1. go to http://people.mozilla.com/~nhirata/html_tp/jumptag.html 2. click on the link "Go to Bottom" Expected: Goes to the bottom of the page Actual: in some cases it doesn't move. HTC Flyer; 20111122; Android 2.3

related to Bug 703620 ?

In the test URL on this bug, I often end up flinging/snapping the page when clicking on the "bottom" link to go back to the top. This was causing the viewport to get clobbered by the FlingRunnable, so I needed this patch to make it work.

Comment on attachment 578056 [details] [diff] [review] (1/2) Force viewport update on anchor jump Review of attachment 578056 [details] [diff] [review]: ----------------------------------------------------------------- r+, but would prefer to see the comment below addressed. ::: mobile/android/chrome/content/browser.js @@ +1185,5 @@ > + gecko: { > + type: "Viewport:Update", > + viewport: JSON.stringify(this.viewport) > + } > + }); Why doesn't this use sendViewportUpdate? I would suggest refactoring sendViewportUpdate so it can be used for things like this, so we don't have these last few lines in multiple places.

Comment on attachment 578057 [details] [diff] [review] (2/2) Abort flings on viewport update Review of attachment 578057 [details] [diff] [review]: ----------------------------------------------------------------- r+, with the comment addressed (either via a rationalisation or a code change, either is fine by me) ::: mobile/android/base/ui/PanZoomController.java @@ +133,5 @@ > public void geometryChanged() { > populatePositionAndLength(); > } > > + public void viewportMetricsChanged() { Can we just have a boolean flag in geometryChanged, something like aCancelTouch or aReset (or some other better name that escapes me)? Having a separate function for what is essentially the same event, but from different sources could become confusing down the line.

Updated with suggested refactoring.

(In reply to Chris Lord [:cwiiis] from comment #5)> > Can we just have a boolean flag in geometryChanged, something like > aCancelTouch or aReset (or some other better name that escapes me)? Having a > separate function for what is essentially the same event, but from different > sources could become confusing down the line. Yeah, that makes more sense, fixed.

https://hg.mozilla.org/projects/birch/rev/ec36c2b415cc https://hg.mozilla.org/projects/birch/rev/cba7acb26616

Samsung Galaxy SII (Android 2.3.4) 20111202040206 http://hg.mozilla.org/projects/birch/rev/e2a54aafac18