Uma investigação sobre a influência da prática de atribuição de rótulos específicos para novatos em repositórios de código aberto

Abstract

Context: Contribute to open source projects for the first time can be challenging. Sometimes, newcomers end up giving up of joining these projects because of the difficulty in finding a task that they feel comfortable to contribute. Aiming to attract new contributors, some projects like Up-ForGrabs are listing open source repositories, hosted on GitHub, that practice assigningment of specific labels to indicate that a task is recommended for picking by newcomers. Since free software are maintained oftentimes by their contributors, practices that can lead to attraction of new contributors are more than welcome and deserve attention. Currently, initiatives such as Up-For-Grabs, assume that the labeling of tasks recommended for newcomers is a practice that can influence positively the attractiveness of newcomers into open source projects. However, to date there is little documented evidence to support this hypothesis. Objective: In view of this, the objective of this work is to investigate if the labeling of tasks as “recommended for first contribution” actually increases the attractiveness of newcomers to the free software projects that use it in their repository. Method: To verify whether the above-mentioned hypothesis is true, comparative analysis was performed through non-parametric statistical tests applied to a set of 1060 projects out of 10 different languages and whose code is open source and available on GitHub. Of the collected samples, the distributions referring to the weekly ingress of newcomers before and after the adoption of practice were statistically compared. Complementarily, the distributions referring to the total number of newcomers from projects adopting the practice were also compared with non-adopters. Results: It had been discovered that there is a wide range of terms being used by the community as labels for newcomers and these terms are not always intuitive. It was also noticed that a significant portion of projects have been using labeling to indicate recommended tasks for newcomers. Statistical tests showed that 63% of the projects adopting the practice had a significant increase in the weekly entry of newcomers after adopting the practice. The tests also showed that projects who use newcomers labels have a greater number of newcomers than projects who don’t. Conclusions: The results obtained in this research showed that the use of labels to indicate tasks for newcomers can significantly increase the attractiveness of newcomers in open source projects. However, there is still room for improvement in the practice, especially regarding to the terms used in labels for newcomers.