Replication Package - Understanding the Time to First Response In GitHub Pull Requests

The pull-based development is widely adopted in
modern open-source software (OSS) projects, where developers
propose changes to the codebase by submitting a pull request
(PR). However, due to many reasons, PRs in OSS projects
frequently experience delays across their lifespan, including
prolonged waiting times for the first response. Such delays
may significantly impact the efficiency and productivity of the
development process, as well as the retention of new contributors
as long-term contributors.
In this paper, we conduct an exploratory study on the time-to-
first-response for PRs by analyzing 111,094 closed PRs from ten
popular OSS projects on GitHub. We find that bots frequently
generate the first response in a PR, and significant differences
exist in the timing of bot-generated versus human-generated
first responses. We then perform an empirical study to examine
the characteristics of bot- and human-generated first responses,
including their relationship with the PR’s lifetime. Our results
suggest that the presence of bots is an important factor contribut-
ing to the time-to-first-response in the pull-based development
paradigm, and hence should be separately analyzed from human
responses. We also report the characteristics of PRs that are more
likely to experience long waiting for the first human-generated
response. Our findings have practical implications for newcomers
to understand the factors contributing to delays in their PRs.