(Programs), Proofs and Refutations (and Tests and Mutants)

on YouTube ↗43m

" summary

An essay-format classroom dialogue probes whether a program is correct by debating binary search specifications and then actually breaking things. It mutates binary_search.c into 85 variants (about 47 fail to compile), runs DeepState random testing (≈327,000 trials in a 30s run) and AFL++ fuzzing, and kills most mutants though some induce infinite loops. The practical claim is that mutation testing (and tools like DeepState and AFL++) exposes specification gaps far better than naive code coverage.

AI-written summary · about

" chapters (7)

1. 0:00Introduction
2. 2:08Why this paper exists
3. 7:33Proving programs correct
4. 11:58Binary search dialogue
5. 22:38Mutation testing demo
6. 36:44Mutation testing answers
7. 39:58Q&A

" related talks

• 01h 06mKill All Mutants! (Intro to Mutation Testing)Dave Aronson · ACCU 2022
• 50mKill All Mutants! Intro to Mutation TestingDave Aronson · NDC Melbourne 2022
• 22mKill All Mutants! (Introduction to Mutation Testing)Dave Aronson · JSConf Hawaiʻi 2020
• 52mKill All Mutants! (Intro to Mutation Testing)Dave Aronson · NDC Sydney 2020
• 01h 00mThere Are Mutants Living in Your Unit TestsMichaël Hompus · NDC Oslo 2025
• 35mKill All Mutants! (Intro to Mutation Testing)Dave Aronson · Heapcon 2024
• 47mKill All Mutants! (Intro to Mutation Testing)Dave Aronson · Voxxed Days CERN 2025
• 44mTesting The Tests: Mutation Testing for C++Seph De Busser · NDC TechTown 2019