Qualitatively Analyzing PR Rejection Reasons from Conversations in Open-Source Projects

Software developers have largely relied on pull requests as a mechanism of collaboration in their projects. Researchers have collected and analyzed pull request data in different ways for different reasons. In particular, we have qualitatively analyzed pull request conversation data to understand the main reasons for pull request rejection from a developer’s perspective. In this paper, we report results from ongoing research on identifying and categorizing pull request rejection factors. Two software developers, co-authors of this paper, manually analyzed 605 rejected PRs from Hexo and ESLint. We found that the most frequent reasons for PR rejection may vary depending on the project size and popularity. Still, some common rejection factors include implementing unnecessary functionality, conflicting PRs, agreement to make PR reattempts, and inactivity. Code quality issues are not among the most frequent reasons.

Transcript

1. Qualitatively Analyzing PR Rejection Reasons from Conversations in Open-Source Projects Tanay Gottigundala, Siriwan Sereesathien, Bruno da Silva bcdasilv@calpoly.edu CHASE Conference, May 2021
2. We focus on all types of PR authors Rejected PRs only Qualitative investigation
3. We manually analyzed 605 rejected PRs RQ1: most frequent reasons why PRs are rejected RQ2: PR rejection reasons on projects in sizes and popularity RQ3: Are code quality issues among the most frequent reasons for rejection?
4. We manually analyzed 605 rejected PRs (450) (155)
5. RESULTS
6. RQ1: most frequent reasons why PRs are rejected PR Reattempt
7. RQ1: most frequent reasons why PRs are rejected PR Reattempt Closed due to inactivity
8. RQ1: most frequent reasons why PRs are rejected PR Reattempt Closed due to inactivity Unnecessary functionality
9. RQ1: most frequent reasons why PRs are rejected PR Reattempt Closed due to inactivity Unnecessary functionality Lack of team consensus
10. RQ1: most frequent reasons why PRs are rejected PR Reattempt Closed due to inactivity Unnecessary functionality Lack of team consensus Unnecessary functionality
11. RQ1: most frequent reasons why PRs are rejected PR Reattempt Closed due to inactivity Unnecessary functionality Lack of team consensus Unnecessary functionality PR Con fl icts
12. RQ2: PR rejection reasons on projects in sizes and popularity Most of the differences in PR rejection reasons stem from distinct project size, popularity, and consequent level oforganization and strictness for managing PRs
13. RQ2: PR rejection reasons on projects in sizes and popularity Most of the differences in PR rejection reasons stem from distinct project size, popularity, and consequent level oforganization and strictness for managing PRs The only common reason among their top fi ve reasons was “Unnecessary functionality”
14. RQ3: Are code quality issues among the most frequent reasons for rejection?
15. Qualitatively Analyzing PR Rejection Reasons from Conversations in Open-Source Projects Tanay Gottigundala, Siriwan Sereesathien, Bruno da Silva bcdasilv@calpoly.edu CHASE Conference, May 2021