Feature Friday: Space. The Final Frontier

I sat down this morning wanting to build something with Three.js after having seen a lot of demos of people manipulating 3D objects via the camera. This is another case of forcing myself to train as I didn't have a specific purpose for this app, I just wanted to learn how to build with the tools.

Vague is...vague

I started off by letting the AI choose the topic and we initially built a particle cloud which moves when my hands move. It was both boring and riddled with bugs and I had hard time keeping it working between edits. I got it mostly working, but in the end, because I started vague, it vaguely worked.🙂

The same happens in the dojo, training without focus or concentration can feel like wasted time and is if you want to improve. Play time, can be valuable though, I'll come back to that later.

For now, play time was over as I had experimented enough to understand how to use the tools. To start fresh, I had the AI generate a build plan for the app I had finally decided on which was a solar system simulator.

Build Plan: Gesture-Controlled Solar System
This document outlines the step-by-step process
build a Gesture-Controlled 3D Solar System from scratch.
It combines Three.js for high-fidelity rendering
and MediaPipe for computer vision-based tracking.

1.  Technology Stack
    Runtime: Node.js & npm
    Build Tool: Vite (Vanilla JS template)
    Graphics Engine: three (Three.js)
    Computer Vision:
            @mediapipe/hands: Hand tracking model.
            @mediapipe/camera_utils: Helper for webcam input.
etc..

---

Structure is the Key to Productivity

I started a completely clean session with this new build plan in hand. The difference was night and day.

While I still had to make adjustments along the way as I still hadn't taken the time to plan out exactly what I wanted in great detail, the work felt controlled rather than chaotic.

The Result

Focusing on a specific solution such as a solar system visualization application - kind of like practicing a kata vs random techniques - I was able to gain more understanding of the tools and tech, and had much better outcomes.

That doesn't mean the unstructured time wasn't useful as it helped me find the shape of the solution I wanted. Play helps you be creative, but structure helps you turn that creativity into something meaningful. I use the play as a time to learn and experiment.

You can see the results of this Feature Friday session and play with the motion tracking yourself at the link below.

https://space.euloth.com/

Domo arigato gozaimashita.