frontpage.
newsnewestaskshowjobs

Open Source @Github

fp.

We don't need more apps. We need better open source

https://jordankrueger.com/blog/better-open-source-not-more-apps/
172•pandemicsoul•1h ago•82 comments

Loopholes Are for Assholes

https://actsofvolition.com/2026/11/loopholes-are-for-assholes/
88•speckx•1h ago•44 comments

Show HN: Alloy CSS, dialogs, menus and tabs that work with JavaScript off

https://alloyhtml.com/example.html
31•prafandy•55m ago•5 comments

New Evidence Shows School Phone Bans Are Gradually Improving Test Scores

https://www.nytimes.com/2026/11/17/upshot/school-phone-bans-study-scores.html
616•donohoe•4h ago•313 comments

First Photos of the Poles of a Rod-Shaped Bacterium

https://nautil.us/check-out-the-first-photos-of-the-poles-of-a-rod-shaped-bacterium-1286664
43•dustfinger•1d ago•9 comments

Carnegie Mellon is putting computer science courses on YouTube for free

https://www.cs.cmu.edu/learn/
1998•pcompeau•19h ago•268 comments

In 27,280 coding agent runs, 64% of inference spend went to runs that failed

https://runta.com/blog/introducing-frontierharness-eval-v1-1/
18•Edward40•1h ago•2 comments

VisualBasic.World: A gallery of Planet Source Code programs from 1997 to 2012

https://visualbasic.world/
155•broast•1d ago•57 comments

Rows vs. Columns: How different are they, really?

https://spiraldb.com/blog/rows-vs-columns
88•eatonphil•4d ago•34 comments

Baidu, Xiaomi CEOs warn of growing gap between AI hype and returns

https://www.scmp.com/tech/big-tech/article/3374774/baidu-xiaomi-ceos-warn-growing-gap-between-ai-...
116•larrysalibra•1h ago•115 comments

Pentagon fails 9th financial audit in a row

https://thehill.com/policy/defense/6224597-pentagon-fails-ninth-audit/
53•zootedjordan•59m ago•27 comments

Optimizely compromised, served malware on Zoom, GitLab

https://blog.report-uri.com/worlds-largest-brands-infected-with-malware-via-optimizely/
525•kristjansson•17h ago•209 comments

Turbo Pascal Compiler Written in Turbo Pascal

https://turbo51.com/compiler-design/tpc16-turbo-pascal-compiler-written-in-turbo-pascal
61•andsoitis•1d ago•25 comments

BugMeNot has closed

https://www.bugmenot.com/
662•lode•18h ago•297 comments

A.I. Just Blew Up a Key Tenet of Computer Science

https://www.nytimes.com/2026/11/06/science/ai-exponential-time-hypothesis.html
17•doener•28m ago•7 comments

Show HN: VennDB: append-only in-memory Rust DB for rows queried using bit flags

https://github.com/plabayo/venndb
20•gdcbe•1d ago•1 comments

Too Many Stadiums with Roofs

https://sharptext.net/2026/too-many-stadiums-with-roofs/
29•colinprince•3d ago•22 comments

Nimony: A New Compiler for Nim

https://nimony.org/
135•Fervicus•1d ago•66 comments

Highly classified US national security communications network goes down

https://www.cnn.com/2026/11/17/politics/highly-classified-national-security-communications-networ...
21•monkeydust•24m ago•3 comments

If an App Asks for Full Disk Access or System and App Protection Override

https://support.apple.com/en-us/149699
192•jshier•13h ago•103 comments

How the Claude Code team thinks about verification

https://claude.dev/blog/how-the-claude-code-team-thinks-about-verification/
124•craigmart•5h ago•128 comments

Why 3dfx 1 cards had such a washed out color output?

https://twitter.com/id_aa_carmack/status/2122478083319992513
235•mfiguiere•15h ago•82 comments

Chemistry is Actually Geometry

https://chillphysicsenjoyer.substack.com/p/chemistry-is-actually-geometry
293•sebg•19h ago•125 comments

Comcast's "fiber-powered Wi-Fi" is cable, AT&T says in false advertising case

https://arstechnica.com/tech-policy/2026/11/comcast-falsely-advertises-cable-internet-as-fiber-po...
209•jaredwiener•19h ago•180 comments

Frame – 31.5" color ePaper display

https://trmnl.com/blog/introducing-frame
492•geerlingguy•23h ago•353 comments

Microsoft blames monster email outage on software defect introduced in July fix

https://www.theregister.com/saas/2026/11/16/microsoft-blames-monster-email-outage-on-software-def...
407•Bender•1d ago•302 comments

Jiff 1.0: A datetime library for Rust

https://burntsushi.net/jiff/
419•burntsushi•23h ago•91 comments

S. Korea to send nationwide emergency alerts over college entrance exam cheating

https://apnews.com/article/south-korea-exam-cheating-ai-college-e25b5612dbf4bd09fdf9197254c4547e
49•parksb•2h ago•51 comments

What is w2rt6byl5nf1opmc-a.com?

https://jack.cab/blog/siterate
390•dverlaeckt80•1d ago•155 comments

Grazing sheep in fields with solar arrays

https://news.cornell.edu/stories/2026/10/sheep-grazing-solar-fields-offers-farmers-economic-ecolo...
310•PaulHoule•22h ago•313 comments
Open in hackernews

In 27,280 coding agent runs, 64% of inference spend went to runs that failed

https://runta.com/blog/introducing-frontierharness-eval-v1-1/
13•Edward40•1h ago

Comments

guanlan•1h ago
Hi HN, I’m Guanlan, founder of Runta. The eval runs on our runtime, so weigh this accordingly.

The short version: when you change the model, think about changing the harness too. On any one model, cost per pass varies by up to 4.4x

What we ran: 22 harness configurations × 5 open-weight models × 62 terminal tasks × 4 attempts. 27,280 runs, 83B tokens. Each harness ran with default settings at maximum reasoning effort, each trial in a fresh sandbox. Costs are Baseten list prices times reported tokens.

What we found: The lowest-cost harness changes with the model. Codex CLI ranks 1st on cost with GLM-5.3 and 15th with Kimi-K3. Paying more doesn’t buy passes. The harness with the highest pass rate ranked anywhere from 4th to 19th on cost. The model still sets most of the bill (93% of cost variation). The harness is the second lever. Starting prompt size doesn’t predict cost. Output tokens do. 64% of inference spend went to runs that didn’t pass.

radlad•12m ago
Hi Guanlan, I vouched for your comment since you involved in this post, but you should be aware that all of your posts have been flagged, likely for AI patterns in your writing.

> Don't post generated text or AI-edited text. HN is for conversation between humans.

https://news.ycombinator.com/newsguidelines.html