[Webkit-unassigned] [Bug 109422] [Qt] Add Page Visibility API support

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=109422





--- Comment #34 from Simon Hausmann <hausmann at webkit.org>  2013-03-11 06:46:59 PST ---
(In reply to comment #29)
[...]
> > I suggest to re-implement showEvent() and hideEvent(), pass on the event to d->page->event and call the base implementation. That gives advanced users the power to control the state if they need to (re-implement and call QWebView::showEvent or not if they like), and it still allows us to centralize the implementation inside QWebPageAdapter.
> > 
> 
> It's an HTML5 requirement, why should we want to give to advanced users the power to control the state and thus change the expected behaviour defined by W3C?

I don't see that requirement. Can you elaborate?

I do see quite the opposite. Take the examples of the "hidden" attribute for example: "The User Agent is not minimized, but the page is on a background tab."

We can _not_ provide a fully comprehensive implementation of this attribute that works out of the box in all cases. We do _need_ to give advanced users control over providing the visibility information to the web application.

> > > Source/WebKit/qt/tests/qgraphicswebview/tst_qgraphicswebview.cpp:140
> > > +    int waitForEvent = 5;
> > 
> > I can almost guarantee you that this is going to fail on the bots. Randomly failing tests on the bot are a PITA.
> > 
> > I suggest to replace this unit test with implementing page visibility support in the Qt DRT. Then we can enable the layout tests for it as well.
> 
> Yes I agree with you but I don't know how to do it differently...
> What is the Qt DRT? Where can I find an example?

It's the shell that is used to run the layout tests. You can find it in Tools/DumpRenderTree with the qt specific bits in Tools/DumpRenderTree/qt.

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