Bug 83771 – Move viewport meta handling til the web process side

• Status: RESOLVED
• Resolution: FIXED
• Component: WebKit2

[Kenneth Rohde Christiansen]

SSIA

[Kenneth Rohde Christiansen]

Created attachment 136878 [details]
Patch

[WebKit Review Bot]

Thanks for the patch. If this patch contains new public API please make sure it follows the guidelines for new WebKit2 GTK+ API. See http://trac.webkit.org/wiki/WebKitGTK/AddingNewWebKit2API

[Jocelyn Turcotte]

Comment on attachment 136878 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=136878&action=review

> Source/WebKit2/UIProcess/API/qt/qquickwebview.cpp:-710
> -    WebCore::restrictMinimumScaleFactorToViewportSize(attr, availableSize);
> -    WebCore::restrictScaleFactorToInitialScaleIfNotUserScalable(attr);

What was the use of those two calls? They disappeared now.

[zalan]

Comment on attachment 136878 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=136878&action=review

> Source/WebKit2/UIProcess/API/qt/qquickwebview.cpp:213
>  }

I'd swap updateViewportSize() and drawingAres()->setSize(). updateViewportSize() calls _q_contentViewportChanged() which sets the visiblerect on the drawingArea. There could be some side effect to set the visiblerect, when the size is not yet set properly.

> Source/WebKit2/WebProcess/WebCoreSupport/WebChromeClient.cpp:-715
> -

Viewport property change callback is now bundled with TILED_BACKING_STORE. Is it intentional?

> Source/WebKit2/WebProcess/WebPage/WebPage.cpp:885
>  }

WebChromeClient::dispatchViewportPropertiesDidChange() and  WebPage::setViewportSize() are very lookalike now. Wouldn't be better to move the common code to a dedicated function?

[Kenneth Rohde Christiansen]

(In reply to comment #3)
> (From update of attachment 136878 [details])
> View in context: https://bugs.webkit.org/attachment.cgi?id=136878&action=review
> 
> > Source/WebKit2/UIProcess/API/qt/qquickwebview.cpp:-710
> > -    WebCore::restrictMinimumScaleFactorToViewportSize(attr, availableSize);
> > -    WebCore::restrictScaleFactorToInitialScaleIfNotUserScalable(attr);
> 
> What was the use of those two calls? They disappeared now.

True and that might need to be revisited later. Currently a lot of additional code in Grob were due to us calling these (the implicit/explicit). That code were related to fit to width.

[Kenneth Rohde Christiansen]

Created attachment 136890 [details]
Patch

[zalan]

lgtm

[Simon Hausmann]

Comment on attachment 136890 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=136890&action=review

> Source/WebKit2/ChangeLog:9
> +        we now do everything on the web process side, and just sends

sends -> send

[Kenneth Rohde Christiansen]

Landed in 113974