digest · hn front page
Fri, Sep 18, 2026
50 stories
-
Android 17 is the first since 3.x to add new APIs without releasing to the AOSP grapheneos.social
The Mastodon web application requires JavaScript to be enabled in the browser to function correctly.
-
Microsoft exec called AI scraping 'the largest theft of labor in human history' techcrunch.com
Unredacted court filings from a copyright lawsuit reveal that executives at Microsoft and OpenAI privately admitted that scraping news content for AI training was equivalent to theft.
-
I don't like passkeys hawksley.dev
Passkeys offer superior protection against phishing and data breaches, but they introduce unacceptable risks for individual users.
-
Cloudflare Quick Tunnels try.cloudflare.com
Cloudflare Quick Tunnels allow developers to expose local servers to the internet via a single command without needing an account, DNS configuration, or opening inbound ports.
-
Claude Code now reads AGENTS.md if there is no Claude.md code.claude.com
The Claude Code updates from September 2026 provide a comprehensive set of bug fixes, performance improvements, and new features across the CLI, VS Code extension, and web interface.
-
OpenJev openjev.com
Comparing direct logit readout versus token generation allows users to measure the performance and accuracy of local AI models in a browser.
-
US Military had close call after using AI for hallucinated intelligence report cnn.com
The US military's rapid integration of artificial intelligence into intelligence reporting has created a dangerous potential for catastrophic miscalculations.
-
A heap overflow and SSO misconfiguration to compromise OpenAI internal repos hacktron.ai
Researchers from Hacktron successfully compromised OpenAI's internal repositories by chaining a heap buffer overflow in the libheif image library with an SSO misconfiguration.
-
Saving another 100TB of RAM blog.cloudflare.com
Cloudflare reduced the memory footprint of its Pingora Backend Router by optimizing the storage and quantity of hashes used in consistent hashing.
-
Inside ZCode: Silently uploading your Git history to the cloud blog.ferstar.org
The ZCode AI coding app secretly packages and uploads a user's entire workspace, including full Git history and sensitive configurations, to a remote server without explicit consent.
-
Jemalloc 5.4.0 github.com
The release of jemalloc 5.4.0 focuses primarily on reducing technical debt through extensive refactoring, bug fixes, and improved test coverage.
-
Minimal Phone 2 minimalcompany.com
The Minimal Phone 2 is a compact, focus-first smartphone designed to reduce screen time and digital distraction by combining a physical QWERTY keyboard with a specialized operating system.
-
Bend 2 and the Vibe-Coding Trap blog.liampwll.com
Vibe-coding allows developers to build complex systems using AI before they understand the underlying problem well enough to realize a more efficient solution already exists.
-
The scourge of x86 emulation fex-emu.com
Emulating the x86 Total Store Ordering (TSO) memory model on ARM's weak consistency model is difficult because x86 requires strict memory ordering and atomicity that ARM does not naturally provide.
-
I vibed a proof of Conway's conjecture overreacted.io
A non-mathematician successfully obtained a Lean proof of Conway's refinement conjecture by using a multi-agent AI workflow.
-
ZCode, the GLM coding agent, silently uploads your Git history tokenstead.ai
The ZCode AI coding application by Z.ai deliberately exfiltrates entire user workspaces, including complete Git histories and sensitive configuration files, without user consent.
-
How SpaceX streamlined the Raptor engine construction-physics.com
The Raptor engine evolved from a complex tangle of external pipes and wires to a streamlined design by internalizing components and removing unnecessary hardware.
-
Show HN: Cactus Needle 3: 8-29MB automation models can match DeepSeek V4 Flash cactuscompute.com
Needle 3 is a highly efficient foundation model designed specifically for mobile devices, wearables, and IoT hardware, prioritizing structured tool calling and data extraction over general conversation.
-
An empirical study of harness design for coding agents arxiv.org
The performance of autonomous coding agents depends heavily on the design of their execution harnesses rather than just the underlying model.
-
Warez: The Infrastructure and Aesthetics of Piracy (2021) archive.org
The Warez Scene is a highly organized, elite underground network that operates differently from public piracy platforms like BitTorrent.
-
Photon-Emission-Guided Laser Fault Injection Enables RP2350 Secure Debug donjon.ledger.com
Researchers successfully bypassed permanent debug-disable settings on the Raspberry Pi RP2350 microcontroller by using photon-emission microscopy to locate a critical control register and laser fault injection to force it into an enabled state.
-
More than 100k people in Japan are now aged 100 or older bbc.com
Japan has reached a historic milestone with over 100,000 people aged 100 or older, driven by advancements in medicine and a healthy traditional diet.
-
Pre-Greek: The lost language hidden within Ancient Greek linguisticdiscovery.com
Ancient Greek contains approximately 1,000 words that cannot be traced to Proto-Indo-European origins, suggesting they were borrowed from an extinct substrate language spoken by Neolithic farmers before the Greeks arrived.
-
The most important product decision is what you don't build liamnugent.me
Successful product management requires a disciplined focus on what not to build to avoid the trap of additive growth and the long-term burden of maintenance.
-
How do we prevent mathemathics from devolving into the Medieval Era of secrecy? mathoverflow.net
The rise of AI-powered proof generation threatens to reverse centuries of open scientific collaboration by incentivizing mathematicians to hide their research to avoid being scooped by well-funded corporations.
-
Xcode 27.1 Beta Release Notes developer.apple.com
The webpage cannot be displayed because the browser lacks active JavaScript.
-
Replacing Pull Requests with Delta zed.dev
The traditional pull request workflow is outdated because AI agents now generate vast amounts of code that make manual diff reviews overwhelming and devoid of context.
-
US Treasuries Have Become Unappetizing for Foreign Central Banks and Governments wolfstreet.com
Foreign central banks and governments are significantly reducing their holdings of US Treasury securities because they have become less attractive due to inflation and shifting economic priorities.
-
Two parallel neural ectoderm progenitors contribute to the developing brain newscientist.com
The human brain evolved from two distinct, primitive nervous systems that merged hundreds of millions of years ago.
-
Second Circuit allows government to search electronic devices at the border knightcolumbia.org
The U.S. Court of Appeals for the Second Circuit has ruled that border agents can search electronic devices without a warrant or suspicion. This decision expands government surveillance powers at the border, threatening the privacy of travelers and the fundamental rights of…
-
Waymo in Singapore waymo.com
Waymo is introducing its all-electric autonomous ride-hailing fleet to Singapore to support the city's goals for sustainable, smart mobility.
-
How Uber Protects Against Retry Storms uber.com
Uber implemented a context-aware 'error ownership' mechanism to prevent retry storms, where multiple services in a deep dependency chain each retry a failed request, leading to exponential traffic growth.
-
When the fractional part of a float fixes your shader crocidb.com
A visual stutter in a Voronoi-diagram shader was caused by a compiler optimization bug in DirectX 11's FXC compiler.
-
Cache-to-Cache: Direct Semantic Communication Between LLMs (2025) arxiv.org
Current multi-LLM systems rely on text-based communication, which loses rich semantic data and slows down response times due to token generation.
-
Shapelearn Qwen 3.8 27B (13.1 GB VRAM) byteshape.com
The ByteShape team released full ShapeLearn quantized models for Qwen 3.8 27B, which outperform their previous Lite versions and several competing quantization methods.
-
Diplodocus, Long Thought Exclusively American, Turns Up in Spain sci.news
Paleontologists have discovered the first confirmed fossils of the Diplodocus genus outside of North America, found in the Teruel province of Spain.
-
Telstra outage: The night a network decided the year was 2006 netnod.se
The 2026 Telstra outage demonstrates that precision time synchronization is a critical design dependency for modern mobile networks, particularly 5G TDD systems.
-
Mathematicians Build Long-Awaited Graph Sandwich quantamagazine.org
Mathematicians have finally proven the 'sandwich conjecture,' demonstrating that complex random regular graphs can be mathematically nested between simpler random binomial graphs.
-
Cyclomatic Complexity in C# blog.ndepend.com
Cyclomatic Complexity (CC) is a vital metric for quantifying a method's structural complexity by counting independent execution paths.
-
The Implications of Linguistic Illegibility for LLM Security arxiv.org
Large Language Models operate using mathematical operations over activation spaces rather than natural language, creating a gap called linguistic illegibility.
-
Apple detectives solved mystery of ancient tree and rewrote the history of fruit scientificamerican.com
Researchers and amateur apple hunters are using DNA fingerprinting to identify ancient, lost apple varieties in North America to recover genetic traits for disease resistance and climate resilience.
-
The Shadows Lurking in the Equations – Underwater Islands gods.art
Conventional graphing only shows where equations are exactly equal, which hides significant mathematical topography.
-
Subnormal floating-point numbers are expensive on Intel processors lemire.me
Subnormal floating-point numbers cause severe performance degradation on modern Intel processors, while AMD and Arm processors handle them with little to no penalty.
-
A search-and-inference database from scratch in pure Zig antfly.io
Antfly transitioned its core engine from Go to Zig to achieve extreme reliability and performance by eliminating external dependencies and gaining total control over resource management.
-
Ask A Monk – A digital wilderness for thoughts with no immediate answer askamonk.online
Ask A Monk provides a digital sanctuary where individuals can seek guidance and support from strangers without the interference of artificial intelligence.
-
Show HN: Scry, programmable internet search w/ congestion pricing scry.io
Scry is a Model Context Protocol (MCP) server that allows AI agents to run complex, Turing-complete SQL-like queries over billions of internet documents from sources like Reddit, Hacker News, and arXiv.
-
NATS publishes preliminary report on technical incident of 8 September nats.aero
A software defect in the National Airspace System caused widespread flight disruptions across the UK on September 8.
-
Speeding up gearhash on ARM64 sam.dev
The gearhash Rust crate has been updated to include a NEON backend for ARM64, achieving nearly twice the speed of the scalar implementation.
-
Size-Specialized Memory Allocation go.dev
Go 1.27 introduces size-specialized memory allocation for objects 80 bytes or smaller to improve performance.
-
Code Scans devin.ai
Cognition is introducing Code Scans, a tool powered by Agentic MapReduce that transforms high-level engineering goals into actionable pull requests.