60-Year-Old Math Puzzle Solved: The Optimal Sofa Size

2025-02-14
60-Year-Old Math Puzzle Solved: The Optimal Sofa Size

A 60-year-old mathematical puzzle – the moving sofa problem – has finally been solved! In the 1960s, mathematicians posed a seemingly simple geometric question: What's the largest area of a sofa that can navigate a unit-width hallway? Recently, Jineon Baek, a postdoctoral researcher at Yonsei University in Seoul, proved in a 119-page paper that the sofa shape proposed by Joseph Gerver in 1992 is the optimal solution, with an area of approximately 2.2195. Baek's proof is remarkable because it didn't rely on computers but used elegant mathematical techniques, offering new approaches to solving other optimization problems. The result also illustrates that even the simplest optimization problems can have surprisingly complex answers.

Read more

Revolutionizing Memory: Atomic-Scale Crystal Defects Unlock New Storage Potential

2025-02-14
Revolutionizing Memory: Atomic-Scale Crystal Defects Unlock New Storage Potential

Researchers at the University of Chicago have achieved a breakthrough in classical computer memory efficiency by harnessing crystal defects. They created memory cells from single missing atoms within a crystal structure, each capable of storing a bit. This innovative approach promises terabytes of data compressed into a cubic millimeter, revolutionizing data storage. The research integrates solid-state physics and radiation dosimetry, offering unprecedented high-density storage for classical non-volatile memory.

Read more

Safe and Efficient printf in Idris: No Macros Required

2025-02-14

This article demonstrates how to implement a safe and efficient printf function in Idris without resorting to unsafe macros or variadics. By cleverly using type-level programming, the author parses the format string into a data structure and dynamically generates the function type signature based on it. This achieves the functionality of C's printf while maintaining memory and type safety. The article also explores handling runtime format strings and points out shortcomings of the implementation, such as unclear error messages, hinting at future improvements.

Read more

Mathics: A Modular Math Environment with Multiple Deployment Options

2025-02-14

Mathics is a modularly designed mathematical computation environment offering various deployment options. Users can quickly deploy a complete environment via a Docker image or install it locally using the Mathics-omnibus Python package. At its core is the Mathics3 kernel, complemented by the mathicsscript command-line client (featuring syntax highlighting, Unicode support, etc.) and a Django-based web server (with MathML output and Three.js graphics). These components have individual dependencies, but the modular design ensures flexibility and scalability.

Read more

Final Fantasy Crystal Chronicles iOS Version Shut Down Due to Unfixable Bug

2025-02-14
Final Fantasy Crystal Chronicles iOS Version Shut Down Due to Unfixable Bug

Square Enix has shut down the iOS version of Final Fantasy Crystal Chronicles and removed it from the App Store due to an unfixable bug preventing access to purchased content. The bug stemmed from changes to the in-app purchase model. Players who made in-app purchases in January 2024 or later can contact Apple Support for a refund. The game remains available on Android, PlayStation, and Nintendo Switch.

Read more

Ricochet: Anonymous Messaging You Can Trust

2025-02-14
Ricochet: Anonymous Messaging You Can Trust

Ricochet is an experimental peer-to-peer instant messaging system built on the Tor Network. It protects your identity, contact list, and communications without relying on any central servers or operators. Your login is your hidden service address, and contacts connect directly to you via Tor. This makes it extremely difficult to trace your identity. Available for Windows, OS X, and Linux, Ricochet is open-source and user-friendly, but users should carefully assess their risks.

Read more

Pinterest Improves Embedding-Based Retrieval for Homefeed Recommendations

2025-02-14
Pinterest Improves Embedding-Based Retrieval for Homefeed Recommendations

Pinterest's engineering team significantly improved its embedding-based retrieval system for personalized and diverse content recommendations on the Homefeed. They achieved this through advanced feature crossing techniques (MaskNet and DHEN frameworks), pre-trained ID embeddings, and a revamped serving corpus with time-decayed summation. Furthermore, they explored cutting-edge methods like multi-embedding retrieval and conditional retrieval to cater to diverse user intents, resulting in increased user engagement and saves.

Read more

Trump Admin's Cuts to Decimate Elite CDC Program

2025-02-14
Trump Admin's Cuts to Decimate Elite CDC Program

The Trump administration's push to shrink the federal civil service is set to severely impact the Centers for Disease Control and Prevention's (CDC) Epidemic Intelligence Service (EIS), a world-renowned training program for applied epidemiologists. Many EIS members, crucial in responding to outbreaks like the 2001 anthrax attacks and the 2014-2016 West African Ebola epidemic, face imminent dismissal. This move is alarming public health experts who warn of significantly reduced capacity to handle future crises, both domestically and internationally. The cuts are seen as shortsighted and potentially catastrophic for global health security.

Read more
Tech

The Secret History of Wari Textiles: Looting and the Transformation of Andean Art

2025-02-14
The Secret History of Wari Textiles: Looting and the Transformation of Andean Art

Wari textiles represent some of the most remarkable examples of Andean fabric art. However, their study is hampered by unclear provenance, with many pieces entering global collections through illicit means. The lack of archaeological context makes it difficult to understand their original function and significance. Experts have documented instances of alteration, including cutting, cropping, and restitching, transforming these garments from multi-sensory ensembles worn on the body into flat art objects for Western consumption. This manipulation obscures their original cultural context and purpose.

Read more

Anthropic's Hybrid AI Model: Deep Reasoning Meets Speed

2025-02-14
Anthropic's Hybrid AI Model: Deep Reasoning Meets Speed

Anthropic, an AI startup, is preparing to release its next major AI model, a hybrid approach blending deep reasoning capabilities with fast response times. This new model will reportedly offer a 'sliding scale' for developers to control costs, as deep reasoning is computationally intensive. Early reports suggest it outperforms OpenAI's o3-mini-high model on certain programming tasks and excels in analyzing large codebases and business benchmarks. Anthropic CEO Dario Amodei recently hinted at the model's impending release.

Read more

Haiku January Development Report: Core Improvements and New Features

2025-02-14

The January Haiku development report covers numerous improvements, including a major refactor of the Tracker file manager adding context menus, cut/copy/paste functionality, and live menu updates. Applications saw additions such as new features in the icon editor, touchpad settings, and styled text editor. Driver support was expanded to include Alder Lake chipsets, AMD temperature monitoring, and the Wacom CTH-470. Kernel-level changes focused on extensive memory management, page mapping, and permission check optimizations, boosting system stability and security. Many bugs were fixed, and the build system and documentation were improved.

Read more
Development Development Report

A Stunning Display of Multilingual Support: A Mysterious Code Snippet

2025-02-14
A Stunning Display of Multilingual Support: A Mysterious Code Snippet

This code snippet showcases an impressive multilingual support, containing the names of almost all known languages. This has sparked speculation about the purpose behind the code; is it an art installation, or a fragment of code from a mysterious project? The simple code structure also raises curiosity about how its function is implemented, and where it will be applied in the future.

Read more

NBA's Apple Vision Pro App Gets a 'Tabletop' View: A New Level of Immersive Sports Viewing

2025-02-14
NBA's Apple Vision Pro App Gets a 'Tabletop' View: A New Level of Immersive Sports Viewing

The official NBA Apple Vision Pro app now features 'Tabletop,' a miniature, diorama-style representation of the live game alongside the standard 2D livestream. While a slight delay exists (around half a second), this dual-view approach offers a unique immersive experience. Currently available for select games, the NBA plans to roll it out to all League Pass games next season. A League Pass subscription ($15/month and up) is required. This innovative feature echoes the now-defunct Lapz F1 app for Vision Pro, highlighting the potential of XR and future AR glasses for remote sports viewing. In contrast, Meta Quest offers free 180-degree immersive streams (though 2D, not 3D) of 52 NBA games via Xtadium, but lacks the unique 'Tabletop' perspective.

Read more

Windows 10's Sunset: 40% of Steam Gamers Face an Upgrade Dilemma

2025-02-14
Windows 10's Sunset:  40% of Steam Gamers Face an Upgrade Dilemma

Microsoft will end free software updates, technical support, and security fixes for Windows 10 on October 14, 2025. This leaves a significant portion of users, over 40% on Steam, facing a difficult choice: upgrading to Windows 11. Windows 11's stringent hardware requirements prevent many from upgrading, pushing gamers to explore alternatives like SteamOS. While Valve plans wider SteamOS adoption, a desktop release remains elusive. Continuing to use the unsupported Windows 10 leaves users vulnerable, forcing many to consider upgrading their hardware or operating system.

Read more
Game

lzbench: An Open-Source Benchmark for Compression Codecs

2025-02-14

lzbench is an open-source benchmark tool for evaluating the performance of various compression codecs. It measures compression ratio, compression speed, decompression speed, and round-trip speed. The tool supports multiple codecs and allows users to add new ones, with raw data available for download and further analysis. The FAQ addresses common questions, including adding codecs, calculation methods, memory usage, multi-threading, chart scaling, and customization options, making it a valuable resource for developers and researchers.

Read more
Development compression codec

German Navy Ships Sabotaged, Raising Concerns About Russia

2025-02-14
German Navy Ships Sabotaged, Raising Concerns About Russia

Germany's Inspector of the Navy announced Tuesday that multiple German warships were sabotaged. This follows a report by Süddeutsche Zeitung detailing metal shavings found in the engine system of a new corvette. While not explicitly accusing any party, the naval chief warned of a growing threat from Russia. The incidents follow a string of suspicious fires and explosions at German ammunition facilities and factories, raising concerns about potential Russian involvement and the escalating threat to German and NATO security. Investigations are ongoing, but the sabotage points to a potential deliberate act of aggression.

Read more

Sea Turtles' Secret Navigation: It's All in the Dance

2025-02-14
Sea Turtles' Secret Navigation: It's All in the Dance

Scientists have discovered that sea turtles use Earth's magnetic field for navigation, expressing memories of food locations through a unique "dancing" behavior. Researchers trained turtles to associate specific magnetic fields with food, and the turtles responded by excitedly "dancing" when they sensed the familiar field. Published in Nature, this study reveals that turtles possess two distinct magnetoreception mechanisms: a magnetic compass and a magnetic map, suggesting these mechanisms may have evolved separately. This provides crucial insights into understanding animal magnetoreception.

Read more

Typst vs. TeX: A Comparison of Layout Models and a Look Ahead

2025-02-14

This article explores the differences in layout models between the typesetting engines Typst and TeX. TeX, based on boxes and glue, is flexible but lacks awareness of precise positions; Typst uses a region model, allowing elements to react to their position but sacrificing some flexibility. The author analyzes the strengths and weaknesses of both models and points out that Typst, by introducing a re-layout mechanism, is expected to balance flexibility and optimization, addressing current shortcomings in handling complex layouts (such as wrap-around images and pageable tables).

Read more
Development typesetting engine

Building a Robust Evaluation Framework for RAG Systems

2025-02-14
Building a Robust Evaluation Framework for RAG Systems

Qodo built a Retrieval Augmented Generation (RAG)-based AI coding assistant and developed a robust evaluation framework to ensure accuracy and comprehensiveness. Challenges included verifying the correctness of RAG outputs derived from large, private datasets. The framework evaluates the final retrieved documents and the final generated output, focusing on 'answer correctness' and 'retrieval accuracy'. To address the challenges of natural language outputs, they employed an 'LLM-as-judge' approach and built a ground truth dataset with real questions, answers, and context. For efficiency, they leveraged LLMs to assist in dataset construction and used LLMs and RAGAS to evaluate answer correctness. Ultimately, they built their own LLM judge and combined it with RAGAS for improved reliability, integrating it into their workflow with regression testing, dramatically reducing the effort to verify code changes' impact on quality.

Read more
Development LLM Evaluation

arXivLabs: Experimenting with Community Collaboration

2025-02-14
arXivLabs: Experimenting with Community Collaboration

arXivLabs is a framework enabling collaborators to develop and share new arXiv features directly on the website. Individuals and organizations involved uphold arXiv's values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners who share them. Have an idea to enhance the arXiv community? Learn more about arXivLabs.

Read more
Development

Asahi Linux Lead Resigns Amidst Burnout and Community Conflict

2025-02-14
Asahi Linux Lead Resigns Amidst Burnout and Community Conflict

Hector Martin, project lead of Asahi Linux, resigned due to developer burnout, demanding users, and Linus Torvalds' handling of Rust integration into the Linux kernel. Martin criticized Torvalds' lack of support and accused the Linux community of hypocrisy and malicious attacks. He cited Torvalds' poor leadership in handling Rust integration, leading to abuse of power by maintainers. This highlights the growing issue of developer burnout and community conflict in open source, and the need for sustainable funding for open source projects.

Read more
Development developer burnout

Google Leverages Machine Learning for Age Estimation to Enhance Child Online Safety

2025-02-12
Google Leverages Machine Learning for Age Estimation to Enhance Child Online Safety

Google is testing a machine learning model in the US to better determine if users are under 18, enabling more age-appropriate experiences. The model uses data like website visits and YouTube viewing habits. Suspected underage users will have settings adjusted and be offered age verification options (selfie, credit card, or ID). This responds to growing US concerns over online child safety, aligning with legislation like KOSA. Enhanced safety features include SafeSearch and restricted YouTube content. Further parental controls are also being rolled out, including limiting calls/messages during school hours, managing contacts via Family Link, and managing payment cards in Google Wallet.

Read more

LLM Plugin for Semantic Line Sorting

2025-02-12
LLM Plugin for Semantic Line Sorting

This command-line plugin leverages a Large Language Model (LLM) to semantically sort lines of text. Similar to the GNU `sort` command, but instead of alphabetical order, it ranks lines based on semantic relevance to a given query. Input can be from files or standard input. Users can customize the sorting method, output limit, model, and prompt template. A default prompt is provided, but customization is supported. Installation is straightforward: clone the code, create a virtual environment, and install dependencies.

Read more

Musk's DOGE Website: Anyone Can Edit the Database!

2025-02-14
Musk's DOGE Website: Anyone Can Edit the Database!

Elon Musk's Department of Government Efficiency (DOGE) website has revealed a major security flaw. Reports indicate the site's database is open for anyone to edit, allowing developers to easily alter website content, leaving messages like, "This is a joke of a .gov site." The site appears to run on Cloudflare Pages, not government servers, raising serious concerns about database security. This incident follows Musk's claim of DOGE's commitment to transparency, highlighting his team's negligence of security measures during website construction. Federal employees are worried about sensitive data leaks, particularly after previous criticism of DOGE's use of personal emails, exposing further chaos and inadequacy in security management.

Read more

Jooki's Ghost in the Machine: Exploiting Vulnerabilities in a Dead Audio Player

2025-02-14

Following the bankruptcy of its manufacturer, many Jooki kid-friendly audio players became unusable. This post details the reverse engineering of the Jooki firmware, revealing multiple vulnerabilities, including a backdoor and remote code execution. The author analyzes the firmware, filesystem, Mender OTA update mechanism, and HTTP server interface, demonstrating how to gain control of the device. Methods for exploiting these vulnerabilities to achieve remote code execution by modifying configuration files, using flags, and leveraging the OTA update mechanism are shown. A call to the creator to open-source their work is included.

Read more
Hardware

State Capture: When Private Interests Hijack the State

2025-02-14

This article delves into the phenomenon of 'state capture,' where private interests significantly influence or even control a state's decision-making processes. Starting with the World Bank's early definition applied to Central Asian countries, it analyzes the systemic nature of this corruption, extending beyond selective enforcement of existing laws to manipulation of the lawmaking process itself. Case studies from Bulgaria, Romania, South Africa, and Kenya illustrate the diverse manifestations of state capture and the potential role of external actors like Russia. The South African 'Gupta family' case is detailed as a prime example, showcasing how powerful elites infiltrate government institutions, resulting in massive economic losses and societal harm.

Read more

Browser Extension Fights Trans Erasure

2025-02-14
Browser Extension Fights Trans Erasure

A browser extension called "Marsha P Johnson" combats the erasure of transgender people by replacing "LGB" with "LGBTQ+🧱." The creator highlights the US government's active removal of trans mentions from government websites, including the removal of "TQ+" from LGBTQ+ on the Stonewall National Monument site. This blatant erasure is actively countered by the extension, allowing users to see and protest the censorship.

Read more
Misc

Western Digital Bets Big on HAMR for 100TB HDDs by 2030

2025-02-14
Western Digital Bets Big on HAMR for 100TB HDDs by 2030

Western Digital announced its roadmap to adopt Heat-Assisted Magnetic Recording (HAMR) technology for its HDDs, starting late 2026, aiming for 80TB-100TB drives by 2030. This marks a shift away from their previously championed MAMR technology. Initial HAMR drives, with 36TB (CMR) and 44TB (UltraSMR) capacities, will launch in 2026, with mass production slated for the first half of 2027. Two hyperscalers are already testing these drives. This breakthrough promises to more than double hard drive storage capacity within the next few years.

Read more

Zed Editor Launches Edit Prediction Powered by Open-Source Model Zeta

2025-02-14
Zed Editor Launches Edit Prediction Powered by Open-Source Model Zeta

Zed editor has released an exciting new feature: edit prediction. Powered by a new open-source model called Zeta, it predicts your next edit, allowing you to apply it with a simple tab press. Zeta, derived from Qwen2.5-Coder-7B, leverages supervised fine-tuning and direct preference optimization for accuracy and efficiency. To address latency challenges, Zed employed techniques like speculative decoding and partnered with Baseten for optimized model deployment. Currently in public beta, users can try Zeta for free with a GitHub account. Its open-source nature allows community contributions to improve the model.

Read more
Development
2 4 5 6 7 8 9 149 150