
New contributors to GNU Emacs over time
I was reading the Wikipedia page for Emacs and found this statement:
>As of 2025, GNU Emacs has had 1,608 individual committers throughout its history.
This made me curious about how first-time contributors to Emacs are distributed over time. I took a local checkout of the Emacs mirror repository and scanned the commit history. For each commit, I recorded the Git author name and email, treated the pair (author_name, author_email) as an identity, and then kept only the first commit where each identity appears. Finally, I plotted the cumulative number of distinct authors over time. I used Git's .mailmap normalization (git log --use-mailmap) to map known name/email variants to the same person when the repository provides such mappings.
This is just a quick/fun experiment, not a rigorous study. There are obvious caveats. The most obvious ones are possible duplicated identities not covered by .mailmap.
Still, the cumulative plot surprised me. It seems to show two very different slopes, with a clear change around 2010: before then, relatively few new author identities appear per year; after then, the number rises much faster.
What is good news for me: Emacs did not decline in popularity. It can still attract new contributors, presumably younger fellow Emacs enthusiasts.
Does anyone know what explains the change around 2010? Was it related to migration away from older version-control workflows, when migrating from CVS to Bazaar and then Git, or simply to a change in how patches were attributed to their original authors?