Advent of Code 2023 Visualizations on the Nintendo Switch

Advent of Code 2023 puzzles visualized on the Nintendo Switch. The Switchbrew examples were used as the starting point, and this repo follows the same Makefile and source file/directory structures from that.

These were tested on a Mariko Switch.

Build Guide

Switchbrew has a guide on setting up a development environment. In my case I used the official Docker image from devkitPro, which also comes with the switchbrew examples.

Assuming you are using the Docker image:

1. Go to your desired project directory and optionally clone this repo.
2. Run the docker image interactively and mount your project directory somewhere in the image's filesystem, say /home/dev:

   docker run -it --mount type=bind,source="$PWD",target=/home/dev --entrypoint=/bin/bash devkitpro/devkita64`
   

3. Go to one of the directories containing a Makefile and a source directory containing the C program's main function.
   • If you cloned this repo to your project directory, navigate to one of the directories inside /home/dev/aoc2023_vis_nx/. The Makefile in this repo is just copied from the switchbrew examples.
   • If you want to build the switch brew examples, they are inside /opt/devkitpro/examples/switch/
4. If successful, a <directory>.nro file should be generated. Congratulations! You can run this directly on the Switch as a homebrew application.
5. To use your own inputs, add an input_dec<daynum>.txt file in the same directory as the .nro file. Due to some puzzle inputs being too large, modify the inputs or the source to read only a subset of the input.

Visualizations

Day 14

This uses the accelerometer readings from the JoyCon. Most of the boilerplate code here comes from the sixaxis example.

• The accelerometer readings and rocks are sampled after every delay to give an "animation" effect to the visualization. So a directional tilt doesn't need to finished, and we can change directions in the middle of a tilting "animation".
• The load on the northern support beam is calculated in sync with the grid currently printed
• TODO: Accelerometer thresholds, and delay values were arbitrary (what felt right), and were hardcoded in the source. Might probably try out dynamically adjusting them via the console inputs.
• TODO: Try out audio examples of when rocks are hitting each other