Etna: An Evaluation Platform for Property-Based Testing
summary
Etna is a multilanguage evaluation platform that runs property‑based testing experiments across Coq and Haskell, uses handcrafted mutants and mutation testing to measure bug‑finding, and ships workloads (data structures, languages, IFC) plus baselines. Experiments show concrete numbers: e.g. a workload with 36 tasks where generators find 14 bugs under 0.1s, six under 1s, four timed out at 60s; QuickCheck type‑based generators drop on red‑black trees versus bespoke and LeanCheck outperforms SmallCheck; fixes to FastCheck yield dramatic improvements.
AI-written summary · about
chapters (12)
- 0:00Property-based testing intro
- 1:27Evaluating testing challenges
- 3:54Introducing Etna platform
- 5:41Etna evaluation strategies
- 8:13Workloads and mutation testing
- 11:23Etna use cases
- 13:05Haskell framework comparison
- 15:13QuickCheck generation strategies
- 18:12Size effects in bug finding
- 20:25Improving FastCheck with Etna
- 22:58Etna conclusion
- 23:41Q&A








