A Constraint Extension to Scalable

S alable Ve tor Graphi s (SVG) is a language developed by the World Wide Web Consortium for des ribing two dimensional ve tor graphi s for storage and distribution on the Web. Unlike raster image formats, SVG-based images s ale ni ely to arbitrary resolutions and sizes. We introdu e a onstraint extension to SVG alled Constraint S alable Ve tor Graphi s (CSVG) that permits a des ription of an image that is more exible. With CSVG, an image an ontain obje ts whose positions and other properties are linearly related to other attributes via onstraints. For example, a re tangle an be spe i ed to remain above a ir le, and a line an be onstrained to onne t their enters. The various onstraints ea h have a spe i ed strength, and we use onstraint hierar hy theory to determine an appropriate solution. CSVG enables better layouts of diagrams for a wider variety of viewing onditions and provides support for de laratively spe i ed animation. We embedded our Cassowary onstraint solving toolkit in an existing SVG renderer to produ e a prototype implementation of a CSVG system.