I had found out about Linux in the last year of school, which was in 2000. I then used Linux in a server role from 2001 through to mid 2004. All this time, I attempted to keep up with the goings on of any accessibility efforts that were under way, which was for the most part, not a lot. There was Emacspeak(1), which provides synthesized speech output for use with Emacs, Speakup(2), which is a console screen reader that is a kernel patch, and finally, the GNOME Accessibility(3) framework, which I first heard about in 2003. Being lucky enough to have some sight, I never required any assistance to get Linux installed and set up, due to the use of a 21 inch monitor. This allowed me to easily evaluate the various Linux accessibility solutions available at the time.

I could never get the hang of emacs, and thought that any new user would also find it difficult, unless they really put their mind to it. Added to that, you also had to learn how to use the screen reader at the same time, which made things even more difficult. Speakup was also a difficult beast to tame, as it meant that I had to patch and compile a kernel with it included. I had compiled plenty of kernels before, but this was slightly different, in that I had to set it up to work with software speech output via the sound card, which meant more software had to be installed and configured. I was a Slackware user at the time, which meant just about all the software I needed had to be compiled from source.

Finally, there was the GNOME accessibility implementation. I managed to use the garnome build system on Slackware to build the latest stable release of GNOME, and have a look. I was quite impressed at the amount of information that was accessible from the desktop, and many of the applications that came with GNOME by default, however there was a lot that still needed work to be more accessible. Added to that, the program used to talk to the accessibility framework, (I will not call it screen reader, as it wasn't actually reading what was on the screen), was designed in a way that was confusing at times, and made some information about various applications difficult to obtain, or just didn't work with some applications at all. I concluded at the time, that while promising, there was nothing yet that was useful for day to day work with GNOME.

When I decided to move to Linux full-time from Windows, the solution I chose was speakup + software speech. This meant a shift from a GUI to a console environment and work-flow, but I soon adjusted. Email and working with documents was no problem, however web browsing would often be an issue, due to the text mode browser I used at the time, having poor HTML implementation. Speakup would also give me grief, as it turned out to be badly coded, and a poor performer, particularly on SMP systems.

Fast forward to the middle of the Gutsy development cycle. Speakup was removed from the gutsy kernel, due to not being able to patch cleanly into 2.6.22, as well as not working, due to some kernel mechanisms having been changed. The GNOME accessibility utility for navigating the GNOME desktop with speech output had changed in the last 6 to 12 months, and was a lot more flexible than the previous utility ever could have been. Added to that, a lot more of the desktop was accessible, including gnome-terminal. Unfortunately, GUI web browsing isn't still quite usable yet, but I have found a better text mode browser, and I still spend most of my time in a terminal. Being able to use the GNOME desktop for the most part, is something that has been well worth waiting for.

Links:
(1) http://emacspeak.sourceforge.net
(2) http://www.linux-speakup.org
(3) http://live.gnome.org/Accessibility