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

Category: Development

LLM Agents: The New DX Standard for API Development

2025-05-20
LLM Agents: The New DX Standard for API Development

LLM-powered agents are becoming tireless junior developers. They read API docs, issue requests, parse errors, and retry until success. However, API developer experience (DX) is crucial. If an agent stalls due to poor documentation or unclear error messages, human developers will likely hit the same roadblocks. Improving API documentation, providing clear and detailed error messages, and ensuring consistency significantly enhances DX and makes agents more efficient. This benefits human developers and allows agents to act as automated testers, catching issues early.

(stytch.com)
Development API Development

Jwno: A Highly Customizable Tiling Window Manager for Windows

2025-05-20

Jwno is a highly customizable tiling window manager for Windows 10/11, built with the Janet programming language. It features a unique "magical parentheses" power for precise window control. While still under development (some documentation is incomplete), it offers installation guides, interactive tutorials, and a reference index. Screenshots showcase its use with applications like Emacs and Sonic Pi.

(agent-kilo.github.io)
Development

RepoRoulette: Randomly Sample GitHub Repositories

2025-05-20
RepoRoulette: Randomly Sample GitHub Repositories

RepoRoulette is a powerful tool for randomly sampling GitHub repositories, offering three distinct methods: ID sampling, temporal sampling, and BigQuery sampling. ID sampling directly selects random IDs from GitHub's ID space, offering speed but with low hit rates. Temporal sampling chooses repositories updated within a specified time range, allowing filtering by stars, languages, etc. BigQuery sampling leverages Google BigQuery's public GitHub dataset, providing powerful capabilities but requiring a GCP account and billing. RepoRoulette is suitable for academic research, learning resources, data science, trend analysis, and security research.

(github.com)
Development Data Sampling

Astra: Blazing-Fast JS-to-Exe Compiler

2025-05-20
Astra: Blazing-Fast JS-to-Exe Compiler

Astra is a fast, reliable, and easy-to-use JavaScript to executable (.exe) compiler available on npm. Boasting an aesthetically pleasing CLI built with signale, inquirer, and chalk, Astra offers a superior user experience. Powered by esbuild, it boasts incredibly fast compilation speeds. It supports compiling ESM-based applications (with workarounds for Node.js SEA limitations) and generates a single executable file containing all dependencies. Customize metadata (icon, name, version, etc.) and benefit from built-in compression for smaller executables. Install globally via npm, yarn, or pnpm, or locally for individual projects.

(github.com)
Development

llm-d: A Kubernetes-Native Distributed LLM Inference Framework

2025-05-20

llm-d is a Kubernetes-native, high-performance distributed Large Language Model (LLM) inference framework offering a streamlined path to serving LLMs at scale. It boasts the fastest time-to-value and competitive performance per dollar for most models across most hardware accelerators. Utilizing cutting-edge distributed inference optimizations like KV-cache aware routing and disaggregated serving, co-designed and integrated with Kubernetes operational tooling in Inference Gateway (IGW), llm-d enables users to operationalize generative AI deployments with a modular, high-performance, end-to-end serving solution. Unlike traditional scaling approaches, llm-d is optimized for the unique characteristics of LLM inference, such as slow, non-uniform, expensive requests, achieving superior performance. Through cache-aware routing, task disaggregation, and adaptive scaling, llm-d significantly improves throughput and efficiency, reduces latency, and supports diverse Quality of Service requirements.

(llm-d.ai)
Development LLM Inference

Running OCaml on a TI-84+ CE Calculator

2025-05-20

This post details the author's journey in compiling an OCaml program to run on a TI-84+ CE calculator. Leveraging Js_of_ocaml, a tool typically used to compile OCaml to JavaScript, the author cleverly repurposed it to generate C code instead. Due to the TI-84+ CE's resource constraints, a simple garbage collector was implemented, along with necessary C functions for interacting with the calculator's hardware. The author successfully ran a simple OCaml program, demonstrating the feasibility of their approach.

(farlow.dev)
Development

JavaFactory: Predictable AI-Powered Java Code Generation

2025-05-20
JavaFactory: Predictable AI-Powered Java Code Generation

JavaFactory is a tool leveraging LLMs to automatically generate repetitive Java code, offering more predictable and stable results than traditional AI code generators. It relies on two core components: Pattern Definition, where units of work (e.g., test generation, implementation generation) are defined in natural language; and Annotation-Based Reference Collection, explicitly specifying required classes using annotations. These defined patterns are reusable for generating various code types (implementations, tests, fixtures). A demo showcases generating 400 lines of code with all tests passing in just 20 seconds. Ideal for developers in repetitive, structured environments (e.g., layered architectures), JavaFactory automates repetitive tasks like dao-repository generation, allowing developers to focus on core logic.

(github.com)
Development

Deno's Future: Not Fading, but Ascending

2025-05-20
Deno's Future: Not Fading, but Ascending

Addressing recent criticism surrounding Deno, Deno Deploy, KV, and Fresh, the Deno team asserts that Deno's momentum is strong, with user numbers doubling. Regarding the reduction in Deno Deploy regions, the team explains this is an optimization for cost and performance, evolving the platform into a full-stack application hosting platform. While Deno KV won't be the central solution for all state management, stronger tools are in development. Fresh 2 is also imminent, bringing significant improvements. The Deno team emphasizes their commitment to building a complete JavaScript platform, focusing on a cohesive system rather than feature parity, and actively participating in improving and growing the JavaScript ecosystem.

(deno.com)
Development

Building a Blog Search Engine from Scratch with Word2Vec

2025-05-20
Building a Blog Search Engine from Scratch with Word2Vec

The authors built a blog search engine from scratch using Python and Word2Vec embeddings. Posts and search queries are embedded into a 300-dimensional vector space, and cosine similarity is used to rank results. To make it web-friendly, the Word2Vec model is split into an index and vectors, with HTTP Range requests used to download only necessary data, reducing web load significantly. An evaluation metric is designed to assess the search engine's accuracy, and future improvements, such as using TF-IDF to reduce noise, are discussed.

(bernsteinbear.com)
Development

Revolutionizing Transparency Ordering in 3D Rendering with Precomputation

2025-05-20

Traditional 3D rendering relies on the CPU to sort translucent objects based on camera position, resulting in inefficient rendering. This post introduces a novel method for precomputing the rendering order of translucent objects. By analyzing the spatial relationship between the faces and the camera position, the algorithm pre-determines the rendering order of some faces, thereby reducing the number of runtime sorting operations. While having a time complexity of O(n²), the precomputation is performed only once, making it suitable for scenes where translucent objects don't move significantly. The algorithm works by splitting each face into two sides and leveraging their spatial relationships to determine the drawing order. In certain cases, it falls back to CPU sorting. This method significantly addresses the challenges of transparency sorting in 3D rendering and improves rendering efficiency.

(jacobdoescode.com)
Development transparency sorting

Cheap and Effective Language Translation Quality Benchmark

2025-05-20
Cheap and Effective Language Translation Quality Benchmark

A developer attempted to build a more scientifically rigorous language translation quality benchmark using pairwise evaluations and a Bradley-Terry model. Initial attempts failed due to high costs, with each experiment requiring hundreds or even thousands of dollars. A compromise system was devised, combining the old scoring system with pairwise evaluations. By iteratively processing sentences, using multiple translation evaluation systems to score, and combining statistical analysis, costs were drastically reduced, yielding reliable results with good p-values. While sacrificing some rigor in blinding, the new system significantly improved efficiency, completing a German test for ~$6.

(nuenki.app)
Development

Biff: A Customizable Full-Stack Clojure Web Framework

2025-05-20
Biff: A Customizable Full-Stack Clojure Web Framework

Biff is a novel Clojure full-stack web framework that curates libraries and tools from across the ecosystem into one polished whole. It features immutable database support, schema enforcement with Malli, and uses hyperscript for rich, interactive UIs without leaving the backend. Passwordless email-based authentication (magic links and one-time passcodes) is included. Deploy via Ubuntu VPS provisioning or a Dockerized Uberjar. Changes are evaluated on file save, and a production REPL allows for live development. Biff boasts strong defaults but is designed to be easily modified as your needs evolve.

(biffweb.com)
Development Full-Stack

Ann: A Decentralized Annotation Server for Empowering Applications

2025-05-20

Ann is a minimal ActivityPub-based decentralized social media built around Web Annotations, essentially comments, recommendations, or likes on any content. It's not a standalone webpage, but a server designed for integration with other applications. Imagine comment sections in Gemini browsers, private research paper annotation systems, article recommendation feeds, browser plugins for adding and viewing comments across the web, or even AI training datasets. Ann's vision is a web independent of JavaScript and trackers, empowering applications with annotation capabilities, giving users choice, privacy, and control over their content consumption.

(mccd.space)
Development web annotation

Memory Consistency Models: A Race Against Time in Multicore Systems

2025-05-20

This tutorial dives into the complexities of memory consistency models in computer science, a particularly challenging problem in multicore systems. It explains how multiple threads accessing shared memory can lead to inconsistencies and explores various approaches to ensuring consistency, including sequential consistency, cache coherence, and relaxed models like TSO. The article uses diagrams and examples to illustrate the challenges and highlights the importance of data race avoidance and synchronization primitives like barriers. The key takeaway? Rely on synchronization libraries to handle low-level complexities and avoid the pitfalls of inconsistent memory.

(jamesbornholt.com)
Development memory consistency multicore programming

AI-Powered Even/Odd Number Checker: The is-even-ai Package

2025-05-20
AI-Powered Even/Odd Number Checker: The is-even-ai Package

The `is-even-ai` npm package leverages OpenAI's GPT-3.5-turbo model to determine if a number is even or odd, along with other numerical comparison functionalities. Developers can easily integrate these features using simple API calls and customize the model and parameters. Inspired by a similar npm package and a tweet, this project showcases how to incorporate AI into a product.

(www.npmjs.com)
Development npm package

Kilo: A Minimalist Text Editor in Under 1K Lines of Code

2025-05-20
Kilo: A Minimalist Text Editor in Under 1K Lines of Code

Kilo is a tiny text editor written in less than 1000 lines of code (using cloc). It boasts a surprisingly useful feature set for its size, including saving (Ctrl+S), quitting (Ctrl+Q), and string searching (Ctrl+F). Importantly, Kilo uses no external libraries and relies on standard VT100 escape sequences. Created by Salvatore Sanfilippo (antirez), it's designed as a learning resource and a starting point for building more advanced command-line interfaces or editors. The project is open source under the BSD 2-clause license.

(github.com)
Development

HashiCorp's Terraform MCP Server: Automating IaC Development

2025-05-20
HashiCorp's Terraform MCP Server: Automating IaC Development

HashiCorp has released the Terraform MCP Server, a Model Context Protocol (MCP) server that seamlessly integrates with Terraform Registry APIs, empowering advanced automation and interaction capabilities for Infrastructure as Code (IaC) development. This server automates Terraform provider and module discovery, extracts and analyzes data from the Terraform Registry, and provides detailed information on provider resources and data sources. Users can run the server via a Docker container or build the binary directly from source code, integrating it with tools like VS Code or Claude Desktop. A suite of tools is available for querying and retrieving documentation and metadata for providers and modules within the Terraform Registry. Crucially, outputs and recommendations are dynamically generated and should be thoroughly reviewed before implementation to ensure alignment with security best practices and compliance requirements.

(github.com)
Development

Deploying a Moose Application to Production with Docker Compose

2025-05-19

This guide provides a comprehensive walkthrough of deploying a production-ready Moose application on a single server using Docker Compose. It covers installing prerequisites, configuring Docker log limits and non-root access, setting up an optional GitHub Actions runner, and a sample Foo Bar Moose application. The guide delves into securely configuring Clickhouse and optional Redpanda, and a phased Temporal deployment (also optional). Finally, it explains setting up a systemd service for automatic Docker Compose startup, and both automated and manual deployment workflows.

(docs.fiveonefour.com)
Development Moose Deployment Production Setup

Don't Use the Obsolete ISO/IEC 14977:1996 EBNF Specification!

2025-05-19

This essay strongly advises against using the ISO/IEC 14977:1996 EBNF specification due to its numerous flaws. The author details the specification's shortcomings, including its lack of support for Unicode characters, character ranges, and common regular expression syntax, as well as its cumbersome "one or more" notation. The author argues that the specification is difficult to understand, lacks readability, and is out of sync with modern software development practices. In contrast, the W3C's EBNF specification is presented as a more concise, user-friendly, and compatible alternative. The author also points out that blindly following ISO standards isn't always correct; choosing the most suitable tool is paramount, rather than being constrained by outdated standards.

(dwheeler.com)
Development

WireGuard Vanity Key Generator: Streamlined Key Generation

2025-05-19
WireGuard Vanity Key Generator: Streamlined Key Generation

This command-line tool, `wireguard-vanity-keygen`, generates WireGuard public keys matching a specified prefix. It offers multi-core processing, case-sensitive/insensitive search, regex support, and the ability to search multiple prefixes concurrently. The tool displays estimated generation time and probabilities, making it easier to generate memorable and manageable WireGuard keys. A significant improvement over existing solutions, offering a more streamlined and efficient key generation experience.

(github.com)
Development Key Generation Command-line Tool

Hacker News Desktop Client Built with Rust

2025-05-19
Hacker News Desktop Client Built with Rust

A sleek Hacker News desktop application built with Rust and egui is now available! Enjoy a clean, modern interface for browsing top stories from various sections: Hot, New, Show HN, Ask HN, Jobs, and Best. Comments are displayed in a threaded, Reddit-style format with auto-folding and adjustable font size. Offline caching, favorites, and powerful search/filtering capabilities enhance the user experience, providing seamless access to Hacker News anytime, anywhere.

(github.com)
Development Desktop App

Run GitHub Actions Locally with act: Faster Feedback Loops

2025-05-19
Run GitHub Actions Locally with act: Faster Feedback Loops

Tired of committing and pushing every time you tweak your GitHub Actions workflows? `act` lets you run GitHub Actions locally for blazing-fast feedback. It mimics GitHub's environment variables and filesystem, enabling local testing and even replacing Makefiles as a local task runner, eliminating repetitive work. A VS Code extension integrates seamlessly into your workflow. `act` reads your GitHub Actions, uses the Docker API to pull or build images, determines execution paths based on dependencies, and finally runs containers, mirroring the GitHub environment.

(github.com)
Development

Microsoft Unveils Lightweight Command-Line Editor 'Edit'

2025-05-19
Microsoft Unveils Lightweight Command-Line Editor 'Edit'

Microsoft launched its new command-line text editor, Edit, at its Build conference. This open-source, sub-250KB editor aims to provide a lightweight default CLI text editor for 64-bit Windows, addressing the lack of a built-in option and aiming to avoid the infamous "how do I exit vim?" problem. Edit boasts keybindings, find and replace functionality, regular expression support, and more. It will be available through the Windows Insider program in the coming months. Microsoft also rebranded Windows Dev Home to Advanced Windows Settings, integrating additional developer-focused toggles into the main Windows 11 settings.

(www.theverge.com)
Development

Anthropic's Claude Code SDK: Powering AI-Driven Coding Assistants

2025-05-19

Anthropic has released the Claude Code SDK, enabling developers to integrate Claude Code into their applications and build AI-powered coding assistants. The SDK currently supports command-line usage, with TypeScript and Python SDKs coming soon. It offers features like multi-turn conversations, custom system prompts, and MCP configuration for extending functionality via external servers. The SDK provides text, JSON, and streaming JSON output formats, along with best practices for error handling, session management, and rate limiting. A real-world example is the Claude Code GitHub Actions, which automates code review and more.

(docs.anthropic.com)
Development

Multilingual Translation Tool Launches

2025-05-19
Multilingual Translation Tool Launches

A new multilingual translation tool has launched, supporting a wide array of languages including Spanish, French, Indonesian, German, Italian, Swedish, Dutch, Danish, Esperanto, Russian, Brazilian Portuguese, Turkish, Polish, Hungarian, Filipino, Slovenian, Croatian, Estonian, Czech, Latvian, Finnish, Catalan, Romanian, Albanian, Armenian, Macedonian, Greek, Ukrainian, Bulgarian, Vietnamese, Korean, Japanese, Simplified and Traditional Chinese, Thai, Persian, and Arabic. Users can easily add new languages and toggle the translation feature on or off.

(ncase.me)
Development tool

Windows 11 Gets a New Built-in Command-Line Text Editor: Edit

2025-05-19
Windows 11 Gets a New Built-in Command-Line Text Editor: Edit

Microsoft introduces Edit, a new lightweight command-line text editor for 64-bit Windows. This open-source editor, under 250KB, boasts features like mouse support, multiple file opening, find and replace, word wrap, and crucially, a modeless design to avoid the steep learning curve of modal editors like Vim. It'll preview in the Windows Insider Program in the coming months before becoming a standard part of Windows 11.

(devblogs.microsoft.com)
Development command-line editor

Microsoft Open Sources Windows Subsystem for Linux (WSL)

2025-05-19
Microsoft Open Sources Windows Subsystem for Linux (WSL)

At its Build developer conference, Microsoft announced it's open-sourcing the Windows Subsystem for Linux (WSL), enabling developers to seamlessly run Linux distributions within Windows. This move aims to solidify Windows as a premier development environment, offering enhanced Linux compatibility. WSL, having evolved from emulation to the native Linux kernel in WSL 2, now boasts significantly improved performance and compatibility. Open-sourcing allows developers to contribute code, further refining WSL's functionality and performance.

(thenewstack.io)
Development

WSL Goes Open Source!

2025-05-19
WSL Goes Open Source!

Microsoft has announced the open-source release of the Windows Subsystem for Linux (WSL)! After years of development, the code powering WSL is now available on GitHub. This allows the community to download the source code, build WSL, add new features and bug fixes, and actively participate in its development. WSL's architecture comprises command-line executables, the WSL service, Linux init and daemon processes, and file sharing components. This open-source release marks a significant step towards faster iteration and community-driven development, highlighting Microsoft's commitment to the open-source community.

(blogs.windows.com)
Development

Glasskube: Open Source Deployment Tools & Hiring!

2025-05-19
Glasskube: Open Source Deployment Tools & Hiring!

Glasskube, a startup building open-source deployment tools to boost engineer productivity, is hiring! They're looking for someone fascinated by how LLMs, MCP servers, and Agent mode are changing software development, enjoys debugging and exploration, and is comfortable automating deployment and monitoring. Prior experience is less important than personal projects and initiative. Tech stack includes Go, TypeScript, and PostgreSQL. If you want to work in Vienna with ties to the San Francisco startup scene, apply!

(www.ycombinator.com)
Development deployment tools

Weekend Hack: Building a ChatGPT Client for Apple Watch with AI

2025-05-19
Weekend Hack: Building a ChatGPT Client for Apple Watch with AI

This post details the author's experience building a ChatGPT client for Apple Watch in a single weekend using OpenAI's API, SwiftUI, CloudKit, and Swift Data. The author leveraged AI to generate initial code, which provided a surprisingly good starting point despite some limitations in understanding watchOS specifics. The process highlighted the strengths and weaknesses of AI in modern development: while AI can quickly generate functional code, human intervention is crucial for refining the design, handling platform-specific quirks, and optimizing performance. The resulting app, WristGPT, is now available on the App Store, showcasing the potential of AI to accelerate the development process while emphasizing the enduring role of human developers in product creation.

(richarddas.com)
Development
1 2 … 82 83 84 86 88 89 90 … 214 215