Re: Introduction and Comments

[Peter Korn <peter korn sun com>]

Hi Keith,

Welcome to the GNOME Accessibility community!

> My primary interest is development of a screen reader and self-voicing
> apps for GNOME. As a blind programmer who really hasn't gone beyond the
> text-based Linux or Unix terminal I'm wondering what my best option for
> developing for the GNOME environment might be.
> 
> My programming experience for the most part has been in that "other"
> popular OS that most of us love to hate. G
> 
> Oh yeah, C++ is my primary development language. I do have some
> experience with others, but I rock in C++.

Hmmm... Maybe it's time for a FAQ for this alias?

There are a number of options for you to pursue.  If you know emacs, then
emacspeak is a nice environment to work in on your Unix system.  Another is a
Windows system (or even a Mac!), a screen reader, and a terminal emulator hooked
in to a Unix box running GNOME.

You'll need the latest GNOME source (instructions on developer.gnome.org should
help you there), and to work with CVS, etc. to get and compile that.  Only the
latest code contains the accessibility work.  There is now also a CVS tree
specifically for the accessibility libraries (see Padraig's e-mail of earlier
today).

Then you need to decide what sorts of things you want to work on.  There is lots
to do!  A quick outline of some of what remains to be done, not exactly in
order:

 - finish implementing the ATK on GTK+ widgets and GNOME widgets
 - define and implement the "platform accessibility service provider interface"
 - write bridges between ATK and the platform SPI; Java and the platform SPI;
   others...
 - go through the "standard" GNOME apps and ensure they implement ATK 
   properly (mostly this is an issue of custom components and following some
   coding rules)
 - write needed building blocks for AT (speech recognition, Braille drivers,
   etc.).
 - write assistive technology

Writing a GNOME screen reader is something we very much need, but may be hard to
start on until we have finished hashing out what the platform accessibility SPI
is, and have an implementation of it.  Writing some "throw-away" code that reads
ATK directly would be an interesting and useful bootstrapping project -
something you might consider.  That work would help prove the ATK (how complete
is it), and as the SPI will probably look a lot like the ATK, much of your code
could be re-used.  It would also allow folks writing GTK+ apps to start doing
early accessibility testing with blind users.


Marc, Bill, Padraig, Owen, & others - any other suggestions for Keith?


Regards,

Peter Korn
Sun Accessibility team