brachiosaurus 🦕

An incredibly basic turtle drawing tool I've built for making generative art with my BrachioGraph pen plotter.

Setup

Build a BrachioGraph and set it up with a Raspberry Pi according to the instructions.

The Python 3 script brachiosaurus.py contained in this repository has no dependencies aside from the standard library. If you run it on a computer whose hostname is raspberrypi, it'll dynamically attempt to import brachiograph and draw the programmed pattern, otherwise it'll generate an SVG version (see above) instead and emit it on stdout.

The Raspberry Pi hostname (and the BrachioGraph configuration – the default probably won't be accurate for your build) are configurable at the top of brachiosaurus.py.

Deployment

Copy brachiosaurus.py to the directory on your Raspberry Pi in which you've installed the BrachioGraph software. With my setup, that's done like this:

scp ~/Dropbox/code/brachiosaurus/brachiosaurus.py pi@raspberrypi.local:/home/pi/BrachioGraph/

Usage

Place the code to generate the image you'd like to draw inside the main function.

Similar to how you'd use the default BrachioGraph software – on your Raspberry Pi, after boot, run sudo pigpiod to set up hardware servo timing. Then:

cd BrachioGraph
source bin/activate
python3 brachiosaurus.py

Alternatively, on your "development" machine where you haven't set up the BrachioGraph software, simply run

python3 brachiosaurus.py > ~/Desktop/preview.svg

and open preview.svg in a web browser or other SVG viewer of your choice. This allows for fast and easy development of a complicated drawing without having to actually plot it until it's right, although I advise not previewing simpler ones – you never know which happy accidents you'd miss.

You want to keep library and drawing code separate?

Not a bad idea at all! Create a file mydrawing.py next to brachiosaurus.py:

import brachiosaurus as bs  # heartbreakingly, python won't accept 🦕 as an alias

c = bs.Canvas()

# turtle drawing commands go here

plotter = bs.AutoPlotter().from_canvas(c)
#plotter = bs.AutoPlotter().from_file("test-patterns/accuracy.json")
plotter.emit()

Then, as above, deploy this new file to your Raspberry Pi:

scp ~/Dropbox/code/brachiosaurus/mydrawing.py pi@raspberrypi.local:/home/pi/BrachioGraph/

And run it:

python3 mydrawing.py.

Or run it without deployment:

python3 ~/Dropbox/code/brachiosaurus/mydrawing.py > ~/Desktop/preview.svg

You're looking for some examples?

Look no further than examples.py – which has the same structure as the mydrawing.py file outlined above and can be deployed and run in the same manner – where I've been collecting the code of some drawings I've made myself.

Alternatively, you can easily create your own examples using UJI, a generative art tool of my own design – it's able to export drawings in the JSON format supported by the AutoPlotter().from_file() function.

You've got a different plotter?

The code is intentionally kept simple and extendible.

• Drawings are represented as lists of lines,
• lines are lists of points, and
• points are 2-ary lists of x and y coordinates.

If your plotter can be driven via Python and you're able to implement a function that translates this representation into whatever representation your plotter expects (see RealPlotter.emit and FakePlotter.emit), you should be golden. (Feel free to send a pull request!)

Related Work

• Converting SVGs into the JSON format expected by this tool and the BrachioGraph software: https://github.com/Findus23/BrachioGraph-Utils
• Raster images can be vectorized using the built-in BrachioGraph tooling, see here: https://github.com/evildmp/BrachioGraph/blob/master/linedraw.py
• Advanced turtle drawing tool for AxiDraw plotters: https://github.com/fogleman/axi