Hacker News: nryoo

New comment by nryoo in "Agentjacking: Fake error reports hijack Claude Code and Cursor into running code"

nryoo — Mon, 15 Jun 2026 18:35:44 +0000

The root problem is that coding agents treat tool output as trusted context.

Agentjacking: Fake error reports hijack Claude Code and Cursor into running code

nryoo — Mon, 15 Jun 2026 18:35:22 +0000

Article URL: https://thenextweb.com/news/agentjacking-ai-coding-agents-sentry

Comments URL: https://news.ycombinator.com/item?id=48545271

Points: 3

# Comments: 1

New comment by nryoo in "Show HN: Philosophy for Kids"

nryoo — Mon, 15 Jun 2026 02:39:44 +0000

Good idea. Worth considering though, philosophy isn't really about right answers, it's about learning to question. Bit worried about an LLM tends to hand kids a tidy conclusion, which is kind of the opposite of what you want.

Monitoring LLM Inference with Prometheus and Grafana (vLLM, TGI, Llama.cpp)

nryoo — Mon, 15 Jun 2026 02:34:15 +0000

Article URL: https://www.glukhov.org/observability/monitoring-llm-inference-prometheus-grafana/

Comments URL: https://news.ycombinator.com/item?id=48535910

Points: 2

# Comments: 0

New comment by nryoo in "Saving family football footage with a Raspberry Pi and a 1928 projector"

nryoo — Sun, 14 Jun 2026 11:35:30 +0000

The FFmpeg/Python step is neat. They avoid stopping on every frame, capture continuous video instead, then split it back out and pick the sharpest frame from each group.

Saving family football footage with a Raspberry Pi and a 1928 projector

nryoo — Sun, 14 Jun 2026 11:20:26 +0000

Article URL: https://www.raspberrypi.com/news/saving-family-football-footage-with-a-raspberry-pi-and-a-1928-projector/

Comments URL: https://news.ycombinator.com/item?id=48526178

Points: 3

# Comments: 1

New comment by nryoo in "Quantum error correction with the toric code on neutral atoms"

nryoo — Wed, 10 Jun 2026 13:21:15 +0000

From Atom Computing. The toric code needs non-local connections, which plays to neutral atoms' all-to-all connectivity over fixed-topology superconducting qubits.

Quantum error correction with the toric code on neutral atoms

nryoo — Wed, 10 Jun 2026 13:20:22 +0000

Article URL: https://arxiv.org/abs/2606.04079

Comments URL: https://news.ycombinator.com/item?id=48475897

Points: 3

# Comments: 1

New comment by nryoo in "Apple's Foundation Models can now use third-party LLMs (Claude, Gemini) [video]"

nryoo — Wed, 10 Jun 2026 02:52:11 +0000

Text summary since it's a video: it's a model abstraction layer. You write against a LanguageModelSession, import a Swift package, initialize the model you want, and pass it to the session. Apple ships SystemLanguageModel (on-device), PrivateCloudComputeLanguageModel, and open-sourced CoreAILanguageModel and MLXLanguageModel for local models; Anthropic and Google are publishing their own packages.

Apple's Foundation Models can now use third-party LLMs (Claude, Gemini) [video]

nryoo — Wed, 10 Jun 2026 02:51:42 +0000

Article URL: https://developer.apple.com/videos/play/wwdc2026/241/

Comments URL: https://news.ycombinator.com/item?id=48470772

Points: 2

# Comments: 1

New comment by nryoo in "Microsoft says it will not pursue security researchers after zero-day backlash"

nryoo — Tue, 09 Jun 2026 04:18:44 +0000

Microsoft dropped the legal threat, but it also dropped the phrase "responsible disclosure." The new statement says coordinated vulnerability disclosure instead. That's the term Microsoft itself switched to back in 2010, specifically so researchers who go public wouldn't be painted as irresponsible. Katie Moussouris, who helped make that switch, said invoking "responsible" this time was "the first strike in my book."

Microsoft says it will not pursue security researchers after zero-day backlash

nryoo — Tue, 09 Jun 2026 03:08:44 +0000

Article URL: https://therecord.media/microsoft-says-it-will-not-pursue-security-researchers-disclosure

Comments URL: https://news.ycombinator.com/item?id=48455828

Points: 12

# Comments: 1

Greenwald density limit isn't a hard wall: density-free regime seen on EAST

nryoo — Mon, 08 Jun 2026 10:21:29 +0000

Article URL: https://www.science.org/doi/10.1126/sciadv.adz3040

Comments URL: https://news.ycombinator.com/item?id=48443463

Points: 2

# Comments: 0

A Virtual Processor Brings Back the Free Lunch

nryoo — Tue, 02 Jun 2026 02:58:28 +0000

Article URL: https://arxiv.org/abs/2605.30507

Comments URL: https://news.ycombinator.com/item?id=48365452

Points: 2

# Comments: 0

"What a joke": GitHub Copilot's token-based billing spurs backlash among devs

nryoo — Mon, 01 Jun 2026 04:14:30 +0000

Article URL: https://techcrunch.com/2026/05/30/what-a-joke-github-copilots-new-token-based-billing-spurs-consternation-among-devs/

Comments URL: https://news.ycombinator.com/item?id=48352533

Points: 11

# Comments: 1

New comment by nryoo in "Navier-Stokes fluid simulation explained with Godot game engine"

nryoo — Sat, 30 May 2026 14:32:53 +0000

Nice writeup. One thing worth adding to the limitations: without vorticity confinement, the Gauss-Seidel projection step quietly dissipates the small-scale curl that makes smoke look like smoke.

The 2001 Fedkiw/Stam/Jensen "Visual Simulation of Smoke" paper added it back as a correction force for exactly this reason. At N=16 it doesn't matter much because the grid itself can't represent fine vortices, but the moment you crank N up the missing confinement becomes visible.

New comment by nryoo in "A new register allocator for ZJIT"

nryoo — Sat, 30 May 2026 11:31:46 +0000

How big is the lifetime holes thing in practice? On loops the contiguous-interval model spills way more than it should. Wondering if that alone explains most of the YJIT gap.

Cognition (Devin): $1B Series D at $26B valuation

nryoo — Fri, 29 May 2026 09:20:18 +0000

Article URL: https://cognition.ai/blog/series-d

Comments URL: https://news.ycombinator.com/item?id=48320889

Points: 2

# Comments: 0

New comment by nryoo in "Claude Platform on AWS"

nryoo — Tue, 12 May 2026 02:54:13 +0000

So what's the good point?

New comment by nryoo in "I prompted ChatGPT, Claude, Perplexity, and Gemini and watched my Nginx logs"

nryoo — Mon, 20 Apr 2026 16:08:27 +0000

So the state of AI in 2026: ChatGPT DDoS-lite, Claude the polite one that actually reads the rules, Perplexity maybe shows up, and Google was already in your house.