Section: New Software and Platforms

GitHub Label Analyzer

URL: http://atlanmod.github.io/gila/

Reporting bugs, asking for new features and in general giving any kind of feedback is the easiest way to contribute to an Open-Source Software (OSS) project. In GitHub, the largest code hosting service for OSS, this feedback is typically expressed as new issues for the project managed by an issue-tracking system available in each new project repository. Among other features, the issue tracker allows creating and assigning labels to issues with the goal of helping the project community to better classify and manage those issues (e.g., facilitating the identification of issues for top priority components or candidate developers that could solve them). Nevertheless, as the project grows a manual browsing of the project issues is no longer feasible.

We believe that visualization techniques could be applied here to overcome this challenge. In particular, we have created GiLA, a tool to better understand how labels are being used in GitHub projects, with the aim of providing more insights into how such projects are being managed. GiLA provides three visualizations addressing three different viewpoints, specifically:

V1

Label usage, which helps to identify the most used labels and which ones are commonly used together.

V2

User involvement, which allows discovering the most active and knowledgeable users around each label.

V3

Typical Label timeline, which provides some insights about how issues under that label evolve over time (e.g., time to be treated).

The tool can be used to explore these viewpoints on all the original projects (i.e., projects that are not a fork of a previous project) in GitHub. We believe that the results favour not only a better comprehension of the project but also help in its advancement, e.g., by helping to quickly identify experts on a particular topic/label.