1 – What is ATK and AT-SPI in simple words?

AT-SPI is the acronym for Assistive Technology Service Provider Interface. Its main purpose is to provide a means for an assistive technology to interact with an application. For instance, the Orca screen reader wants to present newly-inserted text, such as a new instant message, to the user. Therefore Orca asks AT-SPI to inform it whenever text gets inserted. When Orca is told what text has just been inserted, it can present that new text to the user in speech and in braille. Similarly, Orca presents each newly-focused object to the user as the user navigates via the keyboard. Orca can do this because AT-SPI tells it each time a new object gains focus.

Orca, of course, is not the only assistive technology out there. Speech recognition tools can make it possible for users to speak the name of something they wish to click on, like the ‘Close’ button in a dialog box, and then do the clicking for them through AT-SPI. A screen magnifier can make sure the user’s location is always visible on the screen by paying attention to changes in the location of the caret, in the selected item, and in the focused item — each of which gets reported to assistive technologies by AT-SPI.

AT-SPI in turn needs a way to get this information from the applications being used. This is typically accomplished through the application’s toolkit. For instance, Gtk+, Clutter, Gecko, and other toolkits implement an accessibility abstraction layer called ATK. If a toolkit implements ATK, then AT-SPI will automatically receive the information from that toolkit thanks to the atk-bridge. It turns out that, in the case of Qt, there is no ATK implementation. Instead Qt implements its own direct bridge to AT-SPI. But that is a special case. What is important is that in all cases, accessible free desktop toolkits expose information to assistive technologies, and assistive technologies get that information via AT-SPI.

2 – How easy is for someone with disabilities to run a computer? How far is Linux from other similar proprietary software?

In short: it is not especially easy. And in some cases the problems start early on in the user experience because not all distros have accessible installers yet. Having said that, we are seeing more and more awareness of accessibility on the part of developers as well as distros. Thus we are getting closer and closer to the point of Linux accessibility which “just works.”

With respect to how far away Linux is from other similar proprietary software, we’ve admittedly got a ways to go. But it is also worth mentioning that in the case of Linux GUIs, accessibility is comparatively new, having been created just 10 years ago.

3 – Do you believe the financial costs make it more difficult?

Yes, in our opinion financial costs do make things more difficult for the end user. Accessibility related software and hardware tend to be quite specialized, and are only required by a very small percentage of the overall user market. As a result these tools are often expensive. For example, the leading, proprietary Windows screen readers cost in the ballpark of $1000 US for a single user license. In the case of the hardware, a braille display with just 24 cells can cost three thousands dollars (or more). Figuring out how to make the hardware more affordable is a difficult problem to solve, but we can at least reduce the overall costs faced by users through Free Software.

Read more from source