Week 1 Dig: The dead repo
Survey — what makes software dead, how to tell, and how to choose a dig site you will spend the semester excavating.
22 February 2027
Before the dig#
Find three candidate dig sites: public repositories on GitHub, GitLab, or SourceForge whose last meaningful commit is at least eighteen months old. “Meaningful” excludes bot-authored dependency bumps and CI config changes — a project with a Dependabot commit last Tuesday and a human commit in 2021 is dead.
For each candidate, note: the language, the approximate size (files, commits, contributors), the licence, and the date range of human activity. Bring all three to the session.
In the session#
Present your three candidates. The class helps you choose: the best dig site is big enough to reward twelve weeks of investigation, dead enough to be unambiguous, and licensed permissively enough that you can fork it for the restoration in week 11.
You will work with this codebase for the rest of the semester. Choose one you can live with.
Afterwards#
Clone your chosen site. Run git log --oneline | wc -l and
git log --format='%ai' | head -1 / tail -1 to confirm the date range. Write
a one-paragraph summary of what the project appears to be, based only on the
README and the file listing — do not read the code yet.
The spec
What you bring to the session. Written so a reader can tell whether it has been met without asking you.
- you have chosen a dig site — a public, abandoned repository with a visible history
- you can clone it, read its log, and name its last active date
- you can articulate why you chose it over the alternatives you considered
Teaching team
- Maren Voss — convenor