[u-a-dev] [g-a-devel] gnome-speech, and audio output, moving forward.

Steve Lee steve at fullmeasure.co.uk
Tue Oct 16 09:25:29 BST 2007


On 16/10/2007, Luke Yelavich <themuso at themuso.com> wrote:
> - From reading a summary of the recent Boston summit, I don't remember seeing anything regarding
> speech. I would really like to see if we as a community can't work out a solution.

My 2cs, a standard speech platform would be excellent for developers
and users. I would like to add a simple self voicing capability to
Jambu but have not done so yet, largely due to the number of
permutations to consider. Charles Chen suggested a number of
possibilities including bundling espeak.

My ideal is a standard Gnome speech API which could be assumed to be
there (so no need to include in build) but still allows end user
flexibility.  So that's similar to SAPI on Windows (the SAPI speech
markup is just icing on the cake). As Linux thrives on choice a more
palatable option might be an API wrapper that works with the various
speech systems.

Could a module be factored out of Orca for this as much work has
already been done there? Python would suit me as it seems to be
becoming the language of choice for Gnome AT.

Then for a simple self-voicing scenario I could just:

  import pygspeak as pgs
  pgs.say('hello world')

I18n would no doubt add some interesting wrinkles.

For general non-visual accessibility making the program fully
accessible to Orca should be the approved approach.

A mechanism is needed to mute self voicing apps when using Orca. A non
server solution could be to check for orca in the say() function and
having orca check at startup.

-- 
Steve Lee
--
Open Source Assistive Technology Software
PowerTalk - your presentations can speak for themselves
www.fullmeasure.co.uk



More information about the Ubuntu-accessibility-devel mailing list