Self-Hosting Roadmap
Voice Infrastructure Decision
The primary blocker for self-hosting is the voice infrastructure choice. The hosted version of Magne uses LiveKit for real-time audio/video, but LiveKit’s self-hosting complexity may not be appropriate for all deployment targets.Options Under Evaluation
| Option | Pros | Cons |
|---|---|---|
| LiveKit (self-hosted) | Battle-tested, already integrated | Complex to deploy, resource-intensive |
| Jitsi Meet | Easier self-hosting, established project | Different architecture, requires integration work |
| Alternative SFU | Could find optimal balance | Unknown integration effort |
Decision Criteria
- Deployment simplicity — Can a sysadmin deploy it with Docker Compose?
- Resource efficiency — Reasonable CPU/RAM for small team deployments
- Feature parity — Must support the same voice/video capabilities as hosted
- Maintenance burden — Ongoing updates and security patches
Timeline
This is not a committed timeline — it’s a rough indication of priorities:- Phase 1: Resolve voice infra decision
- Phase 2: Docker Compose manifests for API + DB + voice
- Phase 3: Configuration reference and deployment guide
- Phase 4: Upgrade tooling and migration scripts