Serialization course lab notebooks
Hands-on companions to the Serialization 101–401 theory track.
Honesty: timings and sizes here are illustrative. Suite language Results own benchmark-runner truth.
How to run
Google Colab
Open any notebook via its Open in Colab badge (links target the master branch on GitHub after merge).
Local
# from repo root
python -m venv .venv-notebooks && source .venv-notebooks/bin/activate
pip install -r docs/theory/notebooks/requirements-colab.txt jupyter
jupyter lab docs/theory/notebooks/Layout
| Phase | Path | Content |
|---|---|---|
| P0 | 401/ |
Wire-format playground + MiniUser lab |
| P1 | 201/ |
Mechanism labs (encode cost, schema, evolution, …) |
| P2 | 101/ |
Data science + engineering mini labs |
| P3 | 301/ |
Production experiment notebooks |
| P4 | companions/ |
JS Node snippets + Go / Rust MiniUser goldens |
Design rules
- Python-first on Colab; multi-language only as thin companions (P4).
- Prefer stdlib + a few pinned deps; avoid pulling the full suite benchmark runner.
- Reuse teaching schemas (MiniUser, small sensor/order records)—not full suite fixtures.
- Link back to the article; do not fork a second curriculum.
- Notebook microbenchmarks never rank libraries globally.
Colab URL pattern
https://colab.research.google.com/github/leo-gan/GLD.SerializerBenchmark/blob/master/docs/theory/notebooks/<level>/<name>.ipynb