I missed this thread, but came to the same conclusion independently today. I also filed ISSUE-621. PFWG-ISSUE-621 (selection conflict): Spec conflict: UAIG needs to not require UAs to fire selection change events unless aria-selected is changed [ARIA 1.0 Implementation Guide] https://www.w3.org/WAI/PF/Group/track/issues/621 FWIW, I also updated the test files today to include managing focus and selection state changes via aria-selected. If you rerun your tests, I expect they will pass. They pass in WebKit now, too. On Sep 19, 2013, at 9:16 AM, Richard Schwerdtfeger <schwer@us.ibm.com> wrote: > Hi Joseph, > > For grid rows and gridcells the ARIA specification requires that the author set the aria-selected state and that selection does not follow focus unless the programmatically sets the aria-selected state. So, there is no requirement that selection follow focus for these roles in the ARIA spec. > > However, what is confusing is that the UAIG in section 5.8.3 has mappings for selection following focus and for when the author sets aria-selected directly. Michael and I discussed this and we agree that you appear to have added an additional requirement beyond the ARIA spec. that selection MUST follow focus for gridcells and rows in a grid. > > Additionally, it appears that tabs in a tablist may also need to have selection follow focus in the UAIG. > > If this is what the UAIG intended (selection must follow focus) for rows, gridcells, and tabs then Firefox fails all these instances for the UAIG test suite. > > The only place where I see that selection follows focus is in a listbox. > > Please provide a decision on this. A clarification might be to add text as to which roles selection must follow focus in the UAIG. I will tell you that the ARIA spec. is very clear on the fact that aria-selected following focus on tablist tabs is a SHOULD vs. a MUST. > > Thanks, > Rich > > Rich Schwerdtfeger >
Received on Tuesday, 1 October 2013 17:18:17 UTC