- 15 Jan 2026 - 2025 in review
- 12 Jan 2025 - 2024 in review
- 31 Dec 2023 - 2023 in review
- 31 Dec 2022 - 2022 in review
- 14 Feb 2022 - Chengyu Wordle Is Not a Love Story
More in the archives.
- 30 Jun 2024 - 🎂 2 years of (f)unemployment - 🥫 #34
- 31 May 2024 - 🌌 Aurora - 🥫 #33
- 30 Apr 2024 - 🌸 Sakura - 🥫 #32
- 31 Mar 2024 - UV indexing - 🥫 #31
- 29 Feb 2024 - 🕴️ Leap day - 🥫 #30
Subscribe to my newsletter!
- pwa-today/pwa-check - Automated PWA health check tool to make sure your PWA looks, feels, and behaves like a real app.
- lucas-martinic/polyfork-catalog - Free low-poly 3D models for three.js and the web: 2,700+ GLB models and 26 themed kits. The public index of polyfork.dev.
- jeantimex/tokyo - A real-time 3D model of Tokyo in the browser, compiled from public data with procedural detail on top. Fly over the city, change the time of day, and watch the traffic, the trains and the birds.
- nouverse/nouride - Lightweight, zero-bloat multi-agent AI engine in a single daemon. Homelab-friendly and built for mini devices (Raspberry Pi, Orange Pi, Geekom, NUC, LXC).
- ppmpreetham/BUI - Blender's GPU Accelerated UI, Decoupled from it's original codebase. Supports Rust, Python, JS, React, C & C++. Twice as good as GPUI
- birkskyum/manby - Builds vector tilesets from geographic features, like tippecanoe, several times faster
- c0nsumer/maplibre-gl-lanes - Draws routes that share paths as ordered parallel lanes in MapLibre GL JS: a transit-map look for any route network, with crossing minimization, smooth junctions, and constant pixel spacing at every zoom.
- chunkithwang/design-playful-app-icons
- longbridge/gpui-fast - GPUI extracted from the Zed monorepo as a standalone workspace, with the layout engine reworked to keep what did not change between frames
- yugalone/collab-sans
- facebookresearch/HighResCanopyHeight - This repository provides inference code to compute canopy height maps from aerial images, as described in the paper "Very high resolution canopy height maps from RGB imagery using self-supervised vision transformer and convolutional decoder trained on Aerial Lidar".
- Tencent/WeVisOne-VL - WeVisOne-VL is a powerful VLM by the WeChat Vision Team at Tencent, unifying object-level and pixel-level perception from points and boxes to masks and alpha mattes.
- dy/jz - JS to WASM compiler
- vicinaehq/vicinae - A focused launcher for your desktop - native, fast, extensible
- FareedKhan-dev/kimi-k3-in-c - A 2.78-trillion-parameter Kimi K3 running inference on a single CPU in 8.24 GB of RAM. Portable C99: no BLAS, no framework, no GPU.
See more of my starred repos.