Popular：
Virtualization DNS security formal verification reachability analysis compiler errors macro conflict web extension development framework Bitmap Graphics API inconsistencies All Tags

Meta's Internal Emails Reveal Obsession with Beating GPT-4

2025-01-15
Meta's Internal Emails Reveal Obsession with Beating GPT-4

Leaked internal emails reveal Meta executives and researchers were fiercely focused on surpassing OpenAI's GPT-4 while developing Llama 3. Messages show a strong desire to outcompete rivals, even dismissing open-source competitors as insignificant. Their ambition led them to use the LibGen dataset, containing copyrighted works, for training, now resulting in multiple copyright lawsuits. While the released Llama 3 proved competitive with leading closed-source models, even outperforming some, Meta's aggressive tactics highlight the intense competition and risks in the AI race.

Read more
(techcrunch.com)
AI AI race

Forging Passkeys: Exploiting the FIDO2/WebAuthn Attack Surface

2025-06-24

This article delves into the security of FIDO2 passkeys. The author reverse-engineered commercial hardware keys and platform authenticators, building a software-only authenticator that mimics a FIDO2 device without kernel drivers. This allowed forging and replaying passkey signatures for headless logins. The process detailed includes capturing real-world traffic, decoding HID handshakes, verifying attestation data, building a software CTAP2 engine, and exploiting Chrome's built-in virtual authenticator. The author successfully logged in without a real security key, highlighting vulnerabilities and proposing mitigations like mandatory sign-counter enforcement, CDP permission restrictions, and relying-party-side checks to enhance passkey security.

Read more
(www.nullpt.rs)
Tech Passkey Security

Svader: A Svelte Library for GPU-Rendered Components

2024-12-14
Svader: A Svelte Library for GPU-Rendered Components

Svader is a library for creating GPU-rendered Svelte components using WebGL and WebGPU fragment shaders. Developers can write programs in fragment shaders to customize pixel colors and control rendering effects through parameter passing. Supporting Svelte 4 and 5, it offers WebGL and WebGPU rendering modes with built-in parameters like resolution, scale, and time. Svader simplifies GPU rendering with easy-to-use components and provides fallback rendering in environments lacking WebGL or WebGPU support.

Read more
(github.com)
Development

Light-Speed Edge Detection: Energy-Efficient Image Processing Revolution

2025-01-30
Light-Speed Edge Detection: Energy-Efficient Image Processing Revolution

Physicists at the University of Amsterdam have developed a novel method for image edge detection using optical analog computing. This technique boasts exceptional speed and energy efficiency, employing a simple stack of thin films to detect edges as small as 1 micrometer. Compatible with various light sources, this breakthrough promises advancements in high-resolution microscopy, biological sample analysis, and even autonomous vehicles, revolutionizing energy efficiency and computational speed.

Read more
(phys.org)
Tech optical computing edge detection

Preserves: A More Expressive Data Language

2025-01-29

Preserves is a data model and serialization format comparable to JSON, XML, and others. It features a syntax-neutral data model and semantics, enabling lossless conversion between various syntaxes. The project provides specifications, tutorials, implementations in multiple languages (Python, Rust, JavaScript, etc.), and tools for developers seeking a more powerful and flexible data representation. It also offers schema and query capabilities for efficient data manipulation. Compared to JSON, Preserves offers richer expressiveness and better scalability.

Read more
(preserves.dev)
Development data language data model

Proxmox Datacenter Manager Alpha Release Announced

2024-12-19
Proxmox Datacenter Manager Alpha Release Announced

Proxmox has released an alpha preview of its Datacenter Manager. This software centralizes management of all nodes and clusters, offering basic features like virtual machine migration without requiring a cluster network. Developed entirely in Rust, from the backend API to the new frontend, it boasts a modern web UI for improved speed and compatibility. This alpha release aims to gather user feedback, test core features, and foster collaboration. While some features are incomplete and bugs are expected, it's already capable of managing thousands of remotes and virtual machines.

Read more
(forum.proxmox.com)
Development Datacenter Management Alpha Release

GPU Performance Tuning: Hitting the Roofline Limits

2025-06-24

This article delves into the performance bottlenecks of GPU architectures, focusing on how memory bandwidth and compute throughput limit application speed. Using the Roofline model, it analyzes memory-bound and compute-bound regimes, detailing strategies to increase arithmetic intensity (AI): operator fusion and tiling. Fusion reduces intermediate memory traffic, while tiling maximizes data reuse through shared memory. The article also covers nuanced topics like shared memory bank conflicts, thread divergence, and quantization for performance gains. The ultimate goal is to push kernel operation points towards the compute throughput ceiling in the Roofline model.

Read more
(damek.github.io)
Hardware Performance Tuning Roofline Model

Command & Conquer: Red Alert Source Code Released!

2025-02-27
Command & Conquer: Red Alert Source Code Released!

The source code for Command & Conquer: Red Alert is now publicly available on GitHub! While the code isn't fully compilable and requires work to replace outdated libraries like DirectX 5 SDK, it's a valuable resource for nostalgic players and developers. This project is for archival purposes only and offers no support; developers are encouraged to fork the repository for modifications and collaboration.

Read more
(github.com)
Game Command & Conquer Red Alert

Small is the New Big: Building for One in the Age of AI

2025-08-17
Small is the New Big: Building for One in the Age of AI

In the era of AI-assisted coding, the cost of building small, personal applications has plummeted. The author shares anecdotes of creating several small utilities: a private Slack workspace for a hundred people, a simple app for sending postcards to his mother, and a small program that calls her to remind her to take medication. These aren't designed for scale, but to meet specific needs for himself and a small circle. The author argues that the real luxury isn't speed or cost, but the freedom to stop, to build something small, useful, and perfectly personal, without the obligation to grow it until it breaks. In a world obsessed with scale, there's quiet satisfaction in leaving 'good enough' alone.

Read more
(derwiki.medium.com)
Development personal applications small is beautiful

Restore Oddly Shaped App Icons in macOS 26 Tahoe

2025-06-15
Restore Oddly Shaped App Icons in macOS 26 Tahoe

macOS 26 Tahoe replaces the unique, oddly shaped app icons in the Dock with iOS-style squarcles, a change many users dislike. This article provides a solution for both users and developers to restore custom icon shapes. Users can replace the .icns file within the application package; developers can use NSApplication.shared.dockTile.contentView to change the icon at runtime. Get your Dock back to its former glory!

Read more
(simonbs.dev)
Development App Icons

Denmark's HPV Vaccination Program Nearly Eradicates Two Major Cancer-Causing Strains

2025-09-17
Denmark's HPV Vaccination Program Nearly Eradicates Two Major Cancer-Causing Strains

Research published in Eurosurveillance shows that Denmark has virtually eliminated infections with the two most prevalent cancer-causing strains of human papillomavirus (HPV) since the vaccine's introduction in 2008. Analysis of cervical cell samples from Danish women aged 22-30 (2017-2024) revealed that HPV16/18 infection rates in vaccinated women plummeted from 15-17% to less than 1%. This demonstrates not only individual protection but also herd immunity, reducing overall HPV16/18 circulation. However, roughly one-third of screened women still had infections with high-risk HPV types not covered by the initial vaccine. This is expected to decrease as women vaccinated with the newer nine-valent vaccine reach screening age, potentially prompting a review of cervical cancer screening guidelines.

Read more
(www.gavi.org)
Tech HPV vaccine cervical cancer herd immunity

Booting Erlang on 16MB: A GRiSP Nano Achievement

2025-07-22
Booting Erlang on 16MB: A GRiSP Nano Achievement

The GRiSP Nano team achieved a significant milestone by successfully booting an Erlang virtual machine on a 16MB STM32U5 microcontroller. Initially aiming for 32MB, a CPU erratum forced a reduction. Through a series of optimizations, including removing the crypto library, aggressive compile/link flags, stripping BEAM files, RTEMS system tweaks, and allocator surgery, they overcame memory constraints. Disabling Unicode temporarily allowed them to reach the Erlang shell prompt. Future plans involve relocating code to internal RAM/Flash, shipping lightweight kernel/stdlib variants, adding energy-aware boot logic, and developing a Unicode-light build.

Read more
(www.grisp.org)
Development

Privacy-Focused Orion Browser Coming to Linux

2025-03-08
Privacy-Focused Orion Browser Coming to Linux

Kagi, the company behind the paid, privacy-focused search engine, announced that its WebKit-based Orion browser is coming to Linux. Orion, known for its speed, low memory usage, and privacy features, is currently available on macOS and iOS and supports Chrome and Firefox extensions. While currently closed-source, Kagi is gradually open-sourcing components and aims for feature parity with the macOS version on Linux by next year. This is good news for Linux users, offering them another powerful browser choice.

Read more
(www.omgubuntu.co.uk)
Development

Pushing the Limits: A New Measurement of Superheavy Nuclei Half-Life

2025-02-01
Pushing the Limits: A New Measurement of Superheavy Nuclei Half-Life

Researchers have pushed the limit of known half-lives of superheavy nuclei by two orders of magnitude by measuring the half-life of a neutron-deficient rutherfordium isotope. The extremely short half-life was measured by exploiting the longer half-life of excited states, providing insights into nuclear fission. The team bombarded a lead target with titanium-50 ions to create rutherfordium-252, measuring its half-life in excited and ground states as 13 microseconds and 60 nanoseconds, respectively. This challenges existing theoretical models and opens avenues for studying heavier superheavy elements.

Read more
(www.chemistryworld.com)
Tech superheavy elements nuclear physics fission

Those White Crystals on Your Cheese: A Delicious Secret

2025-03-31
Those White Crystals on Your Cheese: A Delicious Secret

Confused by white stuff on your cheese? Don't throw it away! This article reveals the secret of those white crystals. They're not mold, but rather calcium lactate, tyrosine, or leucine crystals – signs of a well-aged cheese, adding unique texture and flavor. Learn about the different types, their formation, appearance, and taste. This guide helps you distinguish them and identify high-quality aged cheese. Next time you see white crystals, confidently savor the delicious reward of time and craftsmanship.

Read more
(www.thephcheese.com)
Misc cheese crystals

Chile Air Quality Map: Real-time Monitoring, Protecting Health

2024-12-27

The Chile Air Quality Map is a real-time air quality monitoring platform providing accurate and reliable air pollution information to Chilean citizens. Users can visually see Air Quality Index (AQI) levels for different regions via the map interface and take appropriate precautions based on pollutant concentrations. This platform enhances public environmental awareness and provides data to support government policies on air pollution control, ultimately aiming to protect public health and create cleaner air.

Read more
(calidaddelairecl.vercel.app)
Tech air quality environmental monitoring

ISO 8583: The Secret Language of Credit Cards

2024-12-18
ISO 8583: The Secret Language of Credit Cards

Every time you tap your card or pay online, you're interacting with the ISO 8583 protocol. This 1987 standard defines the format of real-time transaction messages between banking networks. It includes core fields like message type indicators, bitmaps, and data elements, but networks vary in their extensions and serialization, leading to compatibility challenges. This article delves into the complexities of ISO 8583's structure, field encoding, nested message handling, and demonstrates building a robust ISO 8583 parser to handle network variations and error scenarios.

Read more
(increase.com)
Development credit card payments data parsing

Sherlock Project: Hunt Down Social Media Accounts Across 400+ Networks

2024-12-25

The Sherlock Project is a powerful tool allowing users to search for social media accounts across 400+ networks using only a username. It's easy to get started with simple installation and usage instructions, and supports a wide range of sites. Community contributions are welcome, enabling users to add new sites and improve functionality.

Read more
(sherlockproject.xyz)
Development social media search account search Sherlock Project

Git Shallow Clones: Performance Pitfalls and the Depth 2 Optimization

2025-02-12
Git Shallow Clones: Performance Pitfalls and the Depth 2 Optimization

Git shallow clones (`--depth 1`) can significantly impact performance on the first push. This is because shallow clones artificially mark some commits as root commits, preventing the server from using optimizations and requiring the transmission of the entire commit snapshot. Using `--depth 2` deep clones, however, preserves a complete commit history, allowing the server to utilize optimizations even on the first push, reducing data transfer and significantly improving push speed. Subsequent pushes are unaffected.

Read more
(stackoverflow.com)
Development shallow clone

Farewell to Endless Meetings: A New Approach to High-Velocity Software Development

2024-12-15

Tired of endless meetings and lengthy planning? This article introduces a high-efficiency software development method: code-centric, rapid iteration. The author uses baking as an example to illustrate the concept of achieving the optimal solution through rapid experimentation, frequent testing, and continuous improvement. This method emphasizes reducing documentation, expressing ideas directly in code, using mock data and hot-reloading tools to speed up development, and improving code readability through concise code style and naming conventions. The author advocates breaking down projects into independently executable files, minimizing restart time, and using default language tools for debugging. Although this method may seem like a "chaotic lab," it can efficiently complete projects and avoid the redundancy and inefficiency of traditional methods.

Read more
(mihaiolteanu.me)
Development agile development efficient programming

LLMs and the Limits of Automated Code Optimization

2025-01-07
LLMs and the Limits of Automated Code Optimization

David Andersen experimented with using Large Language Models (LLMs) to optimize code that finds the difference between the smallest and largest numbers whose digits sum to 30 in a list of a million random integers. Initial Python and Rust code ran slowly. While the LLM improved parts, such as the digit summing function, it missed a crucial optimization: checking if a number is relevant *before* the expensive digit sum calculation. Manual intervention, involving a faster random number generator, parallelization, and preprocessing, sped up the Rust code by a factor of 55. This highlights LLMs' limitations in code optimization, particularly for complex problems demanding deep algorithmic understanding and parallelization strategies. Human ingenuity remains crucial.

Read more
(wiredream.com)
Development Performance

OpenAI Pleads with Trump: Loosen Copyright Restrictions or the US Loses the AI Race

2025-03-24
OpenAI Pleads with Trump: Loosen Copyright Restrictions or the US Loses the AI Race

OpenAI warns that the US will lose the AI race to China if it can't access copyrighted material for AI training. They're urging the Trump administration to create more lenient "fair use" rules, allowing AI models to utilize copyrighted data for training. OpenAI argues that China's rapid AI advancements, coupled with restrictive US data access for AI models, will result in American defeat. This move has sparked outrage from copyright holders and publishers, who fear unauthorized use of their works for AI training and increased plagiarism. OpenAI counters that using copyrighted data is crucial for developing more powerful AI, vital for US national security and competitiveness.

Read more
(futurism.com)
Tech

BYD's Gigantic Car Carrier Fuels Global EV Ambitions

2025-01-18
BYD's Gigantic Car Carrier Fuels Global EV Ambitions

BYD launched the world's largest car carrier, the BYD Shenzen, capable of transporting 9,200 vehicles. This is BYD's fourth ro-ro ship, following three others already delivering thousands of NEVs to Europe and South America. Following a record 4.25 million NEV sales in 2024, BYD is aggressively expanding globally, challenging established automakers and seeing significant success in markets like Japan and South Korea. The sheer scale of the Shenzen underscores BYD's ambition to dominate the global EV market.

Read more
(electrek.co)
Tech Global Expansion

Arch Linux Added to url.town Directory

2025-08-04
Arch Linux Added to url.town Directory

url.town, a web directory curated by the omg.lol community, has added Arch Linux. Arch Linux is a lightweight, flexible DIY general-purpose GNU/Linux distribution where users only get what they install. The directory also features a wide range of other resources, spanning blogs, games, art, tech news, and much more.

Read more
(url.town)
Development Linux Distribution

Declassified: The DIY Nuclear Weapon – The 'Nth Country Experiment'

2025-01-24

The National Security Archive has released declassified documents detailing the 'Nth Country Experiment,' a secret mid-1960s project at Lawrence Livermore National Laboratory. A small team of physicists, using only publicly available information, designed a functional nuclear weapon in just three years. This 'do-it-yourself' project demonstrated the feasibility of nuclear weapon development with limited resources, highlighting the dangers of nuclear proliferation. The released documents, while heavily redacted, reveal insights into the experiment's methodology and conclusions, sparking renewed discussion about the protection of nuclear weapons design information and the threat of nuclear terrorism.

Read more
(nsarchive.gwu.edu)
Tech nuclear proliferation nuclear weapons

Open Source License Dispute: A Fight for Software Freedom

2025-02-13
Open Source License Dispute: A Fight for Software Freedom

The Software Freedom Conservancy (SFC) filed an amicus brief supporting a downstream licensee's right to remove “further restrictions” under the Affero General Public License version 3 (AGPLv3) Section 7 in the ongoing Neo4j, Inc. v. PureThink, LLC case. The core issue revolves around whether Neo4j's added “Commons Clause” can be removed. SFC argues that downstream licensees have the right to remove such restrictions under AGPLv3 Section 7, paragraph 4, even if imposed by the original licensor. SFC's brief provides detailed legal analysis of AGPLv3 Sections 7 and 10, arguing that the lower court wrongly sided with Neo4j's interpretation, which could fundamentally alter the community's understanding of adding and removing “further restrictions.” The ruling will have significant implications for software freedom and users' rights.

Read more
(sfconservancy.org)
Development legal dispute

Eki Bright: A Case for DIY Train Routing

2025-01-24
Eki Bright: A Case for DIY Train Routing

Eki Bright, a Tokyo train timetable app, champions a unique 'DIY routing' approach. Instead of automated route suggestions, users manually input each train segment, specifying departure and arrival stations. This empowers power users familiar with their routes, offering real-time updates and easy sharing. The author argues for DIY routing's benefits: precise departure time control, accurate transfer timing, and a streamlined UI free from map clutter. Limitations are also discussed, focusing on its suitability for users with route familiarity and highlighting scenarios where automated routing might be preferable.

Read more
(twocentstudios.com)
Development iOS app route planning train timetable

The Rise of AI Slop: How to Fight Back and Profit

2025-01-26
The Rise of AI Slop: How to Fight Back and Profit

Blogger Ben Congdon observes the proliferation of low-quality AI-generated content, which he terms "AI slop," across the internet. While seemingly convincing at first glance, closer inspection reveals its formulaic nature and lack of originality. He argues against directly copying and pasting AI-generated content, suggesting creators should use AI tools for assistance but meticulously edit and maintain a unique personal voice. He further proposes that creating high-quality content and building a personal brand are key to remaining competitive in the age of AI, and that influencing AI training datasets can even shape the future direction of AI.

Read more
(benjamincongdon.me)
AI

Concurrent Cycle Collection: Garbage-Collected Smart Pointers in Rust for Scheme

2024-12-13

This article details the implementation of a concurrent cycle collector in Rust for garbage-collected smart pointers (Gc) within a Scheme interpreter. Gc functions similarly to Arc>, supporting interior mutability, cloning, and sending across threads. The article thoroughly explains the implementation of Gc, including thread-safe interior mutability using semaphores and read/write locks, and the implementation details of concurrent cycle collection based on the Bacon and Rajan algorithm. This includes the Trace trait, cycle detection, and mechanisms for handling concurrent modifications.

Read more
(maplant.com)
Development Garbage Collection Concurrency
1 2 … 511 512 513 515 517 518 519 … 596 597