Re: [g-a-devel] Status of IBM a11y

[Henrik Nilsen Omma <henrik ubuntu com>]

Janina Sajka wrote:
> It should be clear that we cannot deliver accessibility to end users
> based on wishes or on theory. We need solid, reliable, working
> implementations. Until those are available in dbus, it would be
> irresponsible in the extreme to abandon the existing working Corba
> implementation. Given this obvious truth it's troubling to see Bonobo
> deprecated in Gnome-2.19. What is the meaning of that? If actions speak
> louder than words, how are we to understand a commitment to a11y in
> Gnome?
>   
Janina, this is a rather counter-productive position to take. If we were 
talking about Gnome dropping a specific accessibility library then I 
could see the case for making an ethical argument, but they are 
migrating from one generation of general desktop infrastructure to the 
next. a11y should keep up.

You argument is logically flawed as well (as opposed to just tactically 
unwise). You are implying that a move to dbus would be somehow 
instantaneous, that it's not possible to start work on a dbus version 
without abruptly dropping the current Corba version. The truth is that 
both Corba and Bonobo will be dropped eventually (as will Dbus, X and 
the linux Linux kernel, if you look far enough ahead). By delaying the 
work we are making that inevitable conversion more abrupt and painful. 
It will come at a later stage in Linux deployment, when we will have 
many more disabled users, many of whom will not be very technical.

In addition we loose in two other ways:

We fragment our own community by not being able to work with other 
projects like KDE. This has largely already happened.

What's even worse is that we lose the interest, common ground and good 
will of the wider FOSS developer community. For a start we won't be 
using the same tools they are using on a daily basis, so it will be much 
more difficult just practically to get involved in this work. In 
addition the a11y community will constantly be pushing to stay on old 
frameworks, not for technical reasons by for moral ones. Morality is 
fine, but don't over-play it. It just won't give the results we need.

Let's instead talk about what is needed to move a11y to dbus, set up a 
project, a wiki page and start moving forward. We'll have to do it 
eventually. Waiting just makes it worse. I think that a clear request to 
the wider community from a unified a11y community to help with a 
transition to dbus will yield a much better result for everyone than a 
draw-out and eventually futile struggle to avoid it.

I'm mentoring three a11y summer of code projects this year, 
magnification, colour enhancement and mouse enhancements. These tools 
will interact with their environment via dbus wherever possible.


Henrik