Roadmap

Viora’s roadmap is deliberately short. The app does one thing - private 1:1 calls - and the goal is to do that extremely well and then extend it just far enough to be useful for a small group, without ever becoming a meetings product.

Now

Next

Mesh multi-party (3–5 people)

The big one. A full-mesh topology - every phone holds a PeerConnection to every other phone, no media server - with a hard cap of 5 and a demand-driven video budget so a quiet participant nobody is looking at encodes zero video.

Full design: Multi-party calls.

TURN fallback

Today the ICE config is STUN only. Two peers behind symmetric NAT with no shared LAN can’t connect. A self-hosted coturn (or an embedded relay in the signaling binary) with short-lived credentials would close that gap. This is the single biggest reliability improvement available and doesn’t compromise the privacy story - a TURN relay forwards encrypted SRTP it can’t read.

Dedicated screencast video source

Screen sharing currently reuses the camera video source. Moving it to createVideoSource(isScreencast = true) plus a native-resolution bitrate bump gets sharp text at 60 fps. The change exists on a branch; it needs re-integrating.

Maybe

• Web client parity for 2-person calls. The server/public/ client is a test tool. Bringing it to feature parity for 1:1 would let a phone call a laptop.
• Persistent room links. Right now a room is ephemeral. An opt-in “stable room” (still just a code, still no auth) could make recurring calls easier.
• Call recording - local only, sharer-consented.

Never

• An SFU / MCU. Past a handful of people a mesh is the wrong tool, and so is Viora.
• Accounts, a directory, presence-outside-a-room.
• A hosted SaaS tier.
• Telemetry / analytics of any kind.