OmniDiff – Opensource, super fast, content aware Git difftool, in Rust
Comments on “OmniDiff – Opensource, super fast, content aware Git difftool, in Rust”
Claim-checked. Scored by all three judges, and its checkable claims were compared with the sources it cites. How far stories are checked
OmniDiff, an AGPL-licensed Rust diff tool by Marko Ivankovic, picks a diff method by content: tree-sitter structural diffs for 25 languages, plus pictures, fonts, archives and PDFs. It plugs into git, jj, VS Code and Neovim. The README reports a 4.2ms median over 2,302 fixtures and 442,322 of 442,530 code pairs completed. Its own accuracy goals are unmet: 60–70% of hand-painted cases match exactly. All figures are self-reported, from one laptop, and archives and PDFs are slower.
Why it ranked
No. 22 of 43 with a score of 0.56. The panel rated it highest on new information (0.94) and substantiated (0.88), lowest on significance (0.44). The judges disagreed most on overclaim (spread 0.63), so the story is marked contested. Placed No. 22, No. 39 by score alone: the fourth Programming & dev tools story of the day (−0.020). The placement terms move stories; they are not part of the score.
- AI judgeJudge 2claude-haiku-5-5
- A Rust git difftool with broad format support, detailed benchmarks and integration docs. Useful to developers and substantiated by measurements, but a niche developer tool rather than a landmark.
- AI judgeJudge 3gpt-5.6-luna
- A highly detailed, reproducibly benchmarked launch of a capable developer tool, with practical installation and integration guidance. Its audience is broad among developers but significance is moderate because it is a new niche tool.
Each judge wrote its line after answering the rubric, with the site and the author still hidden. Every answer behind this is in the table under Score.
Score
0.56total under rubric v2.3Contestedjudges differ by 0.63 on overclaim
- Significance0.44
- Reader interest0.67
- Audience breadth0.46
- Practical relevance0.75
- Substantiated0.88
- Technical depth0.85
- New information0.94
- Counts against: Promotional0.17
- Counts against: Overclaim0.42
Each judge’s answer per question (0–1), their mean and spread, and what the question adds to the score (weight × mean).
| Question | Judge 1 | Judge 2 | Judge 3 | Mean | Spread | Weight | Contribution (weight times mean) |
|---|---|---|---|---|---|---|---|
| SignificanceHow much does this change what people in tech know or can do? | 0.55 | 0.30 | 0.48 | 0.44 | 0.25 | +0.24 | +0.106 |
| Reader interestWould a well-informed senior engineer or tech founder want to read this today, and likely pass it on? | 0.67 | 0.55 | 0.78 | 0.67 | 0.23 | +0.26 | +0.173 |
| Audience breadthHow many people who build technology does it concern? | 0.55 | 0.40 | 0.42 | 0.46 | 0.15 | +0.13 | +0.060 |
| Practical relevanceHow useful is it to someone who builds software or tech products? | 0.89 | 0.65 | 0.72 | 0.75 | 0.24 | +0.22 | +0.166 |
| SubstantiatedHow well is it backed? Consider whether it is the primary source, and whether its main claims come with data, code, a demo, a reproducible method or citations. | 0.98 | 0.72 | 0.93 | 0.88 | 0.26 | +0.05 | +0.044 |
| Technical depthHow technically substantive is it? | 0.96 | 0.70 | 0.90 | 0.85 | 0.26 | +0.05 | +0.043 |
| New informationDoes it contain information absent from the stories covered in the last 7 days (list provided)? | 0.92 | 0.90 | 0.99 | 0.94 | 0.09 | +0.05 | +0.047 |
| PromotionalIs it mainly marketing with little substance (a press release, ad or sponsored post with no real information)? A launch post that carries real news is not promotional. | 0.09 | 0.35 | 0.08 | 0.17 | 0.27 | −0.19 | −0.033 |
| OverclaimDo the headline or conclusions claim more than the body supports? | 0.75 | 0.40 | 0.12 | 0.42 | 0.63 , above the contested threshold | −0.11 | −0.047 |
| Sum of contributions | +0.559 | ||||||
| Claim-check penalty | 0.000 | ||||||
| Total, as published | 0.559 | ||||||
Spread is the highest minus the lowest judge answer. marks a spread above 0.40: the judges disagree, and the story is marked contested.
The badge for this ranking, for the author's page or anyone's: one line of HTML, and it links back here.
<a href="https://rubric.hrih.si/editions/2026-11-02/22/"><img src="https://rubric.hrih.si/badge/2026-11-02/22.svg" alt="Ranked #22 on Rubric, 2 Nov 2026" width="238" height="28"></a>Claim check
Claim check: 8 no source given. Each claim is compared with the source the article itself cites; a flag is published only when both checkers agree.
No source givenNumber
OmniDiff can diff 99.95% of code changes: 442,322 of 442,530 readable before/after pairs in 25 languages completed successfully.
The article gives no link or citation for it.
The article says “442,322 (99.95%) completed successfully”
The source says No source to compare with.
No source givenNumber
98.8% of the 2,018,010 file changes across 7,352 repositories are diffed as text or by their content.
The article gives no link or citation for it.
The article says “2,018,010 changes in 7,352 repositories, pictures, fonts and archives included - 98.8% are diffed as text or by their content”
The source says No source to compare with.
No source givenBenchmark result
Over the 2,302 fixtures, OmniDiff measured p50 4.2ms, p90 35.1ms, p99 177ms and slowest 482ms on an Intel Core i7-1370P laptop.
The article gives no link or citation for it.
The article says “p50 4.2ms, p90 35.1ms, p99 177ms, slowest 482ms”
The source says No source to compare with.
No source givenGeneralization
OmniDiff's goals of a median diff of 100ms or less and a 99th-percentile diff of 1000ms or less are both met.
The article gives no link or citation for it.
The article says “Both are met”
The source says No source to compare with.
No source givenNumber
100ms is the 97.6th percentile of fixture times: 55 fixtures take longer than 100ms and none takes longer than a second.
The article gives no link or citation for it.
The article says “100ms is the 97.6th percentile — 55 fixtures take longer than that, and none takes longer than a second”
The source says No source to compare with.
No source givenBenchmark result
With --full, 851 (60.0%) of 1,419 hand-painted fixtures have zero mismatched bytes and 1,241 (87.5%) have at most 1% mismatched; with --minimal the figures are 987 (69.6%) and 1,299 (91.5%).
The article gives no link or citation for it.
The article says “full 851 (60.0%) 1,241 (87.5%) 99,157 of 73.3M (0.14%”
The source says No source to compare with.
No source givenNumber
The Intel Xeon E3-1275 v5 server that measured the research figures takes about 2.5 times as long per fixture as the laptop.
The article gives no link or citation for it.
The article says “the Intel Xeon E3-1275 v5 server that measured the research figures takes about 2.5 times as long per fixture”
The source says No source to compare with.
No source givenCausal conclusion
Only 57% of PDF pairs take under 100ms, since a PDF draws every changed page, while archives miss the 99th-percentile goal.
The article gives no link or citation for it.
The article says “only 57% of them take under 100ms, since a PDF draws every changed page”
The source says No source to compare with.
Also covered by
No other coverage was clustered with this story.
Elsewhere
Elsewhere: HN 196 pts, front page
Recorded for context only, checked 2 Nov 2026, 05:46 UTC. It played no part in the ranking.