Open source news — new project releases, community governance, licensing updates, and open-source AI models including Llama, Mistral, and others.
The Granite 4.1 family consists of dense, decoder-only LLMs (3B, 8B, and 30B parameters) trained from scratch on approximately 15 trillion tokens through a five-phase pre-training pipeline that progressively refines data…
Checkmarx, a security firm, has suffered multiple supply-chain attacks over the past 40 days, including two separate malware distribution incidents via its compromised GitHub account and a subsequent ransomware attack by…
Git 2.54 introduces new features including easier rebasing through the `git history` command, which allows users to reword or split commits without checking out the associated branch first. The update also includes hooks…
A threat actor compromised the open-source package element-data, which has over 1 million monthly downloads, by exploiting a vulnerability in a GitHub action to steal signing keys and account tokens. The attackers publis…
Three scalable web applications—a Document Privacy Explorer, an Image Anonymizer, and a SmartRedact Paste tool—all built using OpenAI's Privacy Filter model and Gradio's Server infrastructure. The Privacy Filter is a 1.5…
The Free Software Legal & Licensing Workshop (LLW) 2026 was held in Berlin's Kreuzberg district, gathering over 100 global legal, compliance, and policy experts to discuss the evolving legal landscape of Free Software. A…
This article describes a Python script that functions as an OpenAI-compatible proxy for the DeepSeek V4 Flash model, designed to optimize API usage through intelligent context compression. The proxy automatically compres…
Here is a factual summary of the article: Fresh 2.3, a web framework for Deno, has been released with over 100 commits from 20 contributors. The update delivers on its "zero JavaScript by default" promise by ensuring pa…
Following the April 2026 update to Ubuntu 24.04 LTS, file thumbnails stopped being generated. The issue can be resolved by relaxing a security restriction, clearing the failed thumbnail cache, and restarting the file man…
NanoClaw is a self-hosted AI assistant built on Anthropic's Claude that runs entirely on a Raspberry Pi. Unlike standard stateless chatbots, it uses a structured memory system that extracts discrete facts into a graph da…
The article announces a new C99 single-header library inspired by @clack/prompts, designed to create beautiful command-line interfaces with zero dependencies, instant compilation, and cross-platform support including bro…
DeepSeek-V4 introduces a new architecture using hybrid attention mechanisms—Compressed Sparse Attention (CSA) and Heavily Compressed Attention (HCA)—to drastically reduce the computational cost and memory requirements of…
Technical guide for developers on integrating Transformers.js into a Chrome extension under Manifest V3 constraints. It details a three-part architecture using a background service worker to host AI models, a side panel …
The article explains that Rust Workers on Cloudflare's platform, which compile Rust to WebAssembly, previously faced reliability issues where panics or aborts could poison the runtime and cause cascading failures across …
The article provides instructions for running Claude Code using a local large language model (LLM) instead of Anthropic's cloud-based models. It recommends downloading specific quantized Qwen3.6 models from the MLX commu…
Here is a 2-3 sentence factual summary of the article: This guide explains how to configure software repositories to work with any AI coding agent without vendor lock-in by organizing agent configurations into three cat…
The AI system Mythos, which combines a large language model with substantial compute power, scaffolding, and autonomy, can rapidly find and patch software vulnerabilities, highlighting that the system's design—not just t…
Here is a factual summary of the article: Bun version 1.3.13 introduces two new flags for `bun test`: `--isolate`, which runs each test file in a fresh global environment within the same process, and `--parallel`, which…
According to a new report from the Free Software Foundation Europe (FSFE), Apple has failed to produce a single concrete solution in response to 56 formal interoperability requests made under the European Union's Digital…
Instructions for using a specific chat template file (`chat_template_gemma_large_fixed.jinja`) with the vLLM inference engine to enable compatibility with the Gemma 4 model and OpenCode. It directs users to download the …