This is an archived snapshot of W3C's public bugzilla bug tracker, decommissioned in April 2019. Please see the home page for more details.
MSAA columns are not clear on what UAs are required to do for aria-hidden. ATK column is blank. This is impacting the generation of testable statements for ARIA.
In the mapping table for aria-hidden, need to specify what the mapping is IF the object is in the tree and refer to section 5.1 for more information.
David, When an object with aria-hidden="true" is in the accessibility tree is aria-hidden mapped to anything or does FF just map everything else as if aria-hidden were not present?
(In reply to comment #2) > David, > > When an object with aria-hidden="true" is in the accessibility tree is > aria-hidden mapped to anything or does FF just map everything else as if > aria-hidden were not present? Currently everything else gets mapped as it would normally. Changes to aria-hidden will result in an object-attribute changed event as well as a reorder event on the parent node.
David, Review all sections in 5.8 for correctness of the way aria-hidden is specified.
aria-hidden is currently not correct. We only fire: IA2_EVENT_OBJECT_ATTRIBUTE_CHANGED This was the ultimate decision reached after lengthy discussion with NVDA. We continue to treat aria-hidden as a hint that AT can then use to decide what they want to convey.
I have made the correction for IA2 indicated in comment #5. But there is another issue with aria-hidden that came up in the ARIA TF meeting. Using an inspection tool, Joseph sees aria-hidden mapped to a property. This is not reflected in the text we have developed for the UAIG state and property mapping table where we just say that if the object is in the accessibility tree, you map all of the other properties as normal. Please clarify what aria-hidden is mapped to.
Opened Bug 14561 for the aria-hidden mapping in the state & property table.