Hacker News on Saturday, 2 May 2026

Stories posted this day, by points. Rank history starts with the October 2026 rebuild.

Screenshot of github.com: VS Code inserting 'Co-Authored-by Copilot' into commits regardless of usage

VS Code inserting 'Co-Authored-by Copilot' into commits regardless of usage (github.com)

1420 points by indrora 5 months ago | 768 comments

A merged GitHub pull request in microsoft/vscode titled 'Enabling ai co author by default', which adds a Copilot co-author trailer to commits by default.

vscode github-copilot git microsoft

Screenshot of noctua.at: Why does it take so long to release black fan versions?

Why does it take so long to release black fan versions?(noctua.at)

701 points by buildbot 5 months ago | 282 comments

Noctua blog post explaining why releasing black versions of its PC fans takes so long; the capture shows only a Vercel security checkpoint.

noctua pc-cooling computer-fans product-development

Dav2d (code.videolan.org)

579 points by dabinat 5 months ago | 166 comments

VideoLAN's GitLab repository for dav2d, an open-source AV2 video decoder aimed at being small, portable and very fast.

av2 video-codec videolan dav2d

Screenshot of donottrack.sh: Do_not_track

Do_not_track (donottrack.sh)

476 points by RubyGuy 5 months ago | 146 comments

DO_NOT_TRACK is a proposed standard environment variable for opting out of telemetry in CLI tools, SDKs, and frameworks, with a table of existing per-tool opt-out methods.

telemetry privacy environment-variables cli-tools

Screenshot of ladybird.org: This Month in Ladybird – April 2026

This Month in Ladybird – April 2026 (ladybird.org)

474 points by richardboegli 5 months ago | 133 comments

Ladybird browser project's April 2026 monthly newsletter covering 333 merged PRs from 35 contributors, new sponsors, and development progress.

ladybird web-browser browser-engine newsletter

Screenshot of theguardian.com: OpenAI’s o1 correctly diagnosed 67% of ER patients vs. 50-55% by triage doctors

OpenAI’s o1 correctly diagnosed 67% of ER patients vs. 50-55% by triage doctors (theguardian.com)

473 points by donsupreme 5 months ago | 419 comments

Guardian news article on a Harvard trial finding OpenAI's o1 model outperformed doctors at diagnosing emergency room patients during triage.

openai-o1 medical-diagnosis emergency-medicine triage

Screenshot of nethack.org: NetHack 5.0.0

NetHack 5.0.0 (nethack.org)

453 points by rsaarelm 5 months ago | 139 comments

Official NetHack DevTeam announcement of the NetHack 5.0.0 release, describing architectural changes, over 3100 fixes, and save-file incompatibility.

nethack roguelike game-release c99

Screenshot of david-smith.org: Six years perfecting maps on watchOS

Six years perfecting maps on watchOS (david-smith.org)

414 points by valzevul 5 months ago | 106 comments

A Design Diary blog post by indie developer David Smith recounting six years of work refining map features in his watchOS apps.

watchos apple-watch maps indie-apps

Screenshot of ask.com: Ask.com has closed

Ask.com has closed (ask.com)

402 points by supermdguy 5 months ago | 205 comments

Ask.com's homepage now shows a farewell notice saying IAC discontinued its search business and Ask.com closed on May 1, 2026 after 25 years.

ask-com iac search-engines shutdown

Screenshot of blog.haskell.org: A couple million lines of Haskell: Production engineering at Mercury

A couple million lines of Haskell: Production engineering at Mercury (blog.haskell.org)

399 points by unignorant 5 months ago | 196 comments

Haskell Blog post by Ian Duncan, first in the 'Haskellers from the trenches' series, on production engineering with a multi-million-line Haskell codebase at Mercury.

haskell mercury production-engineering functional-programming

Screenshot of arxiv.org: AI Self-preferencing in Algorithmic Hiring: Empirical Evidence and Insights