[Webkit-unassigned] [Bug 38443] Flash video in iframe stays on top of everything (no matter what z-index)

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=38443





--- Comment #10 from Artur <arturadib at gmail.com>  2010-05-20 06:59:43 PST ---
Hi Jon,

(In reply to comment #9)
> The behavior of browsers here is varied; Firefox-win 3.6 puts the div on top, but in my testing, Firefox-win 3.5 doesn't, IE8 doesn't, and Opera-win 10.53 doesn't, so I don't see that WebKit on Windows is an outlier. Which versions of IE and Firefox did you test?


Sorry, the URL I gave hasn't been updated for my latest tests.  Please try again.  (The difference is that instead of an iframe and a div, now I'm using two iframes; this only seems to affect IE).

Here's the list of browsers and versions where (unlike WebKit-Win) the red box comes on top:

- IE8-Win (both in standards and compatibility mode, i.e. IE7-like)
- Firefox-Win 3.6.3
- Firefox-Mac 3.6.3
- Firefox-Mac 3.0
- Chrome-Mac 5.0.375
- WebKit-Mac 4.0.5 (5531.22.7)

Opera-Mac 10.10 behaves strangely; sometimes the red box comes on top, sometimes not.  (Every time I alternate between tabs I seem to get a different result...)

Opera-Win 10.53 displays the Flash video on top of the red box, like WebKit-Win.

I didn't mean to use this list as the main argument though; it was a side note.  Of course, the intended behavior should make sense regardless of what other browsers and versions are doing.


> On Windows, Flash can paint with or without its own window. If it uses a window for painting, this window is a child window of the WebView, so it will paint on top of all other content in the WebView. If it paints without using a window, it draws into a layer that WebKit can composite with the rest of the page.
> 
> This is controlled by the wmode attribute; see http://kb2.adobe.com/cps/155/tn_15523.html.


I understand this, and that's perfectly reasonable assuming one has control over wmode.

But that doesn't apply in the context I'm talking about: Since the framed content is from a third party, the wmode is beyond our control.

And *in this case*, I see no reason why a Flash object should behave differently from all other HTML elements.  (If this is meant as a solution to Clickjacking, that's another debate altogether. IMHO the solution should be consistent and part of a grander scheme, and not just limited to plugins).

Don't you agree that a more reasonable behavior is for the browser to override the third-party wmode so as to normalize the behavior across all types of elements?

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