[webkit-reviews] review denied: [Bug 60893] [Qt] tst_QWebFrame::render() failing : [Attachment 94208] Now it waits the content to be loaded before testing and make sure QPicture will expand if its painter is used to render.

[bugzilla-daemon at webkit.org]

Luiz Agostini <luiz at webkit.org> has denied Rafael Brandao
<rafael.lobo at openbossa.org>'s request for review:
Bug 60893: [Qt] tst_QWebFrame::render() failing
https://bugs.webkit.org/show_bug.cgi?id=60893

Attachment 94208: Now it waits the content to be loaded before testing and make
sure QPicture will expand if its painter is used to render.
https://bugs.webkit.org/attachment.cgi?id=94208&action=review

------- Additional Comments from Luiz Agostini <luiz at webkit.org>
View in context: https://bugs.webkit.org/attachment.cgi?id=94208&action=review

> Source/WebCore/ChangeLog:12
> +	   If you use a QPicture's painter to render a themed scrollbar,
> +	   as it does a translation to a certain point before painting,
> +	   QPicture's bounding rect may not be expanded because the new
> +	   clipping region may be inside its bounding rect in the original
> +	   coordinate system (before translation).

This change log is very confusing, specially after reading your comments in the
bug.
Is it a problem with QPicture, with the QStyle in use in your env or with
QtWebKit?

As the problem does not show up in all machines I would say that it is in your
env.

Did you try using QWindowsStyle in the test? I think that this together with
the waitForSignal may solve the problem and we would be sure that the test will
always use the same style in all platforms. 

If you see problems in QPicture or in the QStyle that you use, you should file
bugs for them.

Anyway, this change log needs to be improved a lot. :)