Generational Layered Canvas Mechanism for Collaborative Web Applications

Abstract

We propose a Generational Layered (GL) canvas mechanism to reduce delays of synchronization for collaborative Web applications. The delays consist of network delays and drawing delays. The network delays are well-argued topic by various researches. We focus on the drawing delays, which are a primary bottleneck of synchronization of objects on canvas. In our previous work, we have proposed a Drawing-Frequency based Layered (DFL) canvas mechanism. The DFL canvas mechanism solved the drawing problem but requires to manually assign a parameter how frequently an object is redrawn, respectively. The GL canvas mechanism achieves the automatic assignment of canvas objects to elicit high performance when the drawing-frequency is unknown. The automatic assignment algorithm is inspired by the generational garbage collection. We implemented and evaluated the mechanism, and then the mechanism elicited higher performance up to 3.5 times faster when the canvas had the sufficient number of layers. Developers can create fast Web applications using the mechanism. The mechanism enhances the potential of the Web applications running within low performance devices such as tablets and smartphones.