Ollama 0.40: MLX Is Now Default on Apple Silicon
Ollama 0.40 makes MLX the default backend on Apple Silicon. Decode speed nearly doubles—here is what changed, what to pull, and what you ...
VS Code 1.141: Enable Agent Sandboxing Before Your Next Session
VS Code 1.141 adds cross-platform Copilot agent sandboxing, a worktree cleanup tool, and session grids. ...
Cloudflare Traces Open Beta: Full Request Visibility
Cloudflare just ended the proxy-layer black box. Traces — the company’s distributed tracing system — ...
llama.cpp 0.6.0: 3x Metal Speedup, Decision Models API
llama.cpp v0.6.0 fixes speculative decoding on Apple Silicon (3.4x faster on M3 Ultra), adds a ...
LineageOS 24.0: Android 17 and WebUSB Flashing
LineageOS 24.0 ships Android 17 to 196+ devices with a new browser-based WebUSB flash tool ...
KMP Drops Swift IDE Support: What Changes in 2026.3
JetBrains removes Swift editing from the KMP plugin in IntelliJ IDEA 2026.3. Learn what was cut, what still works, and how Kotlin Toolchain ...
ArtCraft Ships Rust Office Suite Built with Claude AI
ArtCraft releases WordCraft, GridCraft, and DeckCraft — free Rust reimplementations of Word, Excel, and PowerPoint ...
GitHub Rewrote Copilot in Rust: 800K Lines, AI Did It
GitHub migrated the Copilot runtime from TypeScript to 832K lines of Rust in 14.5 weeks. ...
Mockoon Pro Goes Self-Hosted: What Changes for Your Team
Mockoon replaced its cloud plans with self-hosted Mockoon Pro on October 1, 2026. Here is ...
Atlassian AMP: AI Agents Now Have Accountability
Atlassian Agentic Multiplayer Protocol gives AI agents named identities, real-time presence, and audit trails in ...