VC App — WebRTC Learning Project

This is a simple Omegle-like WebRTC learning project. It uses basic WebRTC APIs to establish a peer-to-peer connection between two users and Cloud Firestore as a minimal signaling & matchmaking backend.

Purpose and scope

• Educational: built for learning WebRTC fundamentals (offer/answer, ICE candidates, data channels).
• Signaling: Firestore is used as a lightweight signaling and matchmaking layer.
• Not production-ready: a full Omegle-style product requires additional server-side infrastructure for moderation, monitoring, IP tracking, abuse prevention, and legal compliance — these are intentionally left out for this learning project.

Features

• Simple P2P video between two peers.
• Firestore-based signaling and matchmaking using a connections collection.
• Live local preview on app load.
• Start / Disconnect controls and basic UI placeholders.
• Basic error handling with a modal and automatic cleanup on failures.

Dicecam (Next.js)

cd Next
npm install
npm run dev

Add your Firebase keys to Next/.env.local using NEXT_PUBLIC_ prefix.

Legacy prototype (Vanilla code)

⚠️ Latest changes into the vanilla code were not tested due to the rewrite in Next.

cd Vanilla\ code
npm install
npm run dev

Add your Firebase keys to Vanilla code/.env using VITE_ prefix. Environment variables are read via import.meta.env (Vite). The Firestore collection name used for signaling is connections by default and can be changed with VITE_FIREBASE_DB_COLLECTION.

---

Notes

• Anonymous sign-in must be enabled in Firebase Authentication -> Sign-in method. If it is disabled, the app will receive a 400 from accounts:signUp even if the API key is valid.
• Firestore rules are in Vanilla code/firestore.rules.

Security and privacy

• This project stores ephemeral offers/answers and ICE candidates in Firestore for learning only. Do not use this implementation as-is for production: add proper auth, moderation, rate-limiting and data retention policies.
• The app uses anonymous Firebase Auth. The goal is basic abuse prevention, not strong production security: authenticated users can still be misused if the app is shared publicly.

Minimal abuse prevention

• Enable anonymous authentication in Firebase Authentication.
• Deploy the included firestore.rules file so only signed-in users can read/write the signaling data.
• Keep room and ICE candidate payloads small; the rules cap offer/answer strings and candidate fields to modest sizes.
• Room cleanup is owner-driven: the caller who created the room can delete the room doc plus both ICE candidate collections on disconnect.
• If you later add real moderation or anti-abuse controls, move signaling to a server and add auth, logging, and cleanup policies there.

This project began as a JavaScript-only learning challenge with the goal of understanding WebRTC fundamentals without relying on tutorials. The initial implementation focused on peer connection setup, SDP offer/answer exchange, ICE candidate propagation, and Firestore-based signaling.

After the prototype successfully established peer-to-peer video connections, the project was refactored to TypeScript and received a redesigned user interface, anonymous Firebase authentication, improved error handling, and basic security rules.