Keynote 1: LGTM - Software Sensing and Bug Smelling

Summary form only given. Looks Good To Me (LGTM) is a means to stamp an ´ok´ on code reviews and then have the code moved to production. One key part of this process is a detailed quality analysis by checking code, looking for potential known bugs, and evaluating the design. As code bases are changed almost every day by many developers, we need to devise means for effective, regular, and focused analysis. In the recent past we have looked into approaches of software sensing and bug smelling. Sensing software is employed by audio-visual and multi-touch interfaces to a code base and its change history. Bug smelling leverages bug prediction on the levels of classes, methods, or change types. Combining both can lead to a more effective quality analysis for reviewing tasks such that the signing off by LGTM is facilitated. In this talk we will present approaches and tools such as SmellTagger or Coco Viz for software sensing and bug smelling. We will also discuss current limitations and potential new horizons for software evolution research.