FW: Editorial: The Potential of Open Source for the Visually Impa ired
FYI -- response from JP
-----Original Message-----
From: JP S-C [mailto:jp_sc@yahoo.com]
Sent: Thursday, July 20, 2000 11:41 PM
To: Brent Fulgham
Subject: RE: Editorial: The Potential of Open Source for the Visually
Impa ired
Dear Brent,
Thank you very much for forwarding my message
to Debian-Devel and to GNU.org. I had not yet
contacted GNU/FSF, but in retrospect I should have
since they might be interested in my project. I am
very happy to hear that you think the project's goals
are very important. I can only hope that many other
current and future free software advocates have your
wonderful attitude.
"I'm not sure if any of us have skills or experience
in the specific area of the vision-impaired, but
hopefully a few will have time to visit your site and
see if they could help." (from your reply)
I am grateful to you for helping make the
project visible, which is the first, immensely
important step. In response to your question of
whether or not Debian developers have skills or
experience in the specific area of visual impairments,
I would say that developing software for people with
low or no vision is very similar to developing it for
the sighted. The code writing process is the same and
many of the functions are identical (with some
exceptions, e.g., ones related to appearance or GUI);
the only difference is how you present your program's
functions to a user. My, perhaps overly simplified,
way of testing if your application is usable to the
visually impaired is the following: if you met a
stranger with moderate computer ability (can touch
type), told her some basic commands, and asked her to
close her eyes, walk up to a computer, and use your
software, could she do it? I use the example of a
stranger closing her eyes rather than you (the
developer) doing so because the person who wrote the
program often knows it much better than anyone else
ever will. I think some current pieces of open source
software that can be used by the visually impaired
fail my test because in order to use the software one
must read quite a bit (some would say an excessive
amount) of documentation and information on slightly
esoteric shortcut keys in order to perform simple
functions.
Question 1: "Are there API's that editors or mail
readers should make use of to interface with
assistance devices?"
Answer 1: Unfortunately, interfacing with specialized
adaptive devices is not very standardized. Support
for Braille displays and hardware voice synthesizers
can be spotty and drivers are often written on a
device by device basis. This is one of the problems
that Ocularis distribution of Linux seeks to avoid.
Ocularis is meant to run on inexpensive, readily
available (e.g., at your local computer store rather
than a special order from another state or country)
hardware that already has stable support under Linux
and for which physically interfacing with the device
is no a focus. Although Ocularis may support adaptive
devices, it will focus on using current hardware to
produce an Audio User Interface (AUI) in a more cost
efficient manner.
Question 2: "Is there new software of some kind that
needs to be written?"
Answer 2: This is a good, slightly difficult question.
Yes, new software does need to be written, but this
new code deals primarily with re-expressing commonly
used functions than finding a new way to perform the
functions themselves. For example, a new e-mail client
need not be written with its own built-in speech
synthesizer. The speech synthesizer and basic e-mail
client (or libraries for simple e-mail functions) have
already been written. One need only collect the
header information and body of the message that are
usually displayed on a screen and "pipe" that
information to the speech synthesizer.
There are a few specific pieces of software
that need to be written. The first one that comes to
mind is a Linux installer that will allow a visually
impaired user to install with a generic sound card and
keyboard. Of all the software that Ocularis will
include, this will probably be the one that has the
most new code (rather than incorporating already
written free software, as in the e-mail client
example). The other pieces of software necessary for
a basic, functional workstation for the visually
impaired are a word processor, calendar, calculator,
basic accounting or finance application, file manager,
Internet browser, and e-mail client. In addition,
easy to use, all-purpose screen reading software will
be included to provide access to nearly any
console-based program.
These basic applications will be most likely
written primarily in a high-level language, such as
Python (although the help of developers who use low
level languages, such as C/C++,etc., would still be
appreciated). Because high-level languages already
include libraries that can perform the underlying
functions required for each of the applications
mentioned above, the task is actually smaller than it
may seem. The exception to the minimization of code
due to using high-level languages will probably be
with the installer, which, in order to have sane RAM
requirements, must be written in a lower level
programming language.
Question 3: "What sorts of things would make Debian
easier to install for the blind?"
The features that would make Debian easier to
install for the blind would be the incorporation of a
software text to speech synthesizer and the removal of
installation options that would rarely be used and
those that could not be used by someone with no
monitor and no or little sight. Open Source
software-based speech synthesizers have already been
written and some are quite good although they must be
tweaked somewhat in order to have a reasonable amount
of RAM required for installation. As with using
applications, a visually impaired user who is
moderately computer savvy should not have to read
volumes of documentation about installation. Several
installation steps would obviously be removed, such as
setting up a graphics card, X-windows, and perhaps ink
printers. That leaves the following steps: (automatic
support for generic sound cards), install method,
partitioning, package selection, LAN/WAN/modem setup,
time zone, root password, and boot options.
Question 4: "Can software voice synthesizers (e.g.,
the Emacs voice system) be used to provide useful
utilities for the blind?"
Yes, software voice synthesizers are very
useful, are much more convenient than hardware ones
and allow for the sub $500 [USD] target hardware
configuration on which Ocularis should run well. This
is an important question and one of the fundamental
reasons why Ocularis was created -- to utilize
software voice synthesizers in order to minimize
target system cost. A fuller explanation of this is
in my editorial. I have not yet had much hands on
experience with Emacspeak, although I have it
installed on my computer.
"I'm certain that many Debian developers would be more
than happy to help make their existing software
packages more useful to the blind where possible."
I hope to have answered your questions and that Debian
developers will be "more than happy" to help make
their packages more accessible. Thank you very much
for your e-mail and support of Ocularis. Feel free to
ask me more questions and to forward this message to
Debian-Devel. Ocularis is hosted at SourceForge,
"http://ocularis.sourceforge.net/"; and any input,
suggestions, or questions from Debian developers or
anyone else would be appreciated. Thanks again.
--JP Schnapper-Casteras
jpsc@sourceforge.net
__________________________________________________
Do You Yahoo!?
Get Yahoo! Mail - Free email you can access from anywhere!
http://mail.yahoo.com/
Reply to: