diff options
| author | 2009-06-26 11:36:03 -0700 | |
|---|---|---|
| committer | 2009-06-26 11:36:03 -0700 | |
| commit | 53d3c1e4513135c5d8f70da4938c7d25455263f5 (patch) | |
| tree | 8de7779644de57a9f5576382fc121f5efe5fccda | |
| parent | ef7b5d6da57b434a739c8e8261dab9380a57e2f9 (diff) | |
Fine tune TOUCH_SENT_INTERVAL for the high speed devices.
| -rw-r--r-- | core/java/android/webkit/WebView.java | 13 |
1 files changed, 11 insertions, 2 deletions
diff --git a/core/java/android/webkit/WebView.java b/core/java/android/webkit/WebView.java index 97134d22807c..5f002491865b 100644 --- a/core/java/android/webkit/WebView.java +++ b/core/java/android/webkit/WebView.java @@ -329,9 +329,18 @@ public class WebView extends AbsoluteLayout /** * The minimum elapsed time before sending another ACTION_MOVE event to - * WebViewCore + * WebViewCore. This really should be tuned for each type of the devices. + * For example in Google Map api test case, it takes Dream device at least + * 150ms to do a full cycle in the WebViewCore by processing a touch event, + * triggering the layout and drawing the picture. While the same process + * takes 60+ms on the current high speed device. If we make + * TOUCH_SENT_INTERVAL too small, there will be multiple touch events sent + * to WebViewCore queue and the real layout and draw events will be pushed + * to further, which slows down the refresh rate. Choose 50 to favor the + * current high speed devices. For Dream like devices, 100 is a better + * choice. Maybe make this in the buildspec later. */ - private static final int TOUCH_SENT_INTERVAL = 100; + private static final int TOUCH_SENT_INTERVAL = 50; /** * Helper class to get velocity for fling |