This boilerplate contains both a Flutter project and a Flask app: the Flask app acts as the API/backend and the Flutter project will build as the frontend.
- make sure you have flutter installed on your computer
- make sure you have poetry installed on your computer
- run
sh scripts/setup_dev.shto install flutter and poetry dependencies
- run the backend (Flask) with
sh scripts/start_dev_server.sh - run the frontend (Flutter) with
sh scripts/start_dev_client.sh
In order to use the Flutter hot-reload (by entering "r" in the terminal after starting), these two commands must be run in different terminals. If Flutter ever releases an update for hot-reloading on save, these two commands could be combined.
The Dockerfile has everthing you need to build a Docker image and host the
app. The quick version is that the Dockerfile uses an image with poetry already
installed, flutter is installed, the frontend is build and the result is copied
into the Flask server, and finally - the server starts.
I tested the deployment using Railway (https://railway.app/) and it works great. I have not tested this on any other service (like Azure App Service), but because its Dockerized, it should run just the same.