Section 4.5: Gesture Engine Details

The Gesture Engine follows certain rules when deciding which gestures to fire based on the Multi-Touch input.

First, for Tap gestures, it possible for Taps near the corners and edges of the screen to trigger several different Tap Events. The most specific location is tried first; for example, a tap near the upper-left corner of the screen will first try firing an Upper-Left Tap Event. If an Action is bound to this Event, then that event fires, and Event processing is complete. If no Actions are bound to a specific Event, then the next-most-specific locations are tried, until an Event with a bound Action is encountered; in the example, the Upper Edge, then the Left Edge, then the "Anywhere" locations are tried respecively.

The other rule is similar, and has to do with gesture combinations: whenever a sequence of gestures is recognized, the most specific (i.e., longest) sequence is tried first. So, for example, consider the case of a Swipe-Tap gesture: if the Swipe-Tap is configured to trigger an Action, then the Event fires and the gesture processing is complete. If no Action is bound, however, then the component gestures are tried: first, the Swipe; and then, the Tap.

In any case, if there is no gesture Event that matches and has an associated Action, then no gesture is recorded (and no feedback is provided).