Re: [orca-list] qt apps?

[Joanmarie Diggs <jdiggs igalia com>]

Hey Burt, all.

Last time I checked Qt didn't give us access to the keyboard events. If
that is still the case, what happens when you use an Orca command is:

1. Orca listens for keyboard events
2. You use, say, a flat review command
3. Orca hears crickets because Qt didn't tell Orca about it
4. Orca cannot consume an event it never received
5. The Qt app gets your keystrokes
6. The Qt app possibly consumes the keystroke (adding text, moving you)
7. Qt tells Orca the text was added or the caret and/or focus moved
8. Orca presents the new location

The above would explain why, on a Desktop keyboard, your KP_9 (which is
Page Up if NumLock is not locked) moves you a page. In other words, flat
review isn't working differently in Qt; flat review isn't working at all
in Qt because, again, Orca cannot process a flat review command if it
doesn't know you used a flat review command.

A way to test this theory would be to use Orca + h in a Qt app. If Orca
gives you the Orca version, etc., my theory is wrong and a debug.out for
flat review would be helpful. On the other hand, if Orca doesn't speak
the version, then my theory is right and there's not a thing I can do
about it in Orca.

HTH.
--joanie

On 08/24/2016 04:16 PM, B. Henry wrote:
> I am not familiar with kaccissible, but many qt aps work well enough with orca with out it.
> Others are partially accessible, enough to be usable, and others are not usable at all although orca may 
> read some bits and pieces of the interface.
> Remember that flat review does not work the same in qt as it does with GTK using aps, i.e. you get page up 
> and down buttons for some of the numpad keys, 3 
> and 9 I think it is. 
> Some unity versions use a lot of QT, but one or more use something else, something custom as I understand 
> it, not sure of details, but keeping unity 
> accessible has been quite a job if I understand what I've read correctly.
> Perhaps Luke can elaborate and correct any errors in what I am saying.
> Some developers are not using QT's accessibility features correctly or at all, again I qualify with, "as I 
> understand it...".
> As far as I know, kde is still pretty far from being a realistic option for blind users.
>
>