[webkit-reviews] review denied: [Bug 59200] [Qt][WK2] Implement permission API for Qt port : [Attachment 113285] Fixed: initialization list, class/file names, coding style

bugzilla-daemon at webkit.org bugzilla-daemon at webkit.org
Wed Nov 2 01:09:51 PDT 2011


Simon Hausmann <hausmann at webkit.org> has denied Adenilson Cavalcanti Silva
<adenilson.silva at openbossa.org>'s request for review:
Bug 59200: [Qt][WK2] Implement permission API for Qt port
https://bugs.webkit.org/show_bug.cgi?id=59200

Attachment 113285: Fixed: initialization list, class/file names, coding style
https://bugs.webkit.org/attachment.cgi?id=113285&action=review

------- Additional Comments from Simon Hausmann <hausmann at webkit.org>
View in context: https://bugs.webkit.org/attachment.cgi?id=113285&action=review


Ok, so if this is an intermediate step, what do you envision as the way of
dealing with this on the application side?

I'd rather r+ when we have an idea where this is going.

What I'd like to see is an API that can be dealt with on the QML side.
Something along the lines of:

WebView {

    onPermissionRequested: {
	if (show a dialog)
	    request.accepted = true;
	else
	    request.accepted = false;

}

> Source/WebKit2/ChangeLog:-612
> -

Accidental removal of whitespace/newline?

> Source/WebKit2/UIProcess/API/qt/qdesktopwebview.cpp:509
> +    // For while, set it to true.
> +    request.setAccepted(true);

(it's "For a while" btw :)

> Source/WebKit2/UIProcess/qt/qsecurityorigin_p.cpp:32
> +class QtSecurityOriginPrivate : public QSharedData {
> +public:
> +    WKRetainPtr<WKSecurityOriginRef> m_origin;
> +    QtSecurityOriginPrivate() { };
> +};

So this seems a bit overkill, doesn't it? We have a shared pointer to a data
structure that contains another shared pointer :)


More information about the webkit-reviews mailing list