plottter-daemon

A small networked plot server for AxiDraw-compatible pen plotters (built for the Uunatek iDraw H SE A2). It receives plot jobs (an SVG + settings) over HTTP and drives the plotter with pyaxidraw, so the machine you design on is free during long plots. It is the device side of the Plottter app's remote-plotter feature — Plottter sends jobs; this daemon plots them.

• Stdlib-only HTTP server — no FastAPI/Flask. The only thing you install on the Pi is pyaxidraw.
• Plots are detached — a job runs inside this long-lived service, so closing the laptop or dropping off the network does not stop the plot.
• One job at a time, with pause / resume / stop and live progress.
• The daemon never changes your design — it plots the SVG verbatim with reordering=0; all path optimization happens in Plottter.

Install (on the Raspberry Pi)

git clone <this repo> plottter-daemon && cd plottter-daemon
python3 -m venv .venv && . .venv/bin/activate
pip install -e .                 # the daemon itself (stdlib only)
# Driver for the plotter (not a hard dep, install separately):
pip install https://cdn.evilmadscientist.com/dl/ad/public/AxiDraw_API.zip

Your user needs serial access (usually already true on Raspberry Pi OS): sudo usermod -aG dialout $USER (then re-login).

Run

# Real plotter, auto-generated token (printed once):
python -m plottter_daemon --port 8080 --device-name "iDraw H SE A2"

# Trusted home network, no auth:
python -m plottter_daemon --no-auth

# No hardware yet? Simulate a plotter to test the Plottter connection:
python -m plottter_daemon --fake --no-auth

Then in Plottter → Plot with AxiDraw → Remote Plotter (network): enter the URL (http://<pi-host>:8080) and the token, tick Send to remote device, and click Refresh connection.

Run on boot (systemd)

Edit deploy/plottter-daemon.service (user, paths, --token, --device-name), then:

sudo cp deploy/plottter-daemon.service /etc/systemd/system/
sudo systemctl daemon-reload
sudo systemctl enable --now plottter-daemon

API (v1)

Method & path	Purpose	Auth
GET /api/v1/health	liveness + device + idle/busy + auth_required	open
GET /api/v1/version	daemon / API version	open
POST /api/v1/jobs	{svg, settings} → {job_id}; 409 if busy	token
GET /api/v1/jobs/{id}	status / progress	token
POST /api/v1/jobs/{id}/control	{action: pause|resume|stop}	token
POST /api/v1/manual	{command, settings} (raise/lower pen, motors)	token

settings is exactly the dict Plottter's plot dialog produces (model, speed_pendown/penup, pen_pos_up/down, pen_delay_*, const_speed, …). Pen positions ride with each job/manual call — the daemon keeps no machine state.

Develop / test (no hardware)

pip install -e ".[dev]"
pytest -q          # exercises the server + job manager via a FakeExecutor

plottter_daemon/executor.py has the only hardware code (AxiDrawExecutor, a faithful copy of Plottter's export/axidraw.py) and a FakeExecutor used by the tests.

License

MIT.