Gist 35407c28fd6ba2c2097d1b589630c53f

Gist 20e943b854f7d56bb5c7bf168325aacc

Three users in a networked Hubs room share a virtual space where red laser objects and ring-shaped hubs orbit around avatars. The code uses A-Frame and the Networked-Aframe (NAF) API to synchronize positions and ownership of entities, with `setInterval` driving an animation loop that calls `attachObjAvatars` to reposition objects every 100 milliseconds. It also demonstrates `NAF.utils.takeOwnership` for controlling shared objects and `getElementsFromHash` to select media-loaded assets by their source URL. The scene shows mixed crowns and lasers attached above named staff avatars, with continuous rotation updates on the objects.

81% match

Gist ec675b31a307918b12388d698a79845d

Three avatar-shaped objects in a Hubs room orbit three staff member avatars, their positions and scales updating every 100 milliseconds to follow the users. The animation repeatedly copies each staff avatar’s Three.js position, increments its y-coordinate, and applies a random scale via the A-Frame and Networked-Aframe APIs. Objects are selected by matching a media-loader source hash to a GLB file, and ownership is taken with `NAF.utils.takeOwnership` before transforming the networked element.

79% match

Gist 2da2793cd17e3359327a13d36ea438c6

A networked virtual scene shows media objects continuously following an avatar through space. The code attaches a video element to the avatar “Utopiah” every 100 milliseconds, repositioning the object’s Three.js coordinates to match the avatar’s position while offsetting it upward and rotating it. It uses A-Frame’s `media-loader` and `networked-avatar` components, along with Networked-Aframe (NAF) utilities to take ownership and sync the object’s transform. The animation loops indefinitely, keeping the video attached to the moving avatar in real time.

78% match

Gist 022700cf05d336b0804c94278e541b0f

In this animated Three.js scene, colored rings follow avatars through a networked virtual environment, with each ring positioned roughly 1.8 meters above its assigned avatar. The code uses AFRAME’s scene graph and networked-avatar selectors to track entities, while NAF.utils’ `getNetworkedEntity` and `takeOwnership` manage object control. A 50-millisecond interval repeatedly copies avatar world positions onto the ring objects, approximating y-values for vertical alignment. The visualization demonstrates real-time multi-user object tracking, pulling avatar data from the scene’s media-loader elements matched against a hash ring identifier.

Gist 27b0eae72764e9e50963cdf7592e7849

This animation tracks avatar positions in a Mozilla Hubs virtual room over time. A JavaScript interval samples connected users every second, storing their networked IDs and Three.js position vectors in localStorage. The visualization reconstructs movement trails from these recorded snapshots, showing how participants navigate the shared 3D space. It uses the browser’s `setInterval` and `localStorage` APIs, with the Hubs network avatar attributes providing positional data for each connected client.

Gist b077b41a118451bbcc13bbc94caea681

This example shows a WebVR scene where the user’s viewpoint is repositioned by attaching the active camera-tool entity to an avatar rig. The code selects the last camera-tool element in the A-Frame scene, grabs the avatar rig, and uses Three.js’s `Object3D.attach()` to transfer the camera’s world transform to the rig, effectively making the rig follow the camera’s position and orientation. It demonstrates a simple way to control an avatar’s pose in a virtual environment using A-Frame’s entity-component querying and Three.js’s transform hierarchy.

Gist 7b6be1814abc25f027a85878859ed28a

This example demonstrates a collaborative physics-based game inside a Mozilla Hubs virtual room, where two teams score points by throwing a ball onto targets. It shows real-time networked gameplay with avatars, score tracking, and animated feedback. The code uses A-Frame entities, the networked-aframe (NAF) API for ownership and multiplayer sync, and the Anime.js timeline for star jump-and-scale animations. Assets are loaded from Google Poly via the media-loader component, while target positions are derived from 3D object names in the scene.

Utils functions for Mozilla Hubs scripting

This example demonstrates a collection of utility functions for scripting Mozilla Hubs scenes, showing how to manipulate entities, load media assets, and query the scene graph. The code uses the A-Frame and Networked-AFRAME APIs to interact with the Three.js object hierarchy. It includes functions to get user media elements via DOM selectors, load assets from URLs by creating `a-entity` elements with `media-loader` and `networked` components, and find avatars by display name. It also traverses the Three.js scene graph to match objects by partial names, and queries media elements by URL patterns using `querySelectorAll` and regex matching.