Compatibility Issues in Android: Characterisation and Detection

Android has become the most popular mobile platform with over 2.5 billion active users who use many different languages across many different countries. Since it is open-sourced and rapidly evolved, fragmentation, induced by different device vendors and rapid evolution of itself, has become a serious problem in the whole Android ecosystem, which affects consistent running of Android Apps. In addition, the app per se could also give a bad user-experience if it only provides a single or several different languages. Therefore, we present different approaches to strengthen the quality of Android apps to provide app users a consistent user-friendly experience from two different aspects.

Extensive fragmentation induces severe compatibility issues. We first proposed \androsea{} to identify silently evolved methods, which have the same comment but different implementation between two consecutive Android releases and successfully identified 4,769 silently evolved methods between ten most used Android releases, including 2,271 publicly-accessible methods. To go step further, we also proposed \andromevol{} to successfully quantify 333,952 incompatibility APIs between 11 different releases and 6 different vendors, including 191,546 methods and 142,406 fields. To have a better understanding of the fragmentation, we also investigated how these customizations bring in updates from the official newer releases. We found that only a small portion of version releases of the official Android OS projects were merged into their customization releases.

Project Team

Prof John Grundy, Pei Liu (PhD Candidate), Prof Li Li

Papers

• Project summary
• Liu, P., Zhao, Y., Fazzini, M., Cai, H., Grundy, J., Li, L. Automatically Detecting Incompatible Android APIs, ACM Transactions on Software Engineering and Methodology (TOSEM), 2023
• Liu, P., Li, L., Liu, K, McIntosh, S., Grundy, J, Understanding the Quality and Evolution of Android App Build Systems, Journal of Software: Evolution and Process, 2023
• Liu, P., Xia, Q., Liu, K., Guo, J., Wang, X., Liu, J., Grundy, j., Li, L. Towards Automated Android App Internationalisation: An Exploratory Study, Journal of Systems and Software (JSS), 2022
• Liu, P., Zhao, Y., Haipeng, C., Fazzini, M., Grundy, J, Li, Li.  Automatically Detecting API-induced Compatibility Issues in Android Apps: A Comparative Analysis (Replicability Study).  ISSTA 2022.
• Liu, P., Fazzini, M., Grundy, J, Li, Li. Do Customized Android Frameworks Keep Pace with Android? MSR 2022.
• Liu, P., Li, L., Yan, Y., Fazzini, M., Grundy, J. Identifying and Characterizing Silently-Evolved Methods in the Android API.  ICSE-SEIP 2021.
• Liu, P., Li, L., Zhao, Y., Grundy, J. AndroZooOpen: Collecting Large-scale Open Source Android Apps for the Research Community.  MSR 2021.

Tool

• AndroZooOpen
  AndroZooOpen is a growing collection of Open-Source Android Applications collected from several sources, including Github, Gitlab and Bitbucket. It currently contains 77,406 different projects. We provide this dataset to contribute to ongoing research efforts, as well as to enable new potential research topics on Android Apps.