Week 2: Dating techniques

How to read a commit history as geological layers — construction, maintenance, decay — and what the boundaries between them reveal.

1 March 2027

Open the slides

Learning outcomes#

By the end of this week you should be able to:

1. construct a phase model of a project’s history from version-control data
2. apply the logic of archaeological stratigraphy, and its concept of a sequence, to a non-physical record
3. state the systematic distortions that make commit timestamps unreliable evidence

Outline#

• git log as a primary source: what it records and what it does not
• Commit frequency as a proxy for project health
• Author diversity over time: the bus factor, visible in the log
• Case study: OpenSSL before Heartbleed — years of thin maintenance on one of the most critical libraries in the internet’s stack
• Boundary events: releases, rewrites, the refactor that broke everything
• The limits of dating: clock skew, rebased histories, squash merges

Required reading#

• Harris, E.C. (1989) Principles of Archaeological Stratigraphy, 2nd edn. Academic Press. Chapters 3–5 on the laws of stratigraphy and the Harris matrix.
• Tornhill, A. (2015) Your Code as a Crime Scene. Pragmatic Bookshelf. Part I.

Further reading#

• Schiffer, M.B. (1987) Formation Processes of the Archaeological Record. University of New Mexico Press.
• Lehman, M.M. and Ramil, J.F. (2001) ‘Rules and Tools for Software Evolution Planning and Management’, Annals of Software Engineering 11, pp. 15–44.

Seminar questions#

1. Harris insists that stratigraphic sequence is independent of the material in each layer. What is the software equivalent of that claim, and does it survive git rebase?
2. Schiffer distinguishes cultural from natural formation processes. Which processes deposit commits that no human intended as a record?
3. Your timeline shows a six-month silence. Enumerate four distinct histories that would produce the same silence.