Como novatos agem ao fazer a sua primeira contribuição com teste

Abstract

Context: A widely used method to contribute to a free software project is the pull requests. Writing tests is a way to make it more reliable. A newcomer in a project may not be aware of this, and will not write test, causing their pull request to be rejected. Objective: Try to understand how newcomer actually relate to tests. Method: Data collection of pull requests, classification according to the characteristics involved, such as being a newcomer and having test. Finding in this data, numbers that represent how much test the newcomers do. Subsequently, perform a manual analysis of a random sample of pull requests from these beginners and try to find common characteristics. Finally, to launch a questionnaire to receive the direct opinion of those who were once newcomer. Results:The use of tests varies greatly from one language to another, although when the test exists, the acceptance rate of pull requests is much higher than the general acceptance rate of pull requests. Although they are new to the analyzed projects, these users are not, for the most part, newcomers to GitHub, having previously contributed to another project. Because of this, newcomers who wrote tests, in the majority, wrote their tests on their own, without another person having to ask him to do it. Conclusions: Those who do not really know the importance of testing, proved to be a relatively small number among the newcomers. They have difficulty testing their first contribution, even if it is not the first of their life, as each project may take a different approach.