Eureka! Institutional Repository

Both the web and the desktop are powerful channels nowadays to satisfy user needs over the Internet and the office/home work respectively. In both areas complexity occurs due to the complexity and dynamism modern applications enforce. In this systematic literature review work we present a range of papers published since 2000 in both fields of testing. Based on our selection protocol a specific set of papers are extracted seeking to provide answer to four research questions. The results of our research interpret the trends in the above questions. The first question found that only 48 of the 125 documents (38%) had any testing tools. In both areas there are studies focusing on reverse engineering techniques (from code or script to a model) that reverses approximately to the original design (33%) for error identification and classification. In the second question, there is a wider range of specialized testing techniques in desktop (functions, interfaces, durability, and performance) rather than in web applications. The third question states that empirical studies have been carried out in both areas however for different purposes. As in the fourth research question, there are few studies focusing on describing state-of-the-art technology related to software testing. There is also a significant discrepancy between objectives in the two areas (e.g. tools, processes and future technological challenges in web applications versus creating test cases in desktop applications). Interestingly we conclude that web testing dominates research in contrast to desktop testing. Results also provide useful lessons for the types of test models, the fault models/bug taxonomy for the web applications, test tools for web application testing as well as test tools for desktop application testing and their capabilities, procedures and methodologies for performing desktop applications, metrics used for evaluating cost and time constraints in both types of testing, state of the art and future trends in both types.