Mobile Testing

Abstract

Mobile apps are everywhere. Some apps entertain and others enable business transactions. Apps increasingly interact with complex IT landscapes. For example, a banking app on a mobile device acts as a front end that invokes services on a back-end server of the bank, which might contact even more servers. Mobile testing becomes crucial and challenging. This paper follows a user-centric testing approach. The app's architecture matters for testing, as does its user base and usage context. Addressing these factors ensures that test cases cover all relevant areas. Most apps need test automation for two reasons: agility and compatibly. Agile projects test frequently, such as every night, to detect bugs early. Compatibility tests ensure that an app runs on all relevant devices and operating system versions on the market. Thus, testers execute test scripts on many devices. This demands for a private device cloud and a mobile test automation framework. Swisscom IT Services followed this path, enabling us to address the major quality issues we identified for mobile apps: pre-usage failures (installation fails, app crashes during startup) and lack of basic regression testing (upgrades buggier than predecessor).

Mobile Application Testing Matrix and Challenges

“…Haller in [20] proposed agile and compatibility testing for one a case study from Swisscom IT services to identify whether test on wild device and failure during the start focused on regression testing process. Before decide to adopt any test techniques on the mobile apps, it is necessary to have testing order to meet user's requirements, specifications and to avoid negative feedbacks from app's users.…”

“…Haller in [20] proposed agile and compatibility testing for one a case study from Swisscom IT whether test on wild device and failure during the start-up application and focused on regression testing process.…”

“…Figure 4 ment preparation to the Haller in [20] proposed agile and compatibility testing for one a case study from Swisscom IT up application and Amalfitano et al highlighted the results of automated testing experiment for Android mobile application platform [1]. The author specifically presents a technique for rapid crash testing and regression testing based on crawler that automatically builds models of application in GUI.…”

“…Therefore, the comprehensive mobile application testing is crucial to direct high quality of applications and satisfies user needs, whereas studies indicated that developers are more focusing on the application back end and functionality rather than use experiences. In fact, a user feedback is one of the fundamental parts of application's reputation to ensure app's owners with successful or failure of their application [20]. Commonly, users can easily drop interesting in problematic mobile app, and will abandon it after only one or two failed attempts.…”

Automating UI tests for mobile applications with formal gesture descriptions

“…Device-specific factors include properties that are inherently bound to different handsets, such as operating system, form factor (tablet computer vs. smartphone), display size and orientation. These factors multiply the testing effort, as the typical testing process (unit-tests, integration tests, smoke tests, system tests, acceptance tests and regression tests) is extended by device tests and testing in the wild [9], where the AUT is deployed on a set of handsets and executed under the appropriate conditions. This may include visiting off-site locations, provoking roaming scenarios and traveling long distances.…”