Bun, the fast all-in-one JavaScript runtime, has undergone a complete rewrite of its core from Zig to Rust, promising significant performance and stability improvements.
TypeScript 7.0 is now available, featuring a groundbreaking 10x faster native port that promises to dramatically improve type-checking and tooling performance for JavaScript developers.
Cursor announces Grok 4.5, their most intelligent AI model to date, expanding its capabilities beyond software engineering to address a broader range of applications.
Microsoft Research introduces Flint, an open-source visualization language that enables AI agents to create expressive charts from compact, human-editable specifications.
OpenAI introduces GPT-Live, a new generation of advanced voice models designed to power more natural and fluid human-AI interactions within ChatGPT Voice.
Explore the design and WebGL implementation behind 'Ten Years Away,' an interactive, scroll-driven comic created to celebrate Studio375's tenth anniversary.
GitHub's Aspire team showcases how AI agentic workflows can automate documentation updates across repositories, bridging the gap between code changes and published docs.
GitHub Copilot now enables users to deploy GitHub Pages with custom domains and HTTPS in minutes, eliminating the need for manual DNS record configuration.
Hugging Face announces a new native-speed backend for vLLM transformers, promising significant performance enhancements for large language model inference and deployment.
Learn how to make interactive elements invisible but fully accessible using the `hidden="until-found"` attribute in HTML, combined with a clever CSS trick to reserve space.
OpenAI's new analysis uncovers significant issues in SWE-Bench Pro, a popular coding benchmark, raising concerns about the reliability and accuracy of AI model evaluations.
Deno v2.9.2 introduces new features for Deno Desktop, including React Router autodection and HMR support for Vite and Nuxt, alongside various fixes and enhancements.
Kenton Varda expresses concerns about AI-generated change descriptions, noting they often lack the high-level framing needed for effective code reviews, making them 'worse than useless'.
OpenAI outlines its principles for engaging with government and national security partnerships, focusing on responsible AI use, democratic accountability, and public safety.
