WearWare

An application to turn a Raspberry Pi with an LED Matrix into a wearable LED screen t-shirt.

Setup

Note: Some of the toolchain (Build tasks etc) are currently Windows-specific (WSL - Windows Subsytem for Linux is a requirement). If running a different OS, very minor adaptations will need to be made (Convert WSL commands to native linux)

On the Pi:

Dependencies

• .Net 10 SDK wget https://packages.microsoft.com/config/debian/13/packages-microsoft-prod.deb -O packages-microsoft-prod.deb sudo dpkg -i packages-microsoft-prod.deb rm packages-microsoft-prod.deb sudo apt update && sudo apt install -y dotnet-sdk-10.0 You probably need the .NET SDK 8 on the Pi too, to build the RGB LED Matrix library
  Installing .Net 10 SDK is probably best, as that covers all scenarios.
• (For debugging) VSDBG
  curl -sSL https://aka.ms/getvsdbgsh | /bin/sh /dev/stdin -v latest -l ~/vsdbg

Disk setup

Pick a folder to deploy to (eg /root/dev/WearWare) This folder should contain the following subfolders:

• bin - WearWare binaries should be in here.
  When developing, deploy to this folder, not to the parent folder. This is because the rsync will wipe everything in the deploy folder (Wipe your library and config etc)
• config - This is where configuration of the app will be stored
• incoming - Uploaded files will go here
• library - Imported (Converted) files will go here
• playlists - Playlists will be stored here
• quickmedia - QuickMedia items will be stored here
• tools - Tools (eg the led-image-viewer executable) will go here

Deploy assets

• Build the rpi-rgb-led-matrix library and utils.
  The provided build tasks can be used to do this
  (remote-build-native-lib followed by remote-build-lib-utils)
• Place a compiled copy of led-image-viewer in the tools folder
  This can be done with the copy-util-remote-to-remote task

Dev machine setup

• .NET 10 SDK must be installed
• Set the environment variable WEARWARE_REMOTE_MACHINE to point to the hostname and username used to SSH into the Pi (eg root@DietPi)
• Set the environment varibale WEARWARE_REMOTE_DIR to point to the folder name where WearWare is to be deployed to on the Pi (eg /root/dev/WearWare)
• Set the environment variable WEARWARE_REMOTE_LIB_DIR to point to the folder where the RGB LED Matrix repo lives, eg /root/dev/rpi-rgb-led-matrix
  Ensure the bin folder is created in this folder (eg /root/dev/WearWare/bin)

Remote Dev (VSCode on local machine, debug on Pi)

• Passwordless SSH to the Pi as root must be configured

  Dev on Windows

  Launch configurations are included in the repo

• WSL must be installed, and Rsync must be added to WSL
  This is to allow VSCode to transfer the built files to the Pi

• The Command Variable VSCode extension must be installed
  This allows us to build linux style paths for use in tasks.json

• Select the Debug (Remote Pi) - WSL configuration

• Hit F5 to debug - code will be run on the Pi

  Dev on Linux / Mac using VSCode

  Launch configurations are not included in the repo
  However, the majority of tasks are absolutely compatible with linux with linux / mac
  You should only need to:

• Re-implement the wsl-rsync-ww-to-pi task (Very simple, just remove WSL)

• Duplicate the wsl-deploy-pi-debug compound task, replace wsl-rsync-ww-to-pi with your new task.

• Duplicate the Debug (Remote Pi) - WSL configuration and point it at your new compound task

Local Dev (VSCode on local machine, Pi functionality mocked out)

In this mode, you can work on business logic etc without needing to run on the Pi
It should work fine on any OS

• Select the Debug (Local with mocks) configuration
• Hit F5 to run
• A debug interface will be available on http://localhost:5000/mocks
  This will allow you to simulate GPIO button presses.
• In this mode, all LOOP type animations will only last for 1 second, as it has no way of knowing how long the animation is
• What the Pi would be playing on the screen is shown in the Debug Console of the IDE

Sundry build tasks

• remote-build-native-lib will build the RGB LED Matrix library on the Pi via SSH
• remote-build-lib-utils will build the RGB LED Matrix utils on the Pi via SSH
• copy-util-remote-to-remote will copy the led-image-viewer utility to the tools folder of WareWare on the Pi
• remote-build-bindings will build the C# bindings on the Pi via SSH
• wsl-deploy-pi-release (Windows only) will deploy to the Pi in RELEASE mode
  Be sure that the above two build tasks have been performed at least once before

Hardware notes

Physical buttons

This app supports playing media on press of a button.
Buttons are wired via free GPIO pins.
Pull-up / pull-down resitors are not required, as the button handler will ignore all input for 100ms after initialization to allow for the voltage level to settle.
There is currently no UI to configure which pins are used, instead config/buttonpins.json is used.
This will be created on startup if it does not exist, or you can create it yourself. Format is eg [2, 3, 16, 14]
If using an Electrodragon Active-3 matrix driver board, channel 3 can be disabled by desoldering the chip next to channel 3, freeing up 6 GPIO pins - 2, 3, 14, 16, 21 + 26
If also using a Petrockblock Powerblock (ATX power switch), ** DO NOT ** try and use pins 2 and 3 for the PowerBlock! It will shut down as soon as it starts up (Because these pins default to high). Recommend using pins 21 (Status) and 26 (Shutdown)
In this manner, you can have two panels, four buttons, and an ATX style power switch.

Using the app

• Upload some GIFs or images to the incoming folder of the Pi
• Point a web browser at port 5000 (eg http://dietpi:5000)
• Go to the Options page, click the Matrix Options button and set the matrix options for your panel (You may need to restart the app after)
• Go to the Import page and import files into the library. This will convert them to a stream
• Go to the Playlist page, create a playlist, and add items to it from the library
• Optionally, go to the QuickMedia page and assign items to the buttons (You may need to edit ButtonPins.cs to set which pins your buttons are connected to)