Re: Extending GtkMenuItemAccessible to allow applications to set alternative accessible names for menu items.

Doesn't UI provide a tooltip in this case which is usually mapped to
accessible name if name is not provided? At least that happens on the
web.

But technically it's sort of weird that application creates a widget
but it can't control the properties of widget accessible object.

Thanks.
Alex.

On Fri, Dec 2, 2011 at 10:32 AM, Luke Yelavich
<luke.yelavich <at> canonical.com> wrote:
> Hi all,
> One feature request that the Ubuntu accessibility team receives from users, is the ability to find out the
signal strength and security of a wireless network that appears in the network list. In the case of
network-manager, the icon is used to convey network signal strength and security to the user. After
looking through the code of GtkMenuItem, and GtkMenuItemAccessible, I've determined that there is
currently no way to set an alternative accessible name for a menu item, in the case where the icon is
conveying extra information to the user. Yes, one could put sed information in the label, but that would
bloat the menu item visually, which is not desired.
>
> I was talking to a few GTK developers on IRC earlier today about GtkMenuItem, GtkMenuItemAccessible, and
somehow allowing applications to set alternative accessible names for menu items, whether it be via an
extension to the GtkMenuItem API, or extending the GtkMenuItemAccessible object to allow applications
to set a different accessible name using the atk_object_set_name method call. Benjamin did suggest that
the icon could possibly be exposed via another atk object which is a child of the menu item, with a
description set in the icon atk object to describe the icon, however this gets messy when submenus are put
into the picture.
>
> I'll let Benjamin et al reply as to what their thoughts are about this proposal, but I'd be interested in

Re: [orca-list] Thunderbird launching very slowly if Orca running. Why happening this?

On Thu, Dec 08, 2011 at 12:46:27PM -0500, Joanmarie Diggs wrote:
> Hey all.
> 
> >>TOTAL PROCESSING TIME: 24.5515
> >
> >what unit is this? if its seconds that's really suprising.
> 
> Yeah, it's seconds. And looking at Attila's debug output there are

hmm, it seemed to me like that was the processing time for just one of
the events, but I guess I don't understand the format that well 

> over 10,000 lines (yes ten thousand!!!) of
> object:property-change:accessible-name getting queued up. One line
> per event. In other words, this is an epic event flood from
> Thunderbird or Gecko. I'm really not sure what I can do other than
> to ask Thunderbird/Gecko to just cut it out for pete's sake.

Well, that's simple, thunderbird can just not tell you about the
messages 

More seriously the way a email client works is you have a treeviewish
thing with a list of all your messages (in this case ~25000).  Each item
in the list view should be an accessible right? so we need to add each
of those 2500 accessibles, and then tell at-spi that each of them got a
name.  I'm not sure why its 10000+ instead of 2500+ but that's roughly a
factor of four so with 0 debuging involved I'll gues there is more than
one accessible per message.

SO while I understand not wanting that many events I think you want

Re: [orca-list] Thunderbird launching very slowly if Orca running. Why happening this?

On 12/08/2011 09:15 PM, Trevor Saunders wrote:
>
> SO while I understand not wanting that many events I think you want
> that more than the alternative.
>
> I think the solution here needs to be some sort of api change in atk or
> smart events, so ccing g-a-d

Hi, I was been talking with Joanmarie briefly about this, and also
checking when this notification is sent.

In summary, it is sent each time the accessible name for an object
changes. And this includes when you set the initial value, so the reason
of the low start.

So, at this moment I don't think that this should require an API change,
but as you said, smart events. In that sense, just sent it when
required. Some options:

  * Use the same solution that with roles: default value for role is
UNKNOWN. When it is changed to something meaningful (like BUTTON), atk
doesn't sent that notification. So we could just avoid to sent the
name-change notification when you go from not having a name to having a
name, assuming that this is the initial set. Not sure about that, as
this is a real change on the name.
   * I was wondering if this notification is required always. Do we need
to know that a invisible object have changed his name? So probably other
option could be just sent the notification if the object is in a
specific state (ie: being focused).

Re: [orca-list] Thunderbird launching very slowly if Orca running. Why happening this?

On Fri, Dec 09, 2011 at 02:27:52PM +0100, Piñeiro wrote:
> On 12/08/2011 09:15 PM, Trevor Saunders wrote:
> >
> > SO while I understand not wanting that many events I think you want
> > that more than the alternative.
> >
> > I think the solution here needs to be some sort of api change in atk or
> > smart events, so ccing g-a-d
> 
> Hi, I was been talking with Joanmarie briefly about this, and also
> checking when this notification is sent.

sorry about the late reply.

> So, at this moment I don't think that this should require an API change,
> but as you said, smart events. In that sense, just sent it when
> required. Some options:
> 
>   * Use the same solution that with roles: default value for role is
> UNKNOWN. When it is changed to something meaningful (like BUTTON), atk
> doesn't sent that notification. So we could just avoid to sent the
> name-change notification when you go from not having a name to having a
> name, assuming that this is the initial set. Not sure about that, as
> this is a real change on the name.
>    * I was wondering if this notification is required always. Do we need
> to know that a invisible object have changed his name? So probably other
> option could be just sent the notification if the object is in a
> specific state (ie: being focused).
> 
> But this are only initial ideas, that require refinement. I will

Re: [orca-list] Thunderbird launching very slowly if Orca running. Why happening this?

On 12/14/2011 12:26 PM, Trevor Saunders wrote:
>
>
> But this are only initial ideas, that require refinement. I will
> investigate it.
>  So, my understanding (mostly from
>  https://bugzilla.mozilla.org/show_bug.cgi?id=659018) is that at-spi2
>  uses the signal internally for cache coherency.  So if we want to take
>  this route I think it needs to happen within atk / atk-bridge.  However
>  I'm not convinced this is really the best option here.  Having to
>  create 2500 objects just to fire events on them is rather unfortunate.

Taking into account the explanation from Mike on the atk bug that I created:

https://bugzilla.gnome.org/show_bug.cgi?id=665870#c2

This message and the relation with the cache coherency is only required
when the label updates his name. But from "something" to "something
new". We are proposing to not emit the signal if the change is from NULL
(nothing) to "first value", and seems that Mike is blessing it. Mike are
you here? Could you confirm it? In fact my plan is implement and test it.

And this complain about "having to create 2500 objects just to fire
events", is, IMHO, a different problem. But it is true that needs some
discussion. On the gobject world this could be solved by having
atkobject as a interface [1], but this shouldn't solve the issue on
other toolkits, like the ones based on C++ (Unity, Gecko) or Java. But
as I said, I think that this belongs to a different discussion thread.

[1] https://bugzilla.gnome.org/show_bug.cgi?id=642597

Re: [orca-list] Thunderbird launching very slowly if Orca running. Why happening this?

On Wed, 14 Dec 2011, Piñeiro wrote:

> Taking into account the explanation from Mike on the atk bug that I created:
>
> https://bugzilla.gnome.org/show_bug.cgi?id=665870#c2
>
> This message and the relation with the cache coherency is only required
> when the label updates his name. But from "something" to "something
> new". We are proposing to not emit the signal if the change is from NULL
> (nothing) to "first value", and seems that Mike is blessing it. Mike are
> you here? Could you confirm it? In fact my plan is implement and test it.

I think this makes sense. It would essentially mean that we would be doing 
for names (and descriptions, I guess) what we are already doing for 
roles...

-Mike G-

_______________________________________________
gnome-accessibility-devel mailing list
gnome-accessibility-devel <at> gnome.org
http://mail.gnome.org/mailman/listinfo/gnome-accessibility-devel

Re: GNOME Launches Campaign for Accessibility

Hey Bill.

On 12/16/2011 09:56 PM, Bill Cox wrote:
> I'd feel a lot better pushing gnome accessibility if GTK+ wasn't such
> a complete lost cause in terms of accessibility improvement.

I don't think this is a fair assessment. There has been a lot of 
investment in improving Gtk+ Accessibility, including the recent merging 
of Gail into Gtk+, and the subsequent work being done to find and 
eliminate any remaining regressions. Benjamin Otte has also been working 
hard in recent days to make GtkTreeViews much more performant.

>  Pixmaps should have accessible descriptions.  The calendar should talk.

Bugs should be filed. Developers should be pinged.

> Someone somewhere in GTK+ land should care, and I just don't see it.

http://git.gnome.org/browse/gtk+/log/

> I recommend channeling the few dollars we can influence towards
> projects that have actually made a big difference in the last two
> years: Orca, NVDA, emacspeak, speakup, Ubuntu, Vinux, Debian, and many
> others.  I am listening to this e-mail with NVDA, though if I happened
> to be in Linux, I'd be Orca.  People like Joan Marie and James Teh
> make this possible, at great personal sacrifice.  Other than the
> complete lack of cooperation from the GTK+ team over the last two
> years, what should I be feeling about Gnome that would make me want to
> organize contributors to support Gnome accessibility?  Where are the
> contributors like Joan Marie or Luke?

Re: GNOME Launches Campaign for Accessibility

On Sat, 17 Dec 2011, Joanmarie Diggs wrote:

>>  Pixmaps should have accessible descriptions.  The calendar should talk.
>
> Bugs should be filed. Developers should be pinged.

Bill did file a bug for the first issue at least, if I'm remembering 
correctly. I suspect that part of the issue is that, in an ideal world, 
people would file bugs, and they would always be addressed one way or 
another in a timely manner, but in reality this doesn't happen. People 
forget about bugs, people move on without their bug having been addressed, 
etc., so sometimes bugs languish / fall through the cracks, developers 
need to be pinged, etc. And Benjamin doesn't like Bugzilla by his own 
comments, so it's better to email him or ping him on IRC to get his 
attention.

So it's not an ideal world, unfortunately, and things often don't work the 
way one might think they should, and there's room for improvement in terms 
of processes, but I don't think this means that people "don't care" about 
accessibility issues per se.

-Mike G-

ATK 2.3.3 released

About ATK
=========

GNOME provides support for accessibility devices using the ATK
framework. This framework defines a set of interfaces to which
graphical interface components adhere. This allows, for instance,
screen readers to read the text of an interface and interact with its
controls. ATK support is built into GTK+ and the rest of the GNOME
platform, so any application using GTK+ will have reasonable
accessibility support for free.

Nonetheless, you should be aware of accessibility issues when when
developing your applications. Although GTK+ interfaces provide
reasonable accessibility by default, you can often improve how well
your program behaves with accessibility tools by providing additional
information to ATK. If you develop custom widgets, you should ensure
that they expose their properties to ATK. You should also avoid using
sound, graphics, or color as the sole means of conveying information
to the user.

The GNOME desktop ships with a number of accessibility tools which
enable users with disabilities to take full advantage of their desktop
and applications. Applications that fully implement ATK will be able
to work with the accessibility tools. GNOME's accessibility tools
include a screen reader, a screen magnifier, an on-screen keyboard,
and Dasher, an predictive text entry tool.

News
====

ANNOUNCE: AT-SPI 2.3.3 released

AT-SPI 2.3.3 is now available for download at:

http://download.gnome.org/sources/at-spi2-core/2.3/
http://download.gnome.org/sources/at-spi2-atk/2.3/
http://download.gnome.org/sources/pyatspi/2.2/

(note: There is no pyatspi 2.3.3 release; the newest release is 2.2.1)

What is AT-SPI2
===============

AT-SPI2 is a D-Bus based accessibility framework. It defines a D-Bus
protocol for providing and accessing application accessibility
information. The project includes a library for bridging the D-Bus
protocol to the ATK API, allowing Gtk based applications to be made
accessible. It also contains a client (AT) side library in C and a wrapper
for Python.

What's changed in AT-SPI 2.3.3

* [core/atk] Handle event detail suffixes (such as :system).

* [atk] Fix for BGO#664822 - gnome-shell crash when an AT is launched

* [atk] Fix a reference leak if a child-added signal has no object value.

* [atk] Only deregister objects when they are marked defunct, not when a
   previously defunct object is marked as no longer defunct.

Where can I get more information about AT-SPI2