
javascript and multimedia support. It also uses ORCA OCR to
fetch text from graphical applications IF those fail to have
adaptive access to text fields and espeak for text to speech
(quality varies a lot depending on installed voices).
https://www.knopper.net/knoppix-adriane/index-en.html
According to the German SBL (used by Knoppix) documentation,
the system supports Papenmeier, Handytech, Baum, Alva, Tiemann
and Blazie Braille devices. For text to speech, Apollo2,
Vox700, Festival, TTSynth, Speechd and MNROLA are supported.
Of course, none of this is relevant for DOS browsers. I do
not expect graphical DOS apps to have good compatibility
with anything beyond VGA, but of course I am happy to hear
about DOS GUI apps which do support more output modalities.
I do have some experience with writing a simple adapter
for Dutch text to speech long ago, so I am actually aware
of the limitations of the technology. Even now, youtubers
who prefer to stay anonymous use annoyingly artifically
sounding speech engines. I think I even have a chip from
back when it was important to offload speech output from
the CPU to dedicated hardware somewhere in my collection:
https://en.wikipedia.org/wiki/General_Instrument_SP0256
That thing sounded quite bad even for English. Luckily,
hardware speech output devices today just contain their
own computers, with dedicated optimized TTS software.
If freedos is never going to provide a proper browser,
how can it claim to be a fully functional operating
system where networking is concerned?
Actually DOS does not claim to be a networked operating
system at all. There are some de facto standards for
network drivers for DOS which are in turn used by DOS
applications which implement their own networking, but
this is not something provided by DOS as the operating
system. So I think it would not be better or worse if
you were to find a way to use a Windows browser with
low enough system requirements to run it on Japheth's
HX RT and HX GUI which lets you run some Windows apps
directly inside DOS. Of course this would depend on
whether HX can connect to your adaptive technology. In
any case, using the world wide web in DOS is something
which can be quite frustrating for everybody, but then
DOS does not make any promises about networking either.
Regards, Eric
PS Liam: I am surprised that you have so much experience
with screen readers, so maybe you could share some ideas
about which of the free Linux TTS engines have which
strengths and weaknesses based on what both you and
the users you know think about them in recent years?
And assuming that Knoppix is quite "German", which Linux
distros are your users using with which text output path?
_______________________________________________
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
