Presented at the IEEE ComSoc/Computer Society Boston Chapter Meeting, May 21, 2001
XML, XSL, and application servers enable application developers to separate user interface code from application code. For a quality user interaction, the user interface code needs to know at least the broad parameters of the terminal device. For example, the presentation of a web page on a relatively large monitor can be quite different from the presentation of the same data on a WAP browser (wireless phone). Simply mapping a 1600×1200 bit-mapped display to a two line by forty-character display is not feasible. This issue becomes more pronounced when one introduces voice user interfaces into the mix.
The talk first reviews the application server architecture. A brief introduction to VoiceXML, the DTD of XML for voice user interfaces, follows. The talk will discuss issues with voice user interfaces and why their specification is so different from graphical user interfaces. With this foundation, we demonstrate the power of separating the user interface fully from the application logic by comparing the interfaces for a set of applications that use a traditional web, WAP, and voice browser user interface.