[Webkit-unassigned] [Bug 109937] [EFL][WK2] Additional bits of having Accessibility in WebKit-EFL.

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=109937





--- Comment #11 from Krzysztof Czech <k.czech at samsung.com>  2013-03-08 07:25:09 PST ---
> Also, I'm not 100% sure (again, Pinheiro is the expert in AtkUtil), but I believe an implementation of AtkUtil is something _global_ to the process, not local to every WebPage as it seems to suggest this code, and so get_root() should return always the same and only object regardless of when it's invoked.
> 
> The problem here is that, as far I can understand now, we don't have that "clear root AtkObject" as in GTK (in GTK+ atk_get_root() returns the AtkObject associated to the GTK Application) but "many root objects" instead, one per WebPage. And we can't just create a global AtkObject being the parent of all the topmost AtkObject's (the AtkPlug's) in the WebProcess because that would break the implementation of AtkPlug in the ATK bridge, which expects atk_object_get_parent() to return NULL over the AtkPlug object.
> 
> So, perhaps that's why you added that dummy object in the first place, I don't know... but now I think more of this and see this problem, I wonder whether it was not a good idea after all, or even if this thing of having a "dynamic root object" will work fine.

You are perfectly right that AtkUtil is something global and WebKit should not even know about this. Would also be great if this was somewhere in EFL. I'm just trying to workaround this problem, that's why I'm trying to do it in WebKit.
Generally, more I'm getting into it, I'm closer to the conclusion that WebKit is not a good place for initializing ATK-BRIDGE stuff.

-- 
Configure bugmail: https://bugs.webkit.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.