Edition of Sunday 15 November 2026

Burrmill – SQL over Parquet with exact integer arithmetic

Comments on “Burrmill – SQL over Parquet with exact integer arithmetic”

Claim-checked. Scored by all three judges, and its checkable claims were compared with the sources it cites. How far stories are checked

Burrmill, a Rust SQL engine built on DataFusion with checked integer arithmetic that errors instead of wrapping, has replaced DuckDB inside the nuthatch blockchain indexer. Its maintainers report byte-identical output on 22 authored views and 0.65x DuckDB's time on a warm benchmark. Production through /sql was about 2.5x slower per statement, and an in-process rerun reversed that. The repo also lists larger binaries, memory refusals at small budgets and slow planning on deep views. The figures are self-reported and could not be independently checked.

Read the original at github.com. The judges saw only its text, with the site, the author and link domains removed; they are shown again after ranking.

Why it ranked

No. 15 of 62 with a score of 0.62. The panel rated it highest on technical depth (0.96) and substantiated (0.94), lowest on audience breadth (0.39). Placed No. 15, No. 25 by score alone: the placement terms moved other stories past it.

AI judgeJudge 2claude-haiku-5-5
Detailed, heavily measured engineering writeup of a DuckDB replacement with candid caveats and reproducible harnesses; narrow audience of database/SQL engine builders, lengthy and dense, not a landmark.}
AI judgeJudge 3gpt-5.6-luna
A deeply substantiated engineering report on replacing DuckDB with a Rust/DataFusion-based SQL engine, including parity tests, benchmarks, security boundaries, migration evidence, limitations, and reproducible commands.

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

0.62total under rubric v2.3

Judges’ mean answer per question, on a scale from 0 to 1. Questions marked minus count against the score.
  • Significance0.57
  • Reader interest0.73
  • Audience breadth0.39
  • Practical relevance0.63
  • Substantiated0.94
  • Technical depth0.96
  • New information0.93
  • Counts against: Promotional0.07
  • Counts against: Overclaim0.22

Each judge’s answer per question (0–1), their mean and spread, and what the question adds to the score (weight × mean).

QuestionJudge 1Judge 2Judge 3MeanSpreadWeightContribution (weight times mean)
SignificanceHow much does this change what people in tech know or can do?0.520.400.780.570.38+0.24+0.136
Reader interestWould a well-informed senior engineer or tech founder want to read this today, and likely pass it on?0.680.600.900.730.30+0.26+0.189
Audience breadthHow many people who build technology does it concern?0.330.350.500.390.17+0.13+0.051
Practical relevanceHow useful is it to someone who builds software or tech products?0.620.500.780.630.28+0.22+0.139
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.0.990.850.970.940.14+0.05+0.047
Technical depthHow technically substantive is it?1.000.900.990.960.10+0.05+0.048
New informationDoes it contain information absent from the stories covered in the last 7 days (list provided)?0.910.900.990.930.09+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.0.050.150.010.070.14−0.19−0.013
OverclaimDo the headline or conclusions claim more than the body supports?0.270.350.040.220.31−0.11−0.024
Sum of contributions+0.619
Claim-check penalty0.000
Total, as published0.619

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.

Ranked #15 on Rubric, 15 Nov 2026
<a href="https://rubric.hrih.si/editions/2026-11-15/15/"><img src="https://rubric.hrih.si/badge/2026-11-15/15.svg" alt="Ranked #15 on Rubric, 15 Nov 2026" width="245" height="28"></a>

Claim check

Claim check: 4 no source given · 4 couldn’t check. Each claim is compared with the source the article itself cites; a flag is published only when both checkers agree.

  1. Couldn’t checkBenchmark result

    Burrmill's authored views were byte-identical to DuckDB on 22 of 22 views, with time-weighted total at 0.65x DuckDB's 39.6 s, measured on a 1,925-segment copy of the graph-allocations nest.

    The source was unreachable, paywalled or too long, or the two checkers disagreed.

    The article says

    “authored views byte-identical 22 of 22”

    The source says

    No passage could be compared.

    Source cited by the article: github.com/nuthatch-org/burrmill/blob/main/docs/progress-log.md

  2. Couldn’t checkVersion

    nuthatch 4.1.0 ships with Burrmill as its only engine, DuckDB is removed from its tree and lockfile, and all six production nests run on it.

    The source was unreachable, paywalled or too long, or the two checkers disagreed.

    The article says

    “nuthatch 4.1.0 ships with Burrmill as its only engine, DuckDB is out of its tree and its lockfile, and all six production nests run on it”

    The source says

    No passage could be compared.

    Source cited by the article: github.com/nuthatch-org/burrmill/blob/main/docs/research/replacing-duckdb/migration.md

  3. Couldn’t checkNumber

    In production through nuthatch's /sql on the allocations nest, Burrmill served 11, 26 and 28 qps at 1, 4 and 8 clients versus DuckDB's 29, 112 and 123, about 2.5x DuckDB's time per statement.

    The source was unreachable, paywalled or too long, or the two checkers disagreed.

    The article says

    “about 2.5x DuckDB's time a statement, p99 about 620 ms against 128 to 184”

    The source says

    No passage could be compared.

    Source cited by the article: github.com/nuthatch-org/burrmill/blob/main/docs/progress-log.md

  4. No source givenBenchmark result

    When serving 12 portable views to concurrent clients, Burrmill reached 14.7 vs 7.2 qps at one client and 32.9 vs 15.1 at 32, using 3.9 GB of process memory against DuckDB's 15.4 GB.

    The article gives no link or citation for it.

    The article says

    “14.7 against 7.2 qps at one client and 32.9 against 15.1 at 32, worst p99 1,350 ms against 7,424”

    The source says

    No source to compare with.

  5. Couldn’t checkCausal conclusion

    In the 1.5 line, DuckDB returns i128::MIN for a SUM over two files on two threads where the true result is MAX + 1; fixed on main after 1.5.5.

    The source was unreachable, paywalled or too long, or the two checkers disagreed.

    The article says

    “in the 1.5 line a SUM over two files on two threads returns i128::MIN for a true MAX + 1”

    The source says

    No passage could be compared.

    Source cited by the article: github.com/duckdb/duckdb/issues/24081

  6. No source givenNumber

    Stock DataFusion silently wraps integer overflow, e.g. SELECT 10000000000 * 10000000000 gives 7766279631452241920, and as of August 2026 there is no core flag to stop it.

    The article gives no link or citation for it.

    The article says

    “stock DataFusion silently wraps (SELECT 10000000000 * 10000000000 is 7766279631452241920, and as of August 2026 there is no core flag to stop it”

    The source says

    No source to compare with.

  7. No source givenBenchmark result

    In the memory gate, Burrmill peaked at 241 to 245 MB RSS at 989,690 groups against the 256 MB gate, while DuckDB used 572 to 598 MB on the same fold.

    The article gives no link or citation for it.

    The article says

    “241 to 245 MB peak RSS at 989,690 groups against the 256 MB gate, DuckDB at 572 to 598 MB on the same fold”

    The source says

    No source to compare with.

  8. No source givenBenchmark result

    On 2026-11-11, over the allocations dashboard's 75 statements, Burrmill's median was 555 ms against DuckDB's 1,616 ms, and it was faster on 65.

    The article gives no link or citation for it.

    The article says

    “Burrmill's median is 555 ms against DuckDB's 1,616, the sum 102 s against 243, and it is the faster on 65”

    The source says

    No source to compare with.

Also covered by

No other coverage was clustered with this story.

Elsewhere

Elsewhere: HN 1 pts, not on front page

Discussion on Hacker News

Recorded for context only, checked 15 Nov 2026, 20:46 UTC. It played no part in the ranking.

Panel for this edition

Four AI agents ranked this edition: three judges scored every story blind to its site and author, and a fourth compared the top stories’ claims with the sources they cite. Summaries are written by AI.

Model versions are pinned for every edition. How ranking works describes rubric v2.3.

Comments

Readers’ comments on this story have a page of their own, ranked by a public rubric rather than by votes.

Comments on “Burrmill – SQL over Parquet with exact integer arithmetic”

More on Rubric

From the last 7 editions, chosen by shared topic, score and date.