A German ISP changed their DNS to block my website

799 shaunpud 2025-08-24 10:27:19 UTC lina.sh Llama Share 433

Germany's CUII, a private organization, blocks websites without transparency, and a website cuiiliste.de tracks blocked domains. After Netzpolitik exposed CUII's mistake, ISPs like Telefonica altered their DNS servers to make it harder to confirm CUII blocks.

A visual introduction to big O notation

671 samwho 2025-08-24 07:39:36 UTC samwho.dev Llama Share 192

Big O notation describes a function's performance growth as input size increases, with categories including constant, logarithmic, linear, and quadratic. Understanding time complexity helps improve algorithmic choices and avoid common pitfalls, such as O(n^2) growth, to optimize code performance.

Comet AI browser can get prompt injected from any site, drain your bank account

634 helloplanets 2025-08-24 15:14:34 UTC twitter.com/zack_overflow Full Share 218

Something went wrong, but don’t fret — let’s give it another shot. Some privacy related extensions may cause issues on x.com. Please disable them and try again.

How to build a coding agent

466 ghuntley 2025-08-24 03:21:14 UTC ghuntley.com Llama Share 122

The speaker, Geoffrey Huntley, explains how to build a coding agent using a highly agentic model like Claude Sonnet, and demonstrates the process by creating a coding agent that can read files, list files, execute bash commands, edit files, and search source code. The speaker emphasizes the importance of personal development and investing in oneself to stay ahead in the tech industry, which ...

Everything I know about good API design

420 ahamez 2025-08-24 19:10:09 UTC seangoedecke.com Llama Share 155

API design should focus on simplicity and familiarity for users, with a balance between simplicity and flexibility, and careful consideration of changes to avoid breaking existing software. Key principles for good API design include supporting idempotent operations, using rate limiting and killswitches, and making APIs accessible to non-engineers with simple authentication and clear error handling.

We put a coding agent in a while loop

417 sfarshid 2025-08-24 16:18:53 UTC github.com/repomirrorhq Full Share 303

You have triggered an abuse detection mechanism. Please wait a few minutes before you try again; in some cases this may take up to an hour.

Ghrc.io appears to be malicious

402 todsacerdoti 2025-08-24 23:27:52 UTC bmitch.net Llama Share 73

A typo in ghcr.io led to a malicious credential stealing attack, where users' GitHub credentials were stolen when trying to access the registry. This attack occurred when users stored credentials for the typoed host, allowing attackers to push malicious images or gain direct access to GitHub accounts.

A bug saved the company

377 ingve 2025-08-24 09:55:56 UTC weblog.rogueamoeba.com Llama Share 130

A software bug in Audio Hijack 1.6 accidentally limited the trial period to 15 minutes instead of 15 days, leading to higher sales and saving the company. This fortunate mistake became the basis for the company's trial limitations and helped Rogue Amoeba become a successful business.

Is 4chan the perfect Pirate Bay poster child to justify wider UK site-blocking?

360 gloxkiqcza 2025-08-24 16:30:18 UTC torrentfreak.com Llama Share 813

The UK's Online Safety Act has faced criticism for censoring legitimate content and restricting free speech, with some sites banning UK visitors to avoid fines. The government's approach has been met with resistance from VPN users and platforms like 4chan, which may lead to further escalation and potential intervention at the highest level.

Making games in Go: 3 months without LLMs vs. 3 days with LLMs

352 maloga 2025-08-24 15:01:24 UTC marianogappa.github.io Llama Share 223

User built a card game Truco in free time after 15 years as a software engineer, then used LLM to quickly build a new game Escoba with a refactored backend. User shares a minimalistic primer to make a game with the same stack, including a Tic-Tac-Toe game set of repos for others to fork and start with.

Neuralink 'Participant 1' says his life has changed

352 danielmorozoff 2025-08-24 09:12:14 UTC fortune.com Llama Share 439

Noland Arbaugh, a 31-year-old man with paralysis, became the first person to receive Elon Musk's experimental brain chip, Neuralink, in January 2024. The device allows him to control a computer with his mind, giving him newfound autonomy and purpose.

Uncle Sam shouldn't own Intel stock

330 aspenmayer 2025-08-24 21:40:35 UTC wsj.com Full Share 355

Please enable JS and disable any ad blocker

YouTube made AI enhancements to videos without warning or permission

291 jakub_g 2025-08-24 10:37:15 UTC bbc.com Llama Share 277

YouTube secretly used AI to tweak videos without permission, making skin sharper and ears warp. This move raises concerns about trust and the blurring of reality.

Chinese astronauts make rocket fuel and oxygen in space

290 Teever 2025-08-24 17:54:47 UTC livescience.com Llama Share 131

Chinese astronauts created rocket fuel and oxygen in space using artificial photosynthesis technology on the Tiangong space station. This breakthrough could aid China's proposed moon base, enabling breathable air and fuel for return flights to Earth.

It is worth it to buy the fast CPU

266 ingve 2025-08-24 06:03:15 UTC blog.howardjohn.info Llama Share 510

Corporations are spending $500/year on AI coding tools but could upgrade to top-end CPUs for $170/year, a significant cost difference. Upgrading to top-end CPUs can greatly improve productivity, making them a more cost-effective option.

Valve Software handbook for new employees [pdf] (2012)

242 Michelangelo11 2025-08-24 08:02:12 UTC cdn.akamai.steamstatic.com Llama Share 202

Valve's unique approach to hiring and project management emphasizes self-directed work, collaboration, and constant learning, with a focus on empowering employees to make decisions and take ownership of their projects. The company's flat structure and lack of traditional hierarchy allow for flexibility and innovation, but also require employees to be proactive and responsible in prioritizing ...

US attack on renewables will lead to power crunch that spikes electricity prices

234 rntn 2025-08-24 14:22:36 UTC cnbc.com Llama Share 305

President Trump's attack on solar and wind projects threatens to raise energy prices and worsen a looming power supply shortage. Renewable energy executives warn that blocking permits will harm the electric grid and lead to higher electricity prices for consumers.

Show HN: Clearcam – Add AI object detection to your IP CCTV cameras

230 roryclear 2025-08-24 11:34:15 UTC github.com/roryclear Llama Share 66

Sign up for Clearam premium with iOS app to receive camera streams and notifications. Use iOS-generated User ID on Android for access.

DeepWiki: Understand Any Codebase

229 childishnemo 2025-08-24 07:23:30 UTC aitidbits.ai Llama Share 47

The author uses DeepWiki to understand unfamiliar codebases, generate context for coding agents, and save time in their day-to-day workflow. DeepWiki turns GitHub repositories into instant, navigable wikis, providing answers to questions and clickable citations to source files.

Turning Claude Code into my best design partner

222 scastiel 2025-08-24 08:06:03 UTC betweentheprompts.com Llama Share 83

The user improved their approach to working with Claude Code by asking it to create a plan document for complex tasks, which becomes the source of truth and helps avoid context limit issues. By treating the plan document as a living document and updating it during implementation, the user can efficiently continue the conversation and ensure a clear record of the development process.

Omarchy Is Out

215 kristianp 2025-08-24 04:36:54 UTC world.hey.com Llama Share 133

The user published a Linux setup called Omarchy, an opinionated configuration for Arch Linux and Hyprland. It's designed to make Linux easier to use, especially for those who find Hyprland difficult to configure.

The Annotated Transformer (2022)

206 subset 2025-08-24 09:58:13 UTC nlp.seas.harvard.edu Llama Share 12

The Transformer model uses self-attention to compute representations of input and output sequences without sequence-aligned RNNs or convolution, reducing sequential computation to a constant number of operations. It employs multi-head attention, residual connections, and layer normalization to improve performance.

AGI is an engineering problem, not a model training problem

192 vincirufus 2025-08-24 00:18:52 UTC vincirufus.com Llama Share 421

We've reached an inflection point in AI development where scaling laws are showing diminishing returns. To achieve artificial general intelligence, we need to build engineered systems that combine models, memory, context, and deterministic workflows.

Paracetamol disrupts early embryogenesis by cell cycle inhibition

190 XzetaU8 2025-08-24 18:02:55 UTC academic.oup.com Full Share 195

Halt and Catch Fire Syllabus (2021)

182 Kye 2025-08-24 20:19:55 UTC bits.ashleyblewer.com Llama Share 77

This website offers a 15-class course based on the TV series Halt and Catch Fire, exploring tech history through discussions and episode summaries. The curriculum is designed for self-formed groups to analyze tech history in the 1980s-1990s.

Nvidia DGX Spark

179 janandonly 2025-08-24 22:35:23 UTC nvidia.com Llama Share 191

NVIDIA DGX Spark delivers 1 petaFLOP AI performance with 128GB memory and preinstalled AI software stack. It enables developers to prototype, fine-tune, and deploy AI models locally or in the cloud with up to 200 billion parameters.

Show HN: Sping – An HTTP/TCP latency tool that's easy on the eye

176 zorlack 2025-08-24 23:42:01 UTC dseltzer.gitlab.io Llama Share 24

sping is a modern terminal tool for HTTP/TCP latency monitoring with real-time visualization. It offers interactive charts, outlier detection, and customizable color schemes with terminal color support.

Dynamically patch a Python function's source code at runtime

156 apwheele 2025-08-24 12:28:52 UTC ericmjl.github.io Llama Share 77

The user created a Python function called write_and_execute_code that allows an LLM to execute custom Python code with access to the current runtime environment's variables and functions. This function is part of a tool-calling agent called ToolBot that separates tool selection from execution, enabling more flexible and controllable systems than monolithic agents.

SmallJS: Smalltalk-80 that compiles to JavaScript

149 mpweiher 2025-08-24 09:29:02 UTC small-js.org Llama Share 45

SmallJS is a free, open-source implementation of Smalltalk-80 that compiles to JavaScript for modern browsers or Node.js. It's file-based, object-oriented, and customizable with automatic import of used libraries.

A bubble that knows it's a bubble

141 craigmccaskill 2025-08-24 22:02:52 UTC craigmccaskill.com Llama Share 124

Sam Altman, CEO of OpenAI, warned about AI overexcitement, causing market reactions. The AI bubble is similar to past bubbles, such as Railway Mania and the dot-com bubble, with excessive valuations and speculation.

ICE uses celebrity loophole to hide deportation flights

140 JKCalhoun 2025-08-24 12:40:49 UTC jacobin.com Llama Share 34

The Federal Aviation Administration's LADD program allows private jets to hide flight records from public view, used by celebrities and now ICE for deportation flights. This program was created by the private jet lobby to protect billionaires' and celebrities' jets from public tracking.

Internet Access Providers Aren't Bound by DMCA Unmasking Subpoenas–In Re Cox

138 hn_acker 2025-08-24 19:20:46 UTC blog.ericgoldman.org Llama Share 16

The Ninth Circuit ruled that copyright owners cannot use 512(h) subpoenas to unmask Internet users through Internet access providers due to the providers' 512(a) immunity. The court held that the DMCA does not permit 512(h) subpoenas to issue to 512(a) service providers.

How to make things slower so they go faster

133 neehao 2025-08-24 05:10:10 UTC gojiberries.io Llama Share 38

Synchronized demand causes peaks when a large cohort of clients acts together, leading to queues, timeouts, and retries. To prevent or drain such peaks, use jittered scheduling, pacing admissions, and scaling to allocate delay fairly and minimize overload risk while respecting measurable constraints.

Starship's Tenth Flight Test

126 d_silin 2025-08-24 21:22:07 UTC spacex.com Llama Share 239

The tenth Starship flight test is preparing to launch on Sunday, August 24, with a live webcast starting 30 minutes before liftoff. The test aims to expand the operating envelope of the Super Heavy booster and gather data on future flight profiles and reentry experiments.

Looking back at my transition from Windows to Linux

125 trinsic2 2025-08-24 20:22:26 UTC scottrlarson.com Llama Share 224

A computer professional switched from Windows to Linux after 25 years of troubleshooting, citing Microsoft's aggressive tactics and loss of user control. He now advocates for Linux as a viable alternative and encourages users to take control of their digital lives and resist corporate control.

Show HN: Port Kill – A lightweight macOS status bar development port monitor

125 lexokoh 2025-08-24 03:08:15 UTC github.com/kagehq Llama Share 47

The app monitors and manages development processes on specified ports, allowing real-time detection and process killing. It supports various command-line options for customization and debugging.

In-Memory Filesystems in Rust

124 ingve 2025-08-24 06:33:35 UTC andre.arko.net Llama Share 55

The user searched for a Rust equivalent to the Go package Afero for in-memory filesystem testing but found no suitable alternatives, with vfs and rsfs having limitations such as no symlink support and poor type signatures. The user's benchmarks showed no significant performance difference between using an in-memory filesystem and the regular filesystem on modern SSDs and macOS, suggesting ...

Show HN: Bicyclopedia

122 lemoing 2025-08-24 11:15:26 UTC bicyclopedia.lemoing.ca Full Share 40

If you are using a screen reader or have Javascript disabled, unfortunately you won't be able to see the images or animations of the parts, but you can still read the descriptions and source code here and read more about this project here.

Seed: Interactive software environment based on Common Lisp

120 todsacerdoti 2025-08-24 06:57:19 UTC github.com/phantomics Llama Share 21

Seed is an interactive software environment based on Common Lisp that allows building software locally or remotely and presents programs in various display modes. It integrates ASDF and Quicklisp, and can be installed using a Lisp file and Node.js dependencies.

Turning a Decommissioned iPhone into a UniFi Protect Camera

119 ingve 2025-08-24 06:13:33 UTC caseyliss.com Llama Share 51

User repurposed an old iPhone as a third-party camera in UniFi Protect using IP Camera Lite and Docker. They successfully integrated the iPhone camera into Protect, retiring previous camera management tools.

Evaluating LLMs for my personal use case

118 goranmoomin 2025-08-24 00:40:33 UTC darkcoding.net Llama Share 46

The user conducted an evaluation of various AI models, including reasoning and non-reasoning models, using 130 real prompts from their bash history and found that most models performed well, with speed and cost being key factors. The user recommends using multiple models at once, as seen in their tmux script, and suggests that even expensive models like Sonnet may not be necessary, given the ...

US halts work on almost finished wind farm because national security

117 hvb2 2025-08-24 09:23:46 UTC npr.org Llama Share 83

The Trump administration has ordered a halt to the Revolution Wind project off Rhode Island's coast citing national security concerns. The move is the latest in a series of actions targeting the US renewable energy industry.

Show HN: I Built a XSLT Blog Framework

116 vgr-land 2025-08-24 17:38:30 UTC vgr.land Llama Share 68

The user created a blog publishing framework using XSLT to simplify content publishing and separate presentation from content. This framework allows for dynamic menu generation and RSS creation without requiring a build pipeline or server-side scripting, and is licensed under MIT.

Using information theory to solve Mastermind

115 SchwKatze 2025-08-24 02:54:11 UTC goranssongaspar.com Llama Share 37

A Mastermind strategy involves guessing the code that provides the most information on average, calculated by entropy. This strategy beats the Code Master in 4.47 guesses on average, which is comparable to other Mastermind algorithms.

'Rocks as big as cars' are flying down the Dolomites

114 bookofjoe 2025-08-24 17:03:44 UTC bbc.com Llama Share 57

The Dolomites mountain range in Italy is experiencing frequent landslides and rockfalls due to climate change and erosion. Scientists are monitoring the situation and focusing on alerting the population rather than trying to prevent the natural process of the mountains coming down.

Tesla insiders have sold more than 50% of their shares in the last year

110 MilnerRoute 2025-08-24 00:36:38 UTC electrek.co Llama Share 101

Tesla insiders, excluding CEO Elon Musk, sold over 50% of their shares in the last year. Top executives, including Tom Zhu, significantly reduced their stakes, with Zhu's decreasing by 82%.

Writing with LLM is not a shame

107 flornt 2025-08-24 10:10:29 UTC reflexions.florianernotte.be Llama Share 138

The user discusses the importance of transparency in AI use, questioning whether disclaiming AI-generated content is necessary, and proposes that sourcing content is a more significant issue. They argue that AI-generated content can be valuable and that disclaiming AI use may create bias and a "thought police" mentality.

SQLite (with WAL) doesn't do `fsync` on each commit under default settings

105 Bogdanp 2025-08-24 15:40:38 UTC avi.im Llama Share 90

SQLite's WAL mode offers higher write throughput but loses durability. Using synchronous=FULL in WAL mode ensures transaction durability across power loss.

Ask HN: Best codebases to study to learn software design?

103 pixelworm 2025-08-24 05:08:19 UTC news.ycombinator.com Llama Share 90

You should study publicly accessible codebases like Linux, Apache, and Git, which are considered gold standards for software design. These codebases demonstrate effective design patterns and tradeoffs in various contexts, but keep in mind that "good" design depends on the specific project requirements and team structure.

How to check if your Apple Silicon Mac is booting securely

100 shorden 2025-08-24 23:44:47 UTC eclecticlight.co Llama Share 22

To check your Mac's security, open System Information and select Controller, then check Startup Security Utility in Recovery mode for Secure Boot settings. Enable FileVault and verify XProtect/Gatekeeper checks are enabled in Privacy & Security settings.

DeepConf: Scaling LLM reasoning with confidence, not just compute

98 che_shr_cat 2025-08-24 14:41:58 UTC arxiviq.substack.com Llama Share 34

Researchers introduced DeepConf, a test-time inference method that enhances LLM reasoning capabilities by leveraging internal log-probabilities to derive localized confidence scores. This method achieves state-of-the-art accuracy while reducing computational costs by up to 84.7%.

Agentic Browser Security: Indirect Prompt Injection in Perplexity Comet

96 drak0n1c 2025-08-24 02:52:08 UTC brave.com Llama Share 31

Researchers at Brave discovered a vulnerability in Perplexity's Comet AI assistant that allows attackers to inject malicious instructions through webpage content, enabling cross-domain access and sensitive data exfiltration. The attack highlights the need for new security and privacy architectures in agentic browsing.

A Brilliant and Nearby One-off Fast Radio Burst Localized to 13 pc Precision

95 gnabgib 2025-08-24 19:23:19 UTC iopscience.iop.org Llama Share 31

The CHIME/FRB Collaboration localized a nearby one-off FRB source, FRB 20250316A, to its environment on 13 pc scales using the full CHIME Outrigger array. The source is associated with the nearby galaxy NGC 4141 and has a 5σ upper limit of L9.9 GHz < 2.1 × 1025 erg s−1 Hz−1 on spatially coincident radio emission.

How to Fix Your Context

91 itzlambda 2025-08-24 07:09:53 UTC dbreunig.com Llama Share 26

To mitigate long context failures, consider using Retrieval-Augmented Generation (RAG), Tool Loadout, Context Quarantine, Context Pruning, Context Summarization, and Context Offloading to manage information and prevent context distractions. These tactics help prevent context poisoning, confusion, and clashes by selectively adding relevant information, isolating contexts, removing irrelevant ...

Spending too much time at airports

89 nsoonhui 2025-08-24 11:40:37 UTC thezvi.substack.com Llama Share 138

The author suggests that arriving at the airport one hour before a domestic flight is sufficient, with some extra time added for PreCheck, checking bags, and slow airports. They also recommend considering factors like peace of mind, stress levels, and the importance of the flight when deciding how much time to spend at the airport.

Wildthing – A model trained on role-reversed ChatGPT conversations

89 iamwil 2025-08-24 05:51:43 UTC youaretheassistantnow.com Full Share 36

A model trained on role-reversed ChatGPT conversations

You shouldn't salt a leech that's sucking your blood (2019)

88 pabs3 2025-08-24 03:12:05 UTC cbc.ca Llama Share 69

Leeches can be removed by breaking the seal between their mouth and skin with a fingernail or credit card, or by waiting for them to finish feeding. Leeches prefer to feed in crevices like the groin due to their slow eating pace and need for a safe spot.

Reverse Engineered Raspberry Pi Compute Module 5

87 _Microft 2025-08-24 20:34:49 UTC github.com/schlae Llama Share 23

A user has reverse engineered the Raspberry Pi Compute Module 5 schematic and layout for educational purposes. The design is not meant for fabrication and lacks critical signal integrity parameters.

Bring Back the Blue-Book Exam

84 diodorus 2025-08-24 17:44:55 UTC chronicle.com Llama Share 121

A professor of communication is using handwritten exams to counter AI-generated work, finding students' handwritten essays convey their humanity and restore trust in the student-teacher relationship. This shift towards in-person assessments is part of a larger movement to rethink education in the age of AI.

Neural Nets vs. Cellular Automata

83 todsacerdoti 2025-08-24 06:39:01 UTC nets-vs-automata.net Full Share 12

\[ C_{i,t} = \begin{cases} 1 & \text{if cell } i \text{ is ON}_t\\ 0 & \text{if cell } i \text{ is OFF}_t \end{cases} \] \[ \sum_{i \in \text{neighbor}} C_{i, t} \text{ } \in S \implies \text{ON}_{t+1} \] \[ \sum_{i \in \text{neighbor}} C_{i, t} \text{ } \notin S \implies \text{OFF}_{t+1} \]

US Supreme Court allows NIH to cut $2B in research grants

82 rntn 2025-08-24 13:49:06 UTC nature.com Llama Share 13

The US Supreme Court ruled that researchers cannot challenge grant terminations in district court, but allowed review of NIH guidelines used for terminations. The decision effectively blocks researchers' efforts to reinstate $2 billion in grants cut by the Trump administration.

Germany's Copyright Clearing House now requires courts for website blocks

79 nsdfg 2025-08-24 12:06:17 UTC heise.de Llama Share 23

The Copyright Clearing House for the Internet (CUII) has declared success in fighting online copyright infringement, but will now involve courts in reviewing website restrictions. This change aims to ensure only justified blocks are implemented and guarantee freedom of information.

Adventures in State Space [video]

78 bo0tzz 2025-08-24 09:49:54 UTC youtube.com Embed Share 10

Explanation of the Linux-kernel memory consistency model (2017)

77 Bogdanp 2025-08-24 10:00:04 UTC raw.githubusercontent.com Llama Share 0

The Linux-kernel memory consistency model (LKMM) is a complex model that predicts the values that can be obtained by load instructions in the Linux kernel, taking into account the interactions between multiple CPUs accessing shared memory locations. The LKMM is defined in terms of events, such as read, write, and fence events, and uses various relations, including program order, dependency ...

Show HN: FilterQL – A tiny query language for filtering structured data

75 genshii 2025-08-24 18:55:01 UTC github.com/adamhl8 Llama Share 21

FilterQL is a query language for filtering data, allowing users to specify conditions using a schema and supporting operations like sorting and rounding. The language is designed to be extensible and customizable, with features like field aliases, custom operations, and support for unknown fields.

Equal Earth – Political Wall Map (2018)

75 bjelkeman-again 2025-08-24 05:54:15 UTC equal-earth.com Llama Share 53

The Equal Earth Wall Map is a free, large-scale map showing countries and continents at their true sizes, available for download in three versions. It's designed for schools and organizations, with a professional design and 2,600+ map labels.

The great medieval water myth (2013)

74 apsec112 2025-08-24 17:25:47 UTC leslefts.blogspot.com Llama Share 50

People in the Middle Ages drank water regularly and it was not considered unhealthy unless it appeared bad. Medical authorities of the time did not recommend avoiding water due to disease.

The AI vibe shift is upon us

73 lelele 2025-08-24 10:31:09 UTC cnn.com Llama Share 100

Investors are bracing for a potential AI market correction due to industry stumbles and failed revenue expectations. Companies like Meta and Coreweave are experiencing setbacks, and traders are buying "disaster puts" as a precaution.

ThinkMesh: A Python lib for parallel thinking in LLMs

72 martianlantern 2025-08-24 04:23:33 UTC github.com/martianlantern Llama Share 4

ThinkMesh is a Python library for parallel reasoning and scoring diverse paths with internal confidence signals. It supports offline Hugging Face Transformers and vLLM/TGI, and hosted APIs.

Programming People (2016)

72 saulpw 2025-08-24 03:09:01 UTC leftoversalad.com Full Share 19

special thanks: ScudsCorp, wycats, S. Sever

Mathematical secrets of ancient tablet unlocked after nearly a century of study (2017)

68 surement 2025-08-24 21:31:34 UTC theguardian.com Llama Share 63

A 3,700-year-old Babylonian clay tablet contains the world's oldest and most accurate trigonometric table, which could have been used for surveying and architecture. The tablet, known as Plimpton 322, demonstrates undoubted genius and has potential modern applications in mathematics, surveying, and computer graphics.

The SD Association has an official SD card format utility [Win/OS X/Linux]

66 Almondsetat 2025-08-24 08:55:56 UTC sdcard.org Llama Share 56

The SD Memory Card Formatter formats SD cards according to the SD File System Specification. It's recommended to use this formatter for optimal performance and security.

Playing every game of Wordle simultaneously

63 chriskw 2025-08-24 15:34:26 UTC chriskw.xyz Llama Share 28

A Wordle variant called Hyper Wordle involves solving 2315 words at once, and a strategy combining two suboptimal strategies can outperform a single optimal strategy. By using deduction tricks, the average score can be improved from 7921.5 to 7598.3, beating the best Wordle strategy.

China's Guowang megaconstellation is more than another version of Starlink

62 rbanffy 2025-08-24 17:43:14 UTC arstechnica.com Llama Share 28

China's Guowang satellite network may provide a tactical edge in future conflicts in the Western Pacific, with capabilities similar to the US military's proposed MILNET and Space Development Agency's future constellation. The US military is concerned that China's pace of launching Guowang satellites could give them a head start on completing a celestial kill chain.

From Hackathon to YC

61 rmason 2025-08-24 22:08:26 UTC producthunt.com Llama Share 23

Neha, founder of April, shares her story of creating an AI executive assistant after a hackathon win and joining Y Combinator. April helps users manage their inbox, calendar, and meetings, giving them back their time.

Using acetaminophen during pregnancy may increase childrens autism and ADHD risk

55 spchampion2 2025-08-24 23:26:46 UTC hsph.harvard.edu Llama Share 15

A new study found that using acetaminophen during pregnancy may increase children's risk of autism and ADHD. Researchers recommend judicious use of acetaminophen under medical guidance to minimize risks.

People stuck using ancient Windows computers

53 erickhill 2025-08-24 17:16:09 UTC bbc.com Llama Share 39

Microsoft's legacy continues to impact society through its widespread use in outdated software and devices, including ATMs, trains, and medical facilities. Many organizations rely on these systems due to their reliability and stability, despite the challenges of maintaining and upgrading them.

Integer continued fractions for complex numbers

53 bikenaga 2025-08-24 00:09:57 UTC arxiv.org Llama Share 2

Researchers extended complex numbers to standard continued fractions using Lagrange and Gauss algorithms, resulting in unique representations with useful properties. These new representations also have a geometric cutting sequence interpretation.

Vanguard hits new 'bans-per-second' record

50 Wingy 2025-08-24 23:41:35 UTC playvalorant.com Llama Share 66

Riot's anti-cheat team, Vanguard, saw a brief increase in cheaters in popular regions like Brazil, but they've since been rapidly decreasing with new updates. The team is working to prevent cheating by improving detection, banning repeat offenders, and introducing features like Ranked Rollback to restore lost rank to affected players.

2.5B Gmail users endangered after Google database hack

41 alhazraed 2025-08-24 22:22:26 UTC pcworld.com Llama Share 16

Gmail users are at risk of phishing attempts due to a data leak from Google's Salesforce database. To stay safe, use Google's Security Checkup, Advanced Protection Program, and passkeys, and be cautious of unsolicited contact from alleged support staff.

Molluscs of the Multiverse: molluscan diversity in Magic: The Gathering

41 zdw 2025-08-24 15:53:58 UTC jgeekstudies.org Llama Share 23

Molluscs are the second most diverse animal phylum in sheer number of species, with diverse body plans and shapes, and are found in various forms across the multiverse in Magic: The Gathering. The multiverse features a wide range of molluscan creatures, including snails, slugs, octopuses, squids, and bivalves, with some species exhibiting unique characteristics and abilities not found in our ...

Update on my Racket exit

35 todsacerdoti 2025-08-24 17:27:19 UTC blog.winny.tech Llama Share 18

The user has abandoned their Racket projects, archiving or handing them off to others, and is now focusing on other languages like Python and Bash. They emphasize the importance of letting go of projects and creating a smooth hand-off process for long-term success in software development.

GNU cross-tools: musl-cross 313.3M

34 1vuio0pswjnm7 2025-08-24 22:20:07 UTC github.com/cross-tools Full Share 17

Cross-compile toolchain with musl libc. Contribute to cross-tools/musl-cross development by creating an account on GitHub.

How many paths of length K are there between A and B? (2021)

34 jxmorris12 2025-08-24 20:09:12 UTC horace.io Llama Share 12

The problem is to find the number of paths of length K from node A to node B in a directed unweighted graph with V vertices and E edges, and the solution involves using matrix exponentiation and linear recurrences to find the K-th term of the sequence in V3+V^3 +V3+m \cdot V \log V \log K time. The solution uses the Cayley-Hamilton theorem to convert the matrix exponentiation problem into a ...

Mob Programming (2018)

32 ustad 2025-08-24 05:45:13 UTC mobprogramming.org Llama Share 20

The author introduces the concept of "Fading Problems" where Mob Programming helps eliminate common software development issues such as communication problems, decision-making reluctance, technical debt, and thrashing. By working together in a collaborative and agile way, the team is able to "turn up the good" and eliminate many of these problems, leading to improved productivity and efficiency.

Y Combinator says Apple's App Store has hindered startup growth

30 fauigerzigerk 2025-08-24 17:22:53 UTC techcrunch.com Llama Share 2

Y Combinator filed an amicus brief in support of Epic Games, arguing against Apple's appeal of a ruling that allows developers to offer alternative payment options. The brief claims the Apple Tax stifles competition and innovation, making it easier for Y Combinator to invest in app-based businesses.

A lightweight TypeScript library for assertion-based runtime data validation

30 nayajunimesh 2025-08-24 04:43:34 UTC github.com/nimeshnayaju Llama Share 35

decode-kit is a lightweight TypeScript library for validating runtime data with minimal overhead. It refines types in-place without cloning or transformation.

OOMProf: Profiling on the Brink

26 ingve 2025-08-24 10:00:55 UTC polarsignals.com Llama Share 8

OOMProf is a system that uses eBPF to profile Go programs at the point they are OOM killed, capturing allocations up to the bitter end to give developers a better idea of what went wrong. It can handle up to 60k buckets and is available as a library or as part of the Parca Agent, which can automatically upload OOM memory profiles to Polar Signals cloud or an on-prem Parca instance.

Introducing: VuIO – open-source UPNP/DLNA server written in Rust

25 vyrti 2025-08-24 22:31:22 UTC github.com/vuiodev Llama Share 1

VuIO is a comprehensive DLNA/UPnP media server written in Rust with advanced platform integration and robust database management. It supports multiple media directories, playlist management, and various audio and video formats.

Inflammation predicts 25% higher mortality

24 brandonb 2025-08-24 23:01:44 UTC thelancet.com Llama Share 4

A large study with over 5000 patients found that high sensitivity C-reactive protein (hsCRP) levels are associated with increased long-term cardiovascular risk. Elevated hsCRP levels predicted future cardiovascular events, including myocardial infarction and all-cause mortality, even in individuals without established cardiovascular disease.

What Are Oklch Colors?

21 vinhnx 2025-08-24 03:22:16 UTC jakub.kr Llama Share 4

OKLCH is a color model that provides perceptually uniform colors, making it easier to work with. It consists of three values: Lightness, Chroma, and Hue, and is supported in modern browsers with fallbacks available for outdated browsers.

Show HN: CasCache – multi-generational cache with optimistic concurrency control

20 unkn0wn_root 2025-08-24 11:24:40 UTC github.com/unkn0wn-root Llama Share 2

This is a provider-agnostic CAS cache with pluggable codecs and a pluggable generation store, offering safe single-key reads and optional bulk caching with read-side validation. It supports distributed mode for multi-replica deployments and can use a shared GenStore like Redis for cross-replica correctness and warm restarts.

A Ritzy L.A. Enclave Learned a Bitter Lesson About the Limits of Its Wealth

19 wallflower 2025-08-24 21:15:50 UTC nytimes.com Full Share 43

Please enable JS and disable any ad blocker

"I Hate Almost Everything They Do with a Passion" James Pond Creator on Gameware

16 austinallegro 2025-08-24 22:27:11 UTC timeextension.com Llama Share 1

Chris Sorrell, creator of James Pond, is unhappy with Gameware's handling of the character and their new game Rogue AI, citing poor creativity and excessive use of AI-generated content. He feels hurt and powerless to stop the degradation of the character he created.

DeepCode: Open Agentic Coding

13 pykello 2025-08-24 01:09:06 UTC github.com/hkuds Llama Share 3

DeepCode is an AI-powered development platform that automates code generation and implementation tasks. It leverages a multi-agent system to translate requirements into functional, well-structured code.

YouTube used AI to edit people's videos. The results could bend reality

11 quantified 2025-08-24 21:59:45 UTC bbc.com Llama Share 4

YouTube secretly used AI to tweak videos without permission, making skin sharper and ears warp. This move raises concerns about trust and the blurring of reality.

Not so prompt: Prompt optimization as model selection (2024)

11 neehao 2025-08-24 01:35:54 UTC gojiberries.io Llama Share 0

Define primary metrics like accuracy or BLEU/ROUGE and auxiliary constraints like output format compliance and safety requirements. Determine the required data amount for statistically valid comparisons and use random sampling to represent the true input distribution.