Bug 167484 – Fixed elements should not rubber-band in WK2, nor remain at negative offsets

• Status: RESOLVED
• Resolution: FIXED
• Component: New Bugs

[Simon Fraser (smfr)]

Fixed elements should not rubber-band in WK2, nor remain at negative offsets

[Simon Fraser (smfr)]

Created attachment 299899 [details]
Patch for EWS

[Simon Fraser (smfr)]

rdar://problem/29453068

[Simon Fraser (smfr)]

Created attachment 299993 [details]
Patch

[Simon Fraser (smfr)]

Created attachment 299995 [details]
Patch

[Build Bot]

Comment on attachment 299995 [details]
Patch

Attachment 299995 [details] did not pass ios-sim-ews (ios-simulator-wk2):
Output: http://webkit-queues.webkit.org/results/2961799

New failing tests:
compositing/geometry/fixed-in-composited.html
css3/filters/backdrop/dynamic-with-clip-path.html

[Build Bot]

Created attachment 299997 [details]
Archive of layout-test-results from ews121 for ios-simulator-wk2

The attached test failures were seen while running run-webkit-tests on the ios-sim-ews.
Bot: ews121  Port: ios-simulator-wk2  Platform: Mac OS X 10.11.6

[Dean Jackson]

Comment on attachment 299995 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=299995&action=review

Would be cool to have a test that combines this with scroll snapping.

> Source/WebCore/ChangeLog:18
> +        Also add a way to do unconstrained scrollTo(), so that a test can call window.scrollTo(-10, -10) to
> +        simulate rubberbanding.

👍

> Source/WebCore/page/FrameView.cpp:1812
> +    bool allowRubberbanding = fixedBehavior == StickToViewportBounds;

Isn't rubber-banding two words? allowRubberBanding.

> Source/WebCore/page/FrameView.cpp:1818
> +        bool rubberbandingAtRight = allowRubberbanding && (visualViewport.maxX() - layoutViewport.width()) > stableLayoutViewportOriginMax.x();

Why do you -layoutViewport.width() here? Don't you want to rubberband at the right-most edge? I obviously misunderstand.

> Source/WebCore/page/FrameView.cpp:1830
> +        if (!rubberbandingAtRight && layoutViewportOrigin.x() > stableLayoutViewportOriginMax.x())
> +            layoutViewportOrigin.setX(stableLayoutViewportOriginMax.x());

Just writing this to see if it reads better:

if (!rubberbandingAtRight)
  layoutViewportOrigin.setX(std::min(layoutViewportOrigin.x(), stableLayoutViewportOriginMax.x());

Hmmm... whatever.

[Simon Fraser (smfr)]

https://trac.webkit.org/r211379