Software engineering abstractions for the multi-touch revolution

Multi-touch interfaces allow users to use multiple fingers to provide input to a graphical user interface. The idea of allowing users to touch and manipulate digital information with their hands has been subject of research for more than 25 years. Recently several of these research artifacts have found their way to industry, with examples like the iPhone and the Microsoft Surface. Mainstream programming languages do not offer support to deal with the complexity of these new devices. Unlike the evolution in the hardware technology, the complexity of these new devices has not yet been addressed by adequate software engineering abstractions. Current multi-touch frameworks provide a narrow range of hardcoded functionality like pinch, rotate and move known as multi-touch gestures. There is however a substantial need to develop new and more gestures for domain specific applications. Multi-touch devices are inherently concurrent and provide a continuous stream of events. In many of these frameworks capturing these events to extract gestures is done by means of event handlers. Programming multi-touch devices with event handlers is cumbersome for a number of reasons.