I have been using Colota daily since the closed testing phase and kept running into things that bothered me. The WebView-based maps felt sluggish, switching GPS settings between walking and driving was annoying, and I did not want to record my location while sitting at my desk all day. Version 1.1.0 is the result of fixing all of that.
Native Maps with MapLibre GL
The biggest change is the map engine. In v1.0 the maps were rendered using OpenLayers inside a WebView. Panning had noticeable lag, pinch-to-zoom was not smooth, and memory usage was higher than it should be.
I replaced the entire map stack with MapLibre GL Native via @maplibre/maplibre-react-native. The maps now render on the GPU. Panning and zooming that used to stutter are instant now, even with geofence overlays and accuracy circles drawn on top.
The tile source is OpenFreeMap which provides free vector tiles based on OpenStreetMap data without requiring an API key. This keeps Colota fully FOSS-compatible, which matters for the F-Droid build.
Sorry, the image couldn't be found :(
Dark Mode
With vector tiles and MapLibre I could add a proper dark mode for the map. The app fetches the OpenFreeMap style JSON once, transforms the paint properties to dark colors, and caches the result. No extra network request after the first load.
Sorry, the image couldn't be found :(
The color palette uses a navy/indigo family that fits well with the rest of the dark theme. Water is almost black, buildings are a subtle purple-gray, and text labels use a light gray with a dark halo for readability.
Tracking Profiles
This was the feature I wanted most for my own use. The GPS settings that work well for walking (high frequency, small distance threshold) drain the battery when driving. And the settings that work for driving miss too many points when walking.
Tracking profiles solve this by automatically switching GPS settings based on conditions. You can define profiles that activate when:
- The phone is charging
- Android Auto is connected (car mode)
- Speed is above a threshold
- Speed is below a threshold
Sorry, the image couldn't be found :(
For example, I have a "Driving" profile that activates when Android Auto connects. It sets the GPS interval to 4 seconds with a 20m distance threshold. When I disconnect from the car, it switches back to my default settings (2 second interval, 2m threshold).
The profile system uses priority-based resolution when multiple profiles match. It also has a deactivation delay to prevent rapid toggling when your speed fluctuates around the threshold.
Pause Zones
Sometimes you do not want to record your location at all. I do not need a GPS point every second while sitting at my desk at home or at the office.
Pause zones are geofences that automatically stop location recording when you enter them. You define a center point and a radius on the map, give it a name, and the app handles the rest. When you leave the zone, recording resumes automatically.
Sorry, the image couldn't be found :(
The distance calculation uses the haversine formula. The geofence check runs on every GPS fix inside the foreground service, so it works even when the React Native UI is not active. The zones are also visible on the dashboard map as colored circles with labels.
Speed-Colored Tracks
The location history map got a visual upgrade. Track segments are now colored by speed using a green-to-yellow-to-red gradient. This makes it easy to see at a glance where you were walking, cycling, or driving.
Each point on the track is tappable. A popup shows the exact coordinates, speed, accuracy, altitude, and timestamp. I also added a daily distance counter that shows how far you moved on any given day.
Deep Link Setup
Setting up the app with all the server details (endpoint URL, auth credentials, sync settings) is tedious to do manually. Colota now supports a colota://setup deep link that lets you encode the entire configuration in a base64 payload.
The URL format looks like this:
bashCopy codecolota://setup?config=eyJlbmRwb2ludCI6Imh0dHBzOi8vZXhhbXBsZS5jb20vYXBpL2xvY2F0aW9ucyIsInVzZXJuYW1lIjoidXNlciJ9
The base64 payload decodes to a JSON object with all configuration fields. You can generate a setup link on your server and share it. Scanning or tapping it on the phone configures everything in one step.
Settings and Sync Presets
The settings screen got a cleanup. Sync presets (Instant, Balanced, Power Saver) make it easier to pick the right tradeoff between freshness and battery life without touching individual values.
Sorry, the image couldn't be found :(
Smaller Improvements
- Improved About screen with debug info copy and build variant display
- Better data management with sync progress indicators and database vacuum feedback
- Fixed map panning being blocked during active tracking (the user location overlay was re-rendering every 250ms, removed the animation and memoized the marker coordinates)
- Fixed stale tracking state after Android revokes location permission in the background
- Reliable Android Auto detection using the CarConnection API instead of broadcast hacks
New Icon
The app has a new icon. The old one was a placeholder I threw together in five minutes. The new one follows the Android adaptive icon guidelines and actually looks decent on both light and dark launchers.
Try It Out
The app is available on Google Play, F-Droid (pending review), and as a direct APK download on GitHub. The full source code is AGPL-3.0 licensed.
For setup instructions with different backends (Traccar, Home Assistant, OwnTracks, Dawarich, PhoneTrack) check the documentation. If you run into issues with background tracking being killed by your phone manufacturer, have a look at the battery optimization guide.
If you have been following along from my earlier post about location tracking with OwnTracks and Node.js, Colota is basically the evolution of that setup. The tracking app is now my own, fully open-source, and does not depend on OwnTracks anymore. The server-side stack (PostgreSQL, GeoServer, MapProxy) still works the same way. You just point Colota to your webhook endpoint and it sends the same kind of location payloads.