glutimus

Synthetic / benchmark content for experiments with repository size and commit history.

What this repo did (with codemaximus)

Roughly:

1. Commit count — Used codemaximus hyperdrive (git fast-import) to append a long linear chain of commits on main so the repository reaches on the order of 10,000 commits (one small file updated each time: slop/hyperdrive_signal.txt). This is the fast way to stress commit history without running thousands of separate git commit invocations.

2. Lines of code — Generated about 100 million lines of synthetic multi-language text under output/ (Java, Python, JavaScript/TypeScript, Go, and generic Python-shaped files), via codemaximus --lines 100000000 --lang all (and similar). That tree was then added, committed, and pushed so the slop lives in Git history like normal source.

Together, that inflates commit count and tracked LOC for badges, benchmarks, or demos. It is not meaningful application code.

Generated code

Large portions of the tree (notably under output/) were produced with codemaximus—a CLI that synthesizes plausible-looking multi-language “slop” for demos and load tests, not for production use.

See the codemaximus README for installation, options (--lines, --turbo, hyperdrive, etc.), and policy notes.