Re: Communicating to users what GNOME 3.0 is

[Brian Cameron <Brian Cameron Sun COM>]

Paul:

> The accessibility changes are big - and as it's a core component of
> GNOME (accessible to all) there's probably a lot more we could be doing
> around it.  (I'm not an expert on accessibility in any way, shape or form).

While the planned a11y changes are big, I do not think that they have
much immediate end-user facing impact.  For example, I do not believe
there are any exciting new a11y use-cases that are planned to be
supported in GNOME 3.0, or any new significant accessibility tools.

Switching from bonobo to D-Bus does mean two exciting things:

- GNOME can finally deprecate ORBit2 and bonobo, something that has been
  desired for a long time.
- Switching to D-Bus will help the GNOME and KDE community work more
  closely together moving forward to develop a common a11y
  infrastructure.  Over time, this could lead to some exciting
  innovation, but this will probably happen gradually with end users
  noticing the benefits much later than the initial GNOME 3.0 release.

From an end-user perspective, these changes probably will not be very
visible.  If anything, the switch to GNOME 3.0 will probably be seen as
a step backwards for many a11y users for the following reasons:

- It took many years to tweak the use of ORBit2 and bonobo so that a11y
  features run as fast as they do today.  Switching to D-Bus will
  likely introduce some new performance issues that will likely, again,
  take time to resolve.
- GNOME Shell, other OpenGL/clutter components, and WebKit seem slow
  about addressing a11y issues.  So, there is a real risk that the
  GNOME 3.0 release will have some serious a11y regressions.  I imagine
  that these issues will be addressed over time, but may not be working
  well until a follow-up GNOME 3.x release.  Remember that the GNOME
  2.0 release was held up for a very long time due (in part) to a11y,
  and even so it was not until about GNOME 2.10 that a11y in GNOME
  really started being usable.

Obviously we will need to wait an see, perhaps GNOME 3.0 and a11y
will come together more elegantly than I suggest above.  However, if
not, then it might be hard to promote a11y as being an exciting new
feature of GNOME 3.0.  We may only be able to claim that GNOME 3.0
provides some exciting new non-user-facing a11y infrastructure, which
is probably not very exciting if the end-user experience is actually
worse for users with a11y needs.

Brian