14h 17m 59s logged

Devlog #1

I’m making a server where you can add cameras from different protocols, view them from a web dashboard (like MotionEye’s, but better), route each input to an output, record, and have real-world things happen when things happen in your camera.

I’ve worked 14 hours so far, but I was planning to ship Argus to Macondo, which doesn’t require devlogs, so I’m losing 4 hours now

(Don’t ask what the video is in the screenshot; I found some random HLS test feed 💀)

Protocols

What I’ve written so far (Stardance doesn’t support task lists apparently):

Input
- ☑︎RTMP(S)
- ☑︎RTSP(S)
- ☐WebRTC / WHEP
- ☐SRT
- ☑︎HLS
- ☑︎MJPEG
- ☐MPEG-DASH
- ☐WHIP
- ☐HTTP Progressive
- ☐HomeKit
- ☐MSE
- ☐Command (custom)
- ☐Web client
Output
- ☐RTMP(S)
- ☐RTSP(S)
- ☑︎WebRTC / WHEP
- ☐SRT
- ☐HLS
- ☑︎MJPEG
- ☐MPEG-DASH
- ☐WHIP
- ☐HTTP Progressive
- ☐HomeKit
- ☐MSE

Recording

Always
Never
On trigger (with manual button)

Triggers

Triggers are actions that take place when something happens. Triggers are supplied by extensions. Triggers can start/stop recordings or take actions provided by extensions.
Some examples:

Start recording when a specific cat walks into the room
Stop recording when I arrive home
Send me a notification (via a Pushcuts webhook) when somebody walks onto my porch (a zone)

Extensions

People can make extensions that can

Add triggers
Add trigger actions
Add source and output protocols
Add overlays to recordings, dashboard views, and outputs (like a box around a person)
Extensions that come pre-installed include:
Motion detection
- Anywhere
- In a zone
Face/person recognition
- Detection
  - Anywhere
  - In a zone
- Add a name
- Emotion
- Groups of people
- Overlay (box, name, emotion, and certainty)
Email and webhook (trigger actions)
Custom Overlays
- You can add text/images including
  - Timestamps
  - Protocol
  - Custom text
Animals
- Detection
  - Anywhere
  - In a zone
- Species
- Color
- Add a name
- Overlay