Piko is a side project I've been working on for a while. It's a native macOS app that lets you browse and transfer files from USB-connected Android devices and cameras.
I've just released an update, and Piko is now Apple-signed and notarized, and can be installed via Homebrew - try it out:
brew install --cask morishuz/tap/piko
It's free and open source:
github.com/morishuz/piko
dr.mo retweeted
Video search + dragging objects out of scene
Not sure what dragging things out should actually do yet, maybe bring up some editing interface for that entity across entire video/film?
dr.mo retweeted
Todos los productos de Adobe reimplementados desde cero, gratuitos y de código abierto
→ getartcraft.com/apps
ALT Interfaz de herramientas creativas para edición de imágenes, ilustración, video y desarrollo de bibliotecas.
circle explosion with triangles and realistic physics
using my wasm-based delaunay triangulation algorithm: github.com/morishuz/delaunay…
dr.mo retweeted
🐳 Sanbox now has native Docker support.
Run agents packaged as containers, bring up full-stack apps across multiple containers, or run complex evals with many moving parts.
More details: sanbox.cloud/blog/sanbox-now…
very nice! i had a similar idea. take the source code for gimp, make minimal adjustments and add a usable modern gui to it.
dr.mo retweeted
Jev + OpenCode + Browser Harness + Sanbox = 🔥
Watch how OpenCode + browser harness visit a tetris game website, read the control instructions, use gpt-6 to build a game controller for Jev model, and then hand off to play the game.
This is all a single shot prompt with no special skills or tools.
Jev model chooses from predefined outputs. In this setup keyboard based gameplay actions were not included, so GPT-6 builds a controller on the fly that gives Jev possible moves to choose from and turns those choices into keyboard inputs.
This project started partly as structured procrastination 🤣
I was wondering how far you could actually get “vibe coding” something genuinely difficult with AI and how easily you could add improvements over existing products. So I decided to try building a tiny SQLite-like database.
How hard could it be?
Turns out: quite hard! AI makes writing code dramatically faster, but testing, debugging and architecture decisions don't disappear. And years of fixing edge cases and small optimizations aren't easy to vibe code.
But something came of it, this experiment turned into CoreSQL: a working, but still very experimental, embedded database in C++20, with transactions, indexes, crash recovery and SQL. The novelty: it's built around a small core and an extension system that even its basic data types use.
Now open source:
github.com/morishuz/CoreSQL
🤖 Made with AI
I built Delaunay32: extremely fast delaunay triangulation with polygon holes, and constraints
Serious speed now in your browser: ⚡Try the interactive demo 👇
morishuz.github.io/delaunay3…
C++: github.com/morishuz/delaunay…
