An endless F1 runner, and the arcade it might belong to
App two, and a learning project. Web and mobile from the start, because the point is eventually a room full of them.
A note, not a plan. Parked deliberately.
The thing: a very basic endless runner with an F1 car. Built to learn on rather than to be good — the first one is for finding out what building a game is actually like.
Web and mobile, from the start. Not a port later. That means the web build is the game and the store build wraps it, which is the same architecture question SC-34 raises for Share the Search and has to be argued separately here — a game is not a search box.
Where it points: an arcade on scofti, sitting alongside the jukebox and ScofTV as another room. That is why mobile matters now rather than later; a room of games nobody can play on a phone is half a room.
Rough shape, to be re-argued when it starts
TypeScript with Phaser for 2D or Three.js for 3D, in this stack — same language, same deploy, same Capacitor path already written down. Unity is the better engine and the worse fit: half its work lives in an editor, which makes it a poor thing to build with an assistant that cannot see one.
The tuning is not code. Whether the jump floats, whether speed ramps too early, whether a near miss reads as a near miss — that is played, not reasoned about, and it is Michael's loop rather than something that can be specified up front.
Filing note: the arcade is a room and sits under The Rooms, whose name no longer claims to be only audio and video. The runner itself is an app rather than a room — if the two need separating, the arcade keeps this key and the game gets its own.