Managing developer interruption

Abstract

The high frequency of interruptions during cognitively-intense activities can be annoying and detrimental to deadline-driven work, such as software development. When developers are interrupted they not only lose productivity from the time spent attending to the interruption but also from the time required to recover from it when they resume working. This thesis provides a solution that addresses the recovery process challenges. It focuses on the recovery of momentum based on the understanding that interruption recovery involves knowledge about the interrupted activity, the developer, as well as the context of the work. We designed FastRecovery, a tool in which data is collected while the developer is working normally. Once an interruption is detected, our tool processes the data assigning scores to each task in order to discover important moments. We propose a set of rules to include the most different types of tasks that a developer can perform. Thereafter, a curated video review is created to mitigate the effects of developer interruptions. A user study was conducted to evaluate the efficacy of our solution. Six participants used the tool for two periods of three days while we collected usage feedback along with self-reported impressions about the tool. The study results were positive and indicate that FastRecovery is beneficial with recovery from unexpected interruptions. Due to the small scale of the study these results are best characterized as an initial indication that our approach is promising with respect to interruption recovery.