Research issues in the automated testing of Ajax applications

Abstract

There is a growing trend to move desktop applications towards the web. This move is made possible through advances in web technologies collectively known as Asynchronous JavaScript and XML (ajax). With ajax, the classical model of browsing a series of pages is replaced by a JavaScript engine (running in the browser) taking control of user interaction, exchanging information updates with the web server instead of requesting the complete next page. The benefits of this move include no installation costs, automated upgrading for all users, increased interactivity, reduced user-perceived latency, and universal access, to name a few. ajax, however, comes at a price: the asynchronous, stateful nature and the use of Javascript make ajax applications particularly error-prone, causing serious dependability threats. In this paper, we evaluate to what extent automated testing can be used to address these ajax dependability problems. Based on an analysis of the current challenges in testing ajax, we formulate directions for future research. © 2010 Springer-Verlag Berlin Heidelberg.