New developer

Bruce Couper bruce.couper at
Mon Nov 3 18:04:39 GMT 2008

Hi Francesco, All

A few comments from a quite disabled speech-recognition user and light
user of Ubuntu through VNC (to learn and help with wife's and friends'

Speech recognition is the barrier for some of us adopting GNU/Linux,
Ubuntu.  Echoing an earlier post, as much as I support open source I
would be prepared to pay for the commercial product
(NaturallySpeaking) if it were ported.  Arguments that doing so would
be prohibitively expensive seem refuted by the recent porting of the
basic speech engine to Macs.  The owners of NaturallySpeaking do not
seem inclined to listen to users but if someone has any suggestion for
having someone at Canonical making contact...

Otherwise, there would be a great deal of work, as I understand it,
needed to make what now exists (Sphinx) usable.

Using Wine might provide dictation but no command and control.

Dasher is actually very good and, in my limited experience, does a
pretty good job of word prediction and learning.  However, as you
alluded to, the last build I looked at under Ubuntu no longer
supported outputting text to an active window rather than just the
Dasher text area, making it far less useful than it should be.  Older
versions (at least under Windows) did have the option of outputting
text to another active window and, if I recall, there has been some
recent change involving development of the application.  I think it is
a very promising application for some users and it supports many input
methods (switches, eye tracking etc.).

I'm 55, not a coder and limited now to speech-recognition.  I offer my
support, occasional humble notes and, if I may, whatever have one can
provide in documentation in the future.


On Tue, Oct 28, 2008 at 6:09 AM, Francesco Fumanti
<francesco.fumanti at> wrote:
> Hello Tom,
> There is another accessibility area that lags behind what is already
> available for years on Windows and Macs: a good accessibility tool for
> people that can only use the pointer.
> - There is the gnome onscreen keyboard gok, but it is more geared
> towards switch users. Several modifications would be necessary to
> improve it for pointer only users. Other improvements have been noted on
> this page:
> The current maintainer of gok only has very little time to devote to it,
> so that only the most urgent problems have been looked after.
> - There is the onscreen keyboard onboard: onboard, which has been mostly
> programed in python, is probably the most usable onscreen keyboard
> available on GNOME for pointer only users. However, it is still a basic
> onscreen keyboard that lacks efficiency features like word prediction
> and autoponctuation. Moreover, it does not support switch input yet. I
> suppose that not only the accessibility users, but also the tabletpc
> users would appropriate having somebody that takes onboard to the next
> level.
> - Another starting point to build an efficient onscreen keyboard could
> be the matchbox-keyboard, but as it is not originally geared towards
> - There is kvkbd for KDE, that I have not really look into yet.
> - There is dasher with its unusual input method that looks more like a
> game than an accessibility (this is not intended in a negative way).
> dasher is said to have a good word prediction, but it completely lacks
> desktop control and the written text goes into a pane in dasher instead
> of going into the front window. I don't know whether the word prediction
> of dasher can be used in a "classic" onscreen keyboard.
> As you see, there is also much left to do for pointer only users. It is
> not the area that you talked about in your mail, but one can never know;
> therefore this reply.
> Anyway, if you are going to pick up one of these to take it to the next
> level, please contact me and I will help with what I can (ideas,
> testing, documentation,...).
> Cheers
> Francesco
> Luke Yelavich wrote:
>> On Tue, Oct 28, 2008 at 10:23:43AM EST, Tom Lloyd wrote:
>>> Hello All,
>>> Just wanted to say Hi and to get myself known. I have been using
>>> Ubuntu for= three or so years. I am a 26 year old developer from
>>> the UK trained in Emb= edded / Realtime systems. As a side project
>>> I am intergrating SAPI into Ubu= ntu to gives access to the MS
>>> speech engines using speech dispatcher.
>> This is awesome news, and depending on how far along your work is,
>> may be enough of a reason to push for using speech-dispatcher as the
>> speech backend for orca in the next Ubuntu release.
>>> I use compiz for zoon, so i could be intrested in working to
>>> improve speech= and magnification under ubuntu, suggestions for
>>> projects welcome.=20
>> We certainly need help in improving the state of magnification in
>> Ubuntu. It works, but badly. What really needs doing is integrating
>> the eZoom compiz plugin with Orca, as well as cleaning up Orca's
>> speech-dispatcher support to be of similar quality to its support for
>> gnome-speech. I plan to more tightly integrate speech-dispatcher for
>> Ubuntu next cycle, due to its more flexible nature compared to
>> gnome-speech.
>> In short, integration is what we lack for accessibility in Ubuntu.
>> The tools are there, but they need to be tied to gether in a way that
>> makes them work seemlessly with each other, i.e eZoom/gnome-mag and
>> orca, speech-dispatcher and orca, etc.
>> Thanks for your offer of assistance, it would be much appreciated.
>> Luke -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.9 (GNU/Linux)
>> Mn8AoK7NxudfeWOk9B/uAnIq8NBJ0p2o =t73s -----END PGP SIGNATURE-----
> --
> Ubuntu-accessibility mailing list
> Ubuntu-accessibility at

More information about the Ubuntu-accessibility mailing list