osscontribs

Over 30 years of open source project activity, packaged for R.

This dataset covers FreeBSD, OpenBSD, NetBSD, and PostgreSQL — four foundational projects that have shaped modern computing. The data comes from two sources:

Git repositories (cloned and parsed locally) — every commit, every author, no API limits
FreeBSD Phabricator — account signups on the code review platform

No personal information is included. Authors are counted, not identified.

Why this exists

Most open source research relies on GitHub API data, which caps results at the top 100 contributors per repo. These datasets bypass that limitation entirely by extracting commit logs directly from cloned repositories. The result is a complete picture of project activity going back to the early 1990s.

Good for: time series analysis, growth modeling, changepoint detection, cross-project comparison, or just satisfying curiosity about how these projects evolved over three decades.

Installation

# devtools::install_github("chrislongros/osscontribs")
install.packages("osscontribs", repos = c('https://chrislongros.r-universe.dev', 'https://cloud.r-project.org'))

Datasets at a glance

Dataset	Rows	Granularity	Source	What it measures
`oss_commits`	2,068,717	Per-commit	Git repos	Every commit with timestamp and author
`oss_daily_commits`	44,939	Daily	Git repos	Commits per day
`oss_daily_authors`	2,932	Daily	Git repos	First-time committers per day
`oss_weekly_commits`	6,307	Weekly	GitHub API	Commits per week (top 100 authors)
`oss_contributors`	447	Monthly	Mixed	New signups per month
`oss_contributors_daily`	2,531	Daily	Phabricator	FreeBSD signups per day

From git repositories (complete data)

`oss_commits`

Every individual commit with full timestamp (to the second) and anonymized author ID. The most granular dataset — 2M+ rows across all four projects.

Column	Type	Description
`project`	character	freebsd, openbsd, netbsd, or postgresql
`timestamp`	POSIXct	Commit timestamp in UTC
`author_id`	character	Anonymized 12-character author hash

`oss_daily_commits`

Daily commit counts extracted from cloned repos. This is the most complete commit dataset — it includes every commit, not just top contributors.

Column	Type	Description
`project`	character	freebsd, openbsd, netbsd, or postgresql
`date`	Date	Commit date
`commits`	integer	Commits that day
`cumulative_commits`	integer	Running total

`oss_daily_authors`

Tracks when each unique author made their first commit to a project. Useful for measuring how quickly a project attracts new contributors.

Column	Type	Description
`project`	character	Project name
`date`	Date	Date of first commit
`new_authors`	integer	New committers that day
`cumulative_authors`	integer	Running total of unique committers

From GitHub API and Phabricator

`oss_weekly_commits`

Weekly commit data from GitHub's stats API. Limited to the top 100 contributors per project — superseded by oss_daily_commits for most uses, but kept for backward compatibility.

`oss_contributors` / `oss_contributors_daily`

Signup data — account registrations, not commit activity. Many users sign up but never commit, so these numbers are higher than committer counts.

Quick start

library(osscontribs)

# Cumulative commits over time
data(oss_daily_commits)
projects <- split(oss_daily_commits, oss_daily_commits$project)
plot(NULL, xlim = range(oss_daily_commits$date),
     ylim = c(0, max(oss_daily_commits$cumulative_commits)),
     xlab = "Date", ylab = "Total Commits",
     main = "30 Years of Open Source Commits")
cols <- c(freebsd = "red", openbsd = "orange",
          netbsd = "blue", postgresql = "purple")
for (p in names(projects)) {
  lines(projects[[p]]$date, projects[[p]]$cumulative_commits,
        col = cols[p], lwd = 2)
}
legend("topleft", names(cols), col = cols, lwd = 2)