Skip to content

Dev Environment

How to set up your local machine for editing code and deploying it to the Pi Zero efficiently.

The Problem

You can't comfortably edit code directly on the Pi Zero — no GUI, small RAM, slow SD card I/O. The standard solution is to edit on your local machine and push changes to the Pi for execution.

Edit locally, sync to Pi, run on Pi. Simple, works with any editor, no VS Code dependency.

Workflow

  1. Edit files on your local machine
  2. rsync the changed files to the Pi
  3. SSH in and run the script

Setup

On your local machine, add this to your shell profile or create a Makefile:

# Sync the project to the Pi
alias dilder-sync='rsync -avz --exclude=".git" --exclude="venv" \
    ~/code/dilder/ pi@dilder.local:~/dilder/'

# SSH shortcut
alias dilder='ssh pi@dilder.local'

Example Makefile

# ~/code/dilder/Makefile

REMOTE = pi@dilder.local
REMOTE_DIR = ~/dilder

.PHONY: sync run watch

sync:
    rsync -avz --exclude=".git" --exclude="venv" ./ $(REMOTE):$(REMOTE_DIR)/

run: sync
    ssh $(REMOTE) "cd $(REMOTE_DIR) && source venv/bin/activate && python3 main.py"

watch:
    # Requires: brew install fswatch  (macOS)
    fswatch -o src/ | xargs -n1 -I{} make sync

Run with:

make run       # sync and run
make watch     # auto-sync on file change

Option B — VS Code Remote-SSH

Full IDE experience directly on the Pi. Works best with the Pi Zero 2 W — the original Pi Zero W may be too slow for VS Code's language server.

Setup

  1. Install the Remote - SSH extension in VS Code
  2. Open Command Palette → Remote-SSH: Connect to Host
  3. Enter pi@dilder.local
  4. Open the ~/dilder folder on the remote

You can now edit files, use the integrated terminal, and run scripts — all executing on the Pi.

Pi Zero W performance

The original Pi Zero W (single-core, 512MB) is borderline for VS Code Remote-SSH. Expect slow IntelliSense and occasional timeouts. The rsync workflow is more reliable on the Zero W. The Zero 2 W handles it fine.

Option C — Remote Debugging with debugpy

Run your script on the Pi with a debugger attached, and connect VS Code's debugger from your local machine.

On the Pi

pip install debugpy

# Start your script with debugpy listening on port 5678
python3 -m debugpy --listen 0.0.0.0:5678 --wait-for-client main.py

In VS Code (local machine)

Add to .vscode/launch.json:

{
    "type": "python",
    "request": "attach",
    "name": "Attach to Pi",
    "host": "dilder.local",
    "port": 5678,
    "pathMappings": [{
        "localRoot": "${workspaceFolder}",
        "remoteRoot": "/home/pi/dilder"
    }]
}

Press F5 to connect. You can now set breakpoints locally that trigger on the Pi.

Run at Boot (systemd)

When development is done, run Dilder automatically on boot:

# Create a service file
sudo nano /etc/systemd/system/dilder.service

[Unit]
Description=Dilder virtual pet
After=network.target

[Service]
Type=simple
User=pi
WorkingDirectory=/home/pi/dilder
ExecStart=/home/pi/dilder/venv/bin/python3 /home/pi/dilder/main.py
Restart=on-failure
RestartSec=5

[Install]
WantedBy=multi-user.target

sudo systemctl enable dilder
sudo systemctl start dilder
sudo systemctl status dilder

Useful SSH Aliases

# Add to ~/.ssh/config on your local machine
Host dilder
    HostName dilder.local
    User pi
    ServerAliveInterval 60
    ServerAliveCountMax 3

Then you can just ssh dilder instead of ssh pi@dilder.local.