Re: Two questions: GTK for visually impaired. Windows gtk bundle for apps guidance.

On 04/10/2012 11:14 PM, Evgeniy Philippov wrote:
> 1. I'd like to try to help develop GTK for visually impaired people
> under Windows. I am not sure how my motivation will last long or not,
> but my free time is currently unlimited. 
> I have reports from visually impaired ones that Pidgin client under
> Windows is not very accessibility-features-friendly. These people have
> referred me to Windows UI Automation API ( 
> ). Pidgin is based
> on GTK. Those people prefer to use Miranda which uses WinAPI because
> Miranda is more accessibility-features-friendly. I intend to try to help
> change that state with GTK under Windows.
> The first set of questions is: 
> 1A) Could you tell the current state of accessibility support of GTK+
> under Windows? Current stable Pidgin uses gtk
> +-bundle_2.14.7-20090119_win32 under Windows.

AFAIK, accessibility support of GTK+ under Windows is zero. Current GTK+
accessibility support is based on ATK/AT-SPI2 technologies, not
supported on Windows.

Anyway, it is worth to mention that accessibility APIs status under
windows is somewhat messy, IMHO. At this moment you have available:

 * IAccessible2 (a extension of IAccessible defined on MSAA)
 * Windows UI Automation (the one you mentioned).
 * Java Access Bridge (for Java apps)

Each app would support one or the other depending on how old it is. In
theory Windows UI Automation replaces MSAA but the fact is that some
apps have only MSAA support. That lead that some accessibility tools,
like NVDA, uses whatever is available. As you can see on NVDA about page
[1], it supports all those technologies, in spite of some of them
replacing the others.

AFAIK, Qt only implements MSAA, probably because it is the most used
(disclaimer: I'm not a Qt expert).

> 1B) Is it hard to add support for accessibility features to GTK+ under
> Windows? 

Yes, it is hard, in the sense that it is not a trivial tasks. Some time
ago some people talked about creating an ATK-MSAA wrapper, as ATK
abstraction and MSAA abstraction have a lot in common, but AFAIK, nobody
started or had time for that work.



Alejandro Piñeiro Iglesias

[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]