hackmap

a django app for controlling and visualising stuff going on in the hacklab.

development

either use nix-shell, or download:

• pyenv / python 3.11+
• poetry

and run

# if you want the venv to be in the project dir (e.g. for intellisense)
poetry config virtualenvs.in-project true

# create & activate the venv
poetry env use python

# to install dependencies from `pyproject.toml`
poetry install

# to apply migrations
poetry run ./manage.py migrate

# to start the django app
poetry run ./manage.py runserver

# in a separate terminal,
# to start the websocket server
poetry run ./socket_server.py

formatting should be checked with ruff, and types with pyright.

map

the map itself is an embedded svg file in templates/map.html. static/js/map.js connects to a websocket (socket_server.py), and then receives messages of the form:

@dataclass
class WebsocketMessage:
    display: str | None
    type: str | None
    target: str | None
    state: str | None

if display is set, that HTML gets shown in the log on the right. if type, target, and state are set, then all elements with .{type}.{target} will have their class replaced with {type} {target} {state}

socket_server.py just listens on some MQTT topics and broadcasts messages depending on what it sees. it also keeps track of the last state sent for each type+target combo, and resends that when a new client connects.

you need to edit the svg by hand to add things. you can use static/map_drawing.svg to help get the correct x/y.

deployment

deployment is done via docker, see docker-compose.dev.yml and scripts/deploy.sh.

for now, you need push permission to the github container registry. this should probably be re-arranged to make things less difficult.