OS enhancement for zero-latency UI response

点击翻译按钮获取中文摘要

Users have an expectation of responsiveness when they make input to a computer. The slower the device responds, the greater the impact on the user's performance, and the greater their frustration. Until recently, the developers of interactive computing software, such as applications and operating systems, have worked under an assumption that ever-increasing processor speeds and decreasing costs will enable fast, responsive computers capable of running increasingly complex code. This is based largely on Moore's Law, a prediction made in 1965 by the co-founder of Intel that the number of transistors on a chip will continuously double every two years. While this has proved largely to be true, in 2011, a tipping point was reached: for the first time, worldwide sales of smartphones and tablets outpaced more powerful desktop and laptop computers. This new focus on mobile devices has caused two important changes for the developers of interactive computers. First, because these devices conserve power for the sake of mobility, their processing capabilities are set back several generations of Moore's doubling. Second, because these devices are used for media consumption, they are designed to maximize screen size, which has forced the adoption of input made directly to the screen (commonly called "multi-touch", first developed at the University of Toronto). Multi-touch input compounds slow responsiveness, because users are able to see a physical separation between an object being manipulated and their finger. Our preliminary work has found that the current high levels of latency reduce user performance by over 25%. Tactual's work is in the elimination of all latency from interactive systems. This project has three foci: understanding the limits of human perception of latency (how low must we go?), the design of UI software which hides latency from the user, and, finally, the design of UI software which takes advantage of zero-latency systems developed by Tactual.