Edition of Thursday 12 November 2026

How OpenAI, Canva and Spotify made slow CI steps up to 20x faster

Comments on “How OpenAI, Canva and Spotify made slow CI steps up to 20x faster”

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

BerryBox published a playbook of about 50 CI optimizations for GitHub Actions, grouped into wasted runs, slow setup, builds, tests, test selection, critical path, runners, flaky tests and pull request volume. Each entry names a risk level and companies using it. Examples include cancelling superseded runs, setting job timeouts from p95 runtimes, caching dependencies by lockfile, sharding tests, and gating new tests with 20 to 50 repeat runs. The listing gives no measured speedups, so the "20x" headline isn't backed by figures here.

Read the original at berrybox.io. 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. 13 of 106 with a score of 0.66. The panel rated it highest on practical relevance (0.94) and new information (0.87), lowest on substantiated (0.37). Overclaim counted against it (0.55). Placed No. 13, No. 19 by score alone: the placement terms moved other stories past it.

AI judgeJudge 2claude-haiku-5-5
A practical CI optimization guide with many actionable items, but the title's specific 20x claims are unverified in the excerpt and it reads as a compiled checklist rather than primary evidence.
AI judgeJudge 3gpt-5.6-luna
A highly actionable CI optimization guide with substantial implementation detail and broad relevance. Its main weakness is limited evidence for the headline’s claimed 20x gains; examples are cited but measurements and methodology are largely absent.

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

0.66total under rubric v2.3

Judges’ mean answer per question, on a scale from 0 to 1. Questions marked minus count against the score.
  • Significance0.56
  • Reader interest0.79
  • Audience breadth0.73
  • Practical relevance0.94
  • Substantiated0.37
  • Technical depth0.75
  • New information0.87
  • Counts against: Promotional0.08
  • Counts against: Overclaim0.55

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?0.720.350.620.560.37+0.24+0.135
Reader interestWould a well-informed senior engineer or tech founder want to read this today, and likely pass it on?0.840.700.820.790.14+0.26+0.205
Audience breadthHow many people who build technology does it concern?0.800.600.780.730.20+0.13+0.095
Practical relevanceHow useful is it to someone who builds software or tech products?0.990.900.920.940.09+0.22+0.206
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.0.300.300.520.370.22+0.05+0.019
Technical depthHow technically substantive is it?0.880.700.670.750.21+0.05+0.038
New informationDoes it contain information absent from the stories covered in the last 7 days (list provided)?0.860.900.850.870.05+0.05+0.044
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.0.050.100.080.080.05−0.19−0.015
OverclaimDo the headline or conclusions claim more than the body supports?0.580.400.680.550.28−0.11−0.061
Sum of contributions+0.665
Claim-check penalty0.000
Total, as published0.665

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 #13 on Rubric, 12 Nov 2026
<a href="https://rubric.hrih.si/editions/2026-11-12/13/"><img src="https://rubric.hrih.si/badge/2026-11-12/13.svg" alt="Ranked #13 on Rubric, 12 Nov 2026" width="245" height="28"></a>

Claim check

Claim check: 8 matches source. Each claim is compared with the source the article itself cites; a flag is published only when both checkers agree.

  1. Matches sourceNumber

    A GitHub Actions job without timeout-minutes can hang for six hours before GitHub stops it.

    The cited source says the same thing.

    The article says

    “A job without timeout-minutes can hang for six hours before GitHub stops it”

    The source says

    “A job without timeout-minutes can hang for six hours before GitHub stops it”

    Source cited by the article: berrybox.io/playbooks/wasted-runs/job-timeouts

  2. Matches sourceCausal conclusion

    A workflow that triggers on both push and pull_request runs each pull request commit twice.

    The cited source says the same thing.

    The article says

    “A workflow that triggers on every push and on pull_request runs each pull request commit twice”

    The source says

    “A workflow that triggers on every push and on pull_request runs each pull request commit twice”

    Source cited by the article: berrybox.io/playbooks/wasted-runs/one-run-per-commit

  3. Matches sourceVersion

    GitHub offers a 1-vCPU runner called ubuntu-slim.

    The cited source says the same thing.

    The article says

    “such as GitHub's 1-vCPU ubuntu-slim”

    The source says

    “On GitHub-hosted runners the smallest option is ubuntu-slim: 1 vCPU, 5 GB of RAM and 14 GB of disk, x64 only, running as a container instead of a full VM”

    Source cited by the article: berrybox.io/playbooks/runners/right-size-down

  4. Matches sourceVersion

    TypeScript 7 ships a native compiler that can be used to type-check faster.

    The cited source says the same thing.

    The article says

    “Type-check with TypeScript 7's native compiler and lint with Oxlint or Ruff”

    The source says

    “TypeScript 7 is a native port of the compiler. It type-checks on several threads (four checkers by default), and Microsoft reports 8x to 12x faster full builds than TypeScript 6”

    Source cited by the article: berrybox.io/playbooks/slow-builds/native-tools

  5. Matches sourceNumber

    Newly added or edited tests should be run 20 to 50 times with retries off before they can block merges.

    The cited source says the same thing.

    The article says

    “run just those tests 20 to 50 times with retries off”

    The source says

    “When a pull request adds or edits tests, run just those tests 20 to 50 times with retries off, report the result as its own check”

    Source cited by the article: berrybox.io/playbooks/flaky-tests/gate-new-tests

  6. Matches sourceCausal conclusion

    A merge queue tests each ready pull request on top of the base branch and the pull requests ahead of it, so authors stop updating branches and rerunning CI before every merge.

    The cited source says the same thing.

    The article says

    “A merge queue tests each ready pull request on top of the base branch and the pull requests ahead of it”

    The source says

    “A merge queue tests each ready pull request on top of the base branch and the pull requests ahead of it, so authors stop updating branches and rerunning CI before…”

    Source cited by the article: berrybox.io/playbooks/pr-volume/merge-queue

  7. Matches sourceGeneralization

    Shopify and Spotify, among others, use the practice of running CI once per commit instead of on both push and pull_request.

    The cited source says the same thing.

    The article says

    “Saves runner minuteslow riskShopify, Spotify +1”

    The source says

    “Shopify vscode-shopify-ruby: limit push runs to main so branch commits run CI once”

    Source cited by the article: berrybox.io/playbooks/wasted-runs/one-run-per-commit

  8. Matches sourceCausal conclusion

    Full link-time optimization should be saved for release builds, where the runtime gain pays for the minutes.

    The cited source says the same thing.

    The article says

    “save full link-time optimization for release builds, where the runtime gain pays for the minutes”

    The source says

    “Link Rust and C/C++ builds with LLD or mold instead of the platform's default linker, and save full link-time optimization for release builds, where the runtime gain pays for the…”

    Source cited by the article: berrybox.io/playbooks/slow-builds/faster-linker

Also covered by

No other coverage was clustered with this story.

Elsewhere

Elsewhere: HN 2 pts, not on front page

Discussion on Hacker News

Recorded for context only, checked 12 Nov 2026, 05:47 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 “How OpenAI, Canva and Spotify made slow CI steps up to 20x faster”

More on Rubric

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