Transcript Testing proposes: write test code for humans to read
Understand this story
A developer has proposed a testing method called "Transcript Testing," arguing that software tests should be explicitly written for humans to read, to help people understand and verify AI-generated code. The problem it targets: test code generated by AI coding tools often runs to hundreds of lines, and is easily skipped during manual review. The proposer cites an Apache DataFusion PR as an example, saying it contains four test functions and about 400 lines of code, covering row group contents, write settings, and predicate combinations. Currently this is only a methodological proposal; the report does not give verification results after adopting this style.
AI overview based on source reports · Updated 21 hr ago
Coverage timeline
Follow the coverage to explore different perspectives.
- Hacker News AITranscript Testing: Helping Humans Verify AI Code
The post argues that software tests in the AI era should be written explicitly for humans to read, since AI coding tools now produce large volumes of hard-to-understand tests that reviewers may skim or skip. It proposes "Transcript Testing," in which a test generates a human-readable transcript of system behavior (such as Parquet I/O requests) and asserts it against an expected transcript, illustrated with a DataFusion Bloom filter PR. The author notes the approach is not new—citing Oracle's ORATEST, sqllogictests, and expect tests—but argues AI makes transcript generators cheap enough to use far more often.
Activity for this story
- Current comparable activity
- 6
- Peak comparable activity
- 10Nov 14 01:00
- Change in the past 24 hours
- –
Trends compare the same fully observed participants, which may cover fewer participants than current activity. Move the pointer or tap the chart to inspect hourly activity; use the left and right arrow keys with a keyboard.