Skip to content

Features-Track.md

Latest commit

 

History

History
65 lines (51 loc) · 2.4 KB

Features-Track.md

File metadata and controls

65 lines (51 loc) · 2.4 KB

Riverside Clone Features

Version 1: Core Functionality (P2P Recording MVP)

  • User authentication and room creation (host creates session with session code)
  • Peer-to-peer WebRTC connection between host and receiver
  • Record media locally using MediaRecorder API
  • Upload complete recording to backend endpoint after stopping
  • Backend saves to S3 and stores metadata in DB
  • Dashboard shows uploaded recordings per session

Version 2: Enhanced Session Management & Dashboard

  • Create Sessions, Tracks, JoinSession tables in DB
  • Host can see all session-related recordings in dashboard
  • Participants recordings visible to host
  • User must sign up before joining a session (for access control & dashboard)
  • Upload endpoint creates track entries linked to users and sessions
  • Track who joined which session via JoinSession
  • Display sessions and recordings in dashboard

Optional Additions for v2

  • Show upload progress in UI
  • Allow retry/re-upload if upload failed
  • Track recording status (pending, uploaded, failed)
  • Cache blob temporarily in localStorage or IndexedDB

Version 3: Media Management & Async Upload Improvements

  • Upload in chunks in background
  • Display upload and processing status on dashboard
  • Implement retry logic with exponential backoff
  • Improve S3 metadata tagging (who, when, size)
  • Track device info (optional, for support)
  • Add simple waveform preview (placeholder for audio analysis)

Version 4: Move Towards Riverside-Level Infrastructure

  • Transition to SFU (Selective Forwarding Unit) for 1-to-many calls
  • Parallel media stream recording (record raw streams)
  • Backend media merging (FFmpeg workflows)
  • Server-side transcription or integration with external API (Whisper, Deepgram)
  • Post-processing pipeline (normalize audio, trim silence)
  • Timeline-style editor for aligning tracks

Long-Term Riverside-Like Features

  • Multi-track editing and exporting
  • Green-room flow before session starts
  • Cloud-based recording (backup stream via SFU)
  • Real-time speaker detection (for auto-focus)
  • Calendar integration, scheduling
  • Shareable link with viewer access
  • Progressive uploads and live previews

Milestone Progress

  • Version 1 Complete
  • Version 2 Complete
  • Version 3
  • Version 4
  • Version 5