RE: Implementation questions around HTML structural elements and ARIA Landmarks

HI Steve,
I am working on three things.

1)      UIA implementation of HTML Sections

2)      UIA implementation of ARIA landmarks

3)      Narrator experience for navigating sections and landmarks

For 1 & 2, we will be using a Group Control Type and a Localized Control Type that corresponds to the sectioning element or landmark.  Control type is the UIA equivalent of role.  Localized Control Type is what Narrator says when the user interacts with that control.  It is a user-facing string which will be localized.

3 is not as far along in the design process, but I am thinking of something similar to heading navigation in narrator or the F6 loop in Windows. This is similar to how the screen-readers handle Landmarks in your link below, but would include both HTML Sections and ARIA Landmarks. The user would cycle through the Landmarks and Sections in order, having the Localized Control Type and Name of each Landmark/Section read out as they land on it.  I think HTML Sections and ARIA Landmarks are logically equivalent and should be part of the same user experience.  I also think the names should be consistent, and that HTML has the better names.

As to your question “Are you following all other implementations and mapping the structural elements as per the implementation requirements in HTML5 [1]?” That’s the goal.  However, I’d like to use the word “header” rather than “banner”, “aside” rather than “complementary”, “footer” rather than “content info” for the strings the user actually hears.  The landmark role will (most likely, still discussing) be mapped directly to a new LandmarkType property in UIA, which could use the strings in the table at http://stevefaulkner.github.io/html-mapping-tests/.



I have another question too… Why is the html <form> element not mapped to the ARIA form role?  Why is it not a landmark?  Seems like it should be.  Were there concerns about backward compatibility to “forms mode”?


From: Steve Faulkner [mailto:faulkner.steve@gmail.com]
Sent: Friday, March 20, 2015 2:48 AM
To: Cynthia Shelly
Cc: public-html-a11y@w3.org; PF
Subject: Re: Implementation questions around HTML structural elements and ARIA Landmarks


On 19 March 2015 at 17:43, Cynthia Shelly <cyns@microsoft.com<mailto:cyns@microsoft.com>> wrote:
We’re working on an implementation of accessibility for HTML structural elements in IE.  We have also implemented ARIA landmarks, which I showed at CSUN. (Neither is in a public build yet) These features are closely related and very similar, and we expect that they will be part of the same user experience in Narrator and other AT products.

Hi Cynthia,
these may be helpful, though somewhat old now:
http://www.paciellogroup.com/blog/2013/02/using-wai-aria-landmarks-2013/

http://www.html5accessibility.com/tests/landmarks.html


It is a little unclear to me whether you are talking about what the user will experience when using narrator or what will be exposed via UIA.
 "We’re working on an implementation of accessibility for HTML structural elements in IE."

Are you following all other implementations and mapping the structural elements as per the implementation requirements in HTML5 [1]?



[1] http://stevefaulkner.github.io/html-mapping-tests/

--

Regards

SteveF
HTML 5.1<http://www.w3.org/html/wg/drafts/html/master/>