- From: Joseph Scheuhammer <clown@alum.mit.edu>
- Date: Mon, 28 Oct 2013 14:11:12 -0400
- To: Alexander Surkov <surkov.alexander@gmail.com>, Richard Schwerdtfeger <schwer@us.ibm.com>
- CC: Joseph Scheuhammer <clown@alum.mit.edu>, David Bolter <dbolter@mozilla.com>, W3C WAI Protocols & Formats <public-pfwg@w3.org>
> Hi, Rich. I missed what kind of problem here.
I *think* the problem is that 90, 91, and 92 are testing a "selection
follows focus" scenario. However, as Rich noted, nothing in the test
file changes the focus -- that's why these test files are invalid.
But, if the test file was modified such that focus did change, what is
observed in terms of events?
Using test case 91 as an example: the setup is a multiselectable list
with two options with no option selected. At a certain point in time,
focus moves to the second option. Consequently, according to the
"selection follows focus" model, the second option is also marked
aria-selected="true".
According to the UAIG, for selection follows focus, under MSAA+IA2:
"EVENT_OBJECT_SELECTION and EVENT_OBJECT_STATECHANGE on newly focused
item, but arrange events so state change does not occur on focused item,
to avoid extra selection change announcements".
However, Rich is observing:
> ... both the state change and the selection change occurs on that element.
--
;;;;joseph.
'A: After all, it isn't rocket science.'
'K: Right. It's merely computer science.'
- J. D. Klaun -
Received on Monday, 28 October 2013 18:11:43 UTC