[Webkit-unassigned] [Bug 156774] AX: SVG images are not exposed in the AX api

bugzilla-daemon at webkit.org bugzilla-daemon at webkit.org
Tue Apr 19 19:17:34 PDT 2016


--- Comment #4 from Joanmarie Diggs (irc: joanie) <jdiggs at igalia.com> ---
(In reply to comment #3)
> (In reply to comment #2)
> > Nan: Can you show me where in the SVG spec it states that the property on
> > the parent should cause a shape to be exposed?
> I don't think the spec has included this case. But we have bugs that svg
> root is labelled but shape in it is not exposed. 
> So you think we should not expose it this way?

I think we should follow the spec. If you don't like what the spec says (I have had those occasions myself), then you can raise an issue with the task force (which I've done myself). :)

The particular case I question is Circle2. Looking at the specs which govern this implementation:

5.1.1 Excluding Elements from the Accessibility Tree

States the following:

     The following elements in the SVG namespace should not be included
     in the accessibility tree, except as described under Including Elements
     in the Accessibility Tree:

     * shape elements (circle, ellipse, line, path, polygon, polyline, rect)

Looking at the referenced section: 5.1.2 Including Elements in the Accessibility Tree

It lists criteria for inclusion. But having an *ancestor* with a global ARIA property isn't one of them. That section then points to the Core AAM, but Core AAM doesn't say that an ancestor with a global ARIA property requires insertion.

So unless I'm missing something, you're introducing a change which will cause our implementation to not comply with the W3C specification we are implementing.

Again, that doesn't mean the spec is right. And if you have a valid reason for thinking they are wrong, now is the time to tell them (i.e. while they're still working on the spec). :)

You are receiving this mail because:
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.webkit.org/pipermail/webkit-unassigned/attachments/20160420/47c5bf47/attachment.html>

More information about the webkit-unassigned mailing list