Features

• Minimap images are sent automatically from the client
• Automatic zoom level generation
• Live-mode available (you can see online characters moving on the map)
• Ingame markers can also be automatically sent to the client (including quest givers and custom markers)
• Geolocation available for the client
• Metrics and server state can be monitored through the Humio
• Good performance (easily handled ~40 clients sending updates at the same time at the W11 start)
• Cheap maintenance (or even free - can be run inside free-tier t2.micro aws instance)
• Optimal stability (worked nearly the whole W11 (a year) with only few reboots)
• No more manual merges and updates!

Some examples:

Random W11 map part

Live mode

Geolocation

Limitations

This map is centered around certain point and can receive new minimaps only when client can determine offset The typical use-case for this map is:

• At the start of the world your group find a spawn point
• You define this point as (0; 0) for the map using Admin API (check backend readme for more info)
• Everyone restart clients :)
• Now map will receive minimaps and they immediately will be seen in the Web UI

Configuration and build systemd

Detailed configuration options can be found inside backend/frontend folders

Deployment package can be generated using create-distribution.sh script

To config initial values open the script and edit variables at the top

export DOMAIN_NAME="ec2-1-2-3.eu-central-1.compute.amazonaws.com"  # Domain for your server
export ENDPOINT_URL="http://ec2-1-2-3.eu-central-1.compute.amazonaws.com" # Endpoint for the frontend/api
export HUMIO_TOKEN="abcdef" # Humio token if present (see backend readme)
export MAP_USER="TestUser" # Username to login into the map
export MAP_PASSWORD="TestPassword" # Password to login into the map
export MAP_ADMIN_TOKEN="abcdef" # Admin token for map api
export INSTANCE_MAP_USER="mapuser" # Your VM user, which will be use to run map

After invocation of this script you will get hnh-map folder with all required data inside

• hnh-map/public - built frontend
• hnh-map/map-backend.jar - backend server jar
• hnh-map/application.properties - backend server properties
• hnh-map/start.sh - startup script
• hnh-map/nginx.conf - nginx configuration (should replace default nginx.conf)
• hnh-map/hnhmap.service - systemd service for the backend server
• hnh-map/install.sh - install script

Configuring VPS instance

You don't have very specific restriction about OS. I've tested everything with Ubuntu 18.x

Required software:

• Java (tested with OpenJDK 11)
• Nginx (tested with 1.14)
• PNGQuant (tested with 2.12.5)
• vsftpd (optional ftp server to upload distribution)

Installation process (for Ubuntu 18.x)

1. sudo apt-get update
2. sudo apt install pngquant nginx default-jdk vsftpd
3. sudo adduser mapuser - Add new user for map server
4. usermod -aG sudo mapuser - Set as sudouser
5. Configure vsftpd. You can take any guide, for example DC's one
6. Upload hnh-map folder to the remove server (the structure should looks like this: /home/mapuser/hnh-map/install.sh)
7. su - mapuser - Login as mapuser
8. Move into ~/hnh-map/ folder and run ./install.sh. Script should update nginx conf and systemd service

This should be it. Now try to access your instance on port 80 and you should see frontend application

docker deployment

There is a docker-compose file with everything builded && included.

It also supports let's encrypt to ensure secure connection (https).

You will need docker:

curl https://get.docker.com | bash

And docker-compose

curl -L https://github.com/docker/compose/releases/download/1.25.4/docker-compose-`uname -s`-`uname -m` -o /usr/local/bin/docker-compose
chmod +x /usr/local/bin/docker-compose

Fill .env file and make docker-compose up -d

Amazon ec2 instance doesn't support let's encrypt on it's base dns name, you will need other domain (like .tk), or use http.

docker-compose up -d

Client integration

You can check sample integration against Amber Client here: APXEOLOG/amber#1

All client code is around 1k lines, concentrated into a separate files with few integration points (~20 invocations across client codebase) Should easily integrated into any custom client