Nothing Happens, is a networked online performance in which viewers work together to make a series of objects tip over. The performance consists of three acts, which are centered around staged environments - a high shelf, a cluttered tabletop and an empty floor. Each scene contains a central protagonist, respectively: a cardboard box, a glass full of water and a wooden chair. In all three acts, web-enabled physical devices, controlled by viewer’s clicks, make these objects tip over. When this happens, the performance is over.

The website acts as a go between, allowing physically distant observers a chance to participate. In one direction, the site displays live video stream in real-time of the current act as it unfolds. In the other direction, users are able to click a simple interface in order to manipulate the scene. Web-enabled physical devices translate every click into a minute but noticeable physical change in the scene. In the case of the shelf act, a small-motorized device rotates with each click, pulling the box off of the shelf, until it tips over. In the case of the tabletop, each click causes a miniature device to push the glass over until its content spills. Finally, in the chair scene, a small mechanical device responds to clicks by expanding, tipping the chair further back.

Every users click thus effects the object to move in small steps. Each change is recorded as a snapshot-image, creating not only an archive of the work – so that viewers can browse through the entire history of the performance both during it and after its conclusion – but when each performance is concluded these images are combined to a collective creative result: a stop-motion-animation sequence.

The key aim of interactivity in this performance is to create an immediate understandable form of interaction, so that each click of a user is rightfully perceived as progressing the scene further. My aim is to make the scenes last for at least one thousand clicks, if not more, so that a sense of time is drawn out. The expected duration of each act of the performance is between one and two weeks, although actual timing depends on user participation.