Understanding Effects of Collaborations in Developing Mobile Computing Systems: Popularity, Efficiency, and Quality

Abstract

The development of mobile computing systems is an inherently collaborative activity, and the nature of the collaboration likely affects systems success. Collaboration can lead to faster development of systems and increased creativity, but it may also adversely affect the quality of systems. As one of the most popular and active Web services for developing open source software, Github has been a place of origin of many novel mobile computing systems with the Internet of Things, such as OpenEdge and CloudSim. Meanwhile, Github provides a place for people to collaborate with each other on developing novel mobile service computing systems. In this paper, we explore the effects of collaboration in developing systems hosted by Github and use a model of collaboration to detect bugs of systems aiming to enhance their quality. We conduct an at-scale study of more than five million projects on GitHub to reveal collaboration's effects on the popularity of software systems and speed of development. We find that having more collaborators results in higher popularity but slower development. Projects that use Objective-C are at an even more significant risk for lower development speeds than other projects. We define collaboration networks and co-commit networks, which capture the projects' collaboration patterns and use structural features of these networks to identify files within commits with bugs. By examining 18 large popular projects, we find that involving these features statistically significantly improves bug detection by 6.7% on average.

Obstacle Boundary Point and Expected Velocity-Based Flocking of Multiagents with Obstacle Avoidance

“…Te Bug algorithm is a kind of obstacle avoidance algorithm for an agent without knowing the global information of a map and obstacle shape, such as the Bug2 algorithm [27,28], the IBug algorithm [29], the FuzzyBug algorithm [30], and the RandomBug algorithm [31,32]. In this kind of algorithm, the agent frstly moves along the boundary of obstacles when encountering some, then leaves these obstacles based on certain judgment criteria, and then continues to move towards the target in a straight line.…”

Investigating diversity and impact of the popularity metrics for ranking software packages

“…Previous research has examined the impact of collaborative development style, firm involvement, and software documentation on the popularity of open source software projects. [3][4][5] We want to examine the relationship between development activity and popularity of software packages. Development activity measured by the number of commits describes the action developers take in order to contribute code/documentation in OSS projects.…”