SparkOS/README.md

SparkOS

A minimal Linux distribution designed for simplicity and portability. SparkOS features:

• Minimal footprint: Barebones Linux system with busybox shell
• Portable: dd-able disk image for USB flash drives
• Custom init: Lightweight C init system
• Future-ready: Designed to support Qt6/QML GUI and Wayland
• Root elevation: Uses sudo for privilege management
• Bootstrap networking: Wired networking with DHCP for initial setup
• Minimal installation by default: kernel, init system, busybox, git, and sudo
• DNS configured with fallback to public DNS servers (8.8.8.8, 1.1.1.1)
• WiFi and advanced networking configured later via spark CLI

MVP Status

The current MVP provides:

• ✅ Custom init system written in C
• ✅ Working busybox shell environment
• ✅ dd-able AMD64 image creation scripts
• ✅ Minimal root filesystem structure
• ✅ Build system (Makefile)
• ✅ Wired networking configuration with DHCP
• ✅ DNS configuration with public fallback servers
• ✅ Docker container for testing
• ✅ Automated builds and publishing to GHCR
• ✅ Multi-architecture Docker images (AMD64 and ARM64)
• ✅ CI/CD pipeline for compiled release packages
• ✅ GitHub releases with pre-built binaries

Prerequisites

To build SparkOS, you need:

• GCC compiler
• GNU Make
• Linux system (for building)

To create bootable images (optional):

• Root privileges
• syslinux bootloader
• parted partitioning tool
• losetup for loop devices
• mkfs.ext4 filesystem tools

Quick Start

Using Pre-built Releases (Easiest)

Download the latest release package from the GitHub Releases page:

# Download the latest release (replace VERSION with actual version, e.g., v1.0.0)
wget https://github.com/johndoe6345789/SparkOS/releases/download/VERSION/sparkos-release.zip

# Extract the package
unzip sparkos-release.zip
cd sparkos/

# The init binary is already compiled and ready to use
ls -lh init

# Copy to your rootfs or use directly
cp init /path/to/your/rootfs/sbin/init

The release package includes:

• Pre-compiled init binary (statically linked, ready to use)
• Complete source code
• Build scripts and configuration
• Root filesystem structure
• Full documentation

Using Docker (Recommended for Testing)

The easiest way to test SparkOS is using the pre-built Docker image from GitHub Container Registry:

# Pull and run the latest image (automatically selects the correct architecture)
docker pull ghcr.io/johndoe6345789/sparkos:latest
docker run --rm ghcr.io/johndoe6345789/sparkos:latest

# Or build locally
docker build -t sparkos:local .
docker run --rm sparkos:local

# Build for specific architecture
docker buildx build --platform linux/amd64 -t sparkos:amd64 --load .
docker buildx build --platform linux/arm64 -t sparkos:arm64 --load .

The Docker image includes:

• Pre-built init system binary
• Minimal root filesystem structure
• Test environment for validation
• Multi-architecture support: Available for both AMD64 (x86_64) and ARM64 (aarch64) architectures

Images are automatically built and published to GitHub Container Registry on every push to main branch.

Building the Init System

# Build the init binary
make init

# Or use the quick build script
./scripts/build.sh

Setting Up Root Filesystem

# Create the root filesystem structure
./scripts/setup_rootfs.sh

# Install init to rootfs
make install

Creating a Bootable Image (Advanced)

⚠️ Warning: Creating bootable images requires root privileges and proper tools.

# Install required tools (Ubuntu/Debian)
sudo apt-get install syslinux parted

# Build everything and create image
make all
sudo make image

Installing to USB Drive

Once you have created the sparkos.img file, use the installation script to write it to a USB drive or storage device:

# Use the installation script (RECOMMENDED)
sudo ./scripts/install.sh /dev/sdX

# The script will:
# - Validate the target drive
# - Display warnings about data destruction
# - Require confirmation before proceeding
# - Show progress during installation
# - Verify successful installation

Replace /dev/sdX with your actual USB device (e.g., /dev/sdb, /dev/nvme1n1).

⚠️ WARNING: This will permanently erase all data on the target drive!

Project Structure

SparkOS/
├── .github/             # GitHub Actions workflows
│   └── workflows/
│       └── docker-publish.yml  # Docker build and publish workflow
├── config/              # Build configuration files
│   └── build.conf       # Build parameters
├── scripts/             # Build and setup scripts
│   ├── build.sh         # Quick build script
│   ├── setup_rootfs.sh  # Root filesystem setup
│   ├── create_image.sh  # Image creation script
│   └── install.sh       # Installation script for USB drives
├── src/                 # Source code
│   └── init.c           # Custom init system
├── rootfs/              # Root filesystem (generated)
│   ├── bin/             # Essential binaries
│   ├── sbin/            # System binaries
│   ├── etc/             # Configuration files
│   └── ...              # Standard FHS directories
├── Dockerfile           # Docker image definition
├── Makefile             # Build system
└── README.md            # This file

Architecture

Init System

SparkOS uses a custom init system (/sbin/init) that:

• Mounts essential filesystems (proc, sys, dev, tmp)
• Initializes wired networking via DHCP
• Spawns a busybox sh login shell
• Handles process reaping
• Respawns shell on exit

Root Filesystem

Follows the Filesystem Hierarchy Standard (FHS):

• /bin, /sbin: Essential binaries
• /etc: System configuration
• /proc, /sys, /dev: Kernel interfaces
• /tmp: Temporary files
• /usr: User programs and libraries
• /var: Variable data
• /root: Root user home
• /home: User home directories

Networking

SparkOS provides wired networking for initial bootstrap:

• DHCP: Automatic IP configuration via busybox udhcpc
• DNS: Fallback to public DNS servers (8.8.8.8, 8.8.4.4, 1.1.1.1, 1.0.0.1)
• Interface: Primary wired interface (eth0) configured automatically
• WiFi: Will be configured later via spark CLI after installation

Development

CI/CD and Docker

SparkOS uses GitHub Actions for continuous integration and delivery:

Automated Builds:

• Docker images are automatically built on every push to main/develop branches
• Compiled release packages are automatically built on every push to main/develop branches
• Both are also built for pull requests (testing only, not published)
• Tagged releases automatically create versioned Docker images and GitHub releases with compiled binaries
• Multi-architecture builds: Images are built for both AMD64 (x86_64) and ARM64 (aarch64)

Compiled Releases:

• Pre-compiled init binaries are available as GitHub releases for version tags
• Release packages include: compiled init binary, source code, build scripts, and documentation
• Download releases from the GitHub Releases page
• Build artifacts are available for all workflow runs (retained for 90 days)

Container Registry:

• Images are published to GitHub Container Registry (GHCR)
• Pull images: docker pull ghcr.io/johndoe6345789/sparkos:latest
• Available tags: latest, main, develop, version tags (e.g., v1.0.0)
• Docker will automatically select the correct architecture for your platform

Docker Development:

# Build Docker image locally
docker build -t sparkos:dev .

# Build for multiple architectures (requires Docker Buildx)
docker buildx build --platform linux/amd64,linux/arm64 -t sparkos:multiarch .

# Test the image
docker run --rm sparkos:dev

# Inspect the init binary
docker run --rm sparkos:dev sh -c "ls -lh /sparkos/rootfs/sbin/init"

Building Components

# Build init system only
make init

# Install to rootfs
make install

# Clean build artifacts
make clean

# Show help
make help

Adding Binaries to Root Filesystem

To create a fully functional system, you need to populate the rootfs with binaries:

# Required binaries (statically linked recommended)
# 1. Busybox - provides shell and most utilities including networking
cp /path/to/busybox rootfs/bin/

# 2. Git - for cloning spark CLI
cp /path/to/git rootfs/bin/
# Note: If git is dynamically linked, you'll need to copy its libraries too

# 3. Sudo - for privilege elevation
cp /path/to/sudo rootfs/bin/

# Create busybox symlinks for common utilities
cd rootfs/bin
for cmd in sh ls cat mkdir rm cp mount umount chmod chown ln \
           ip ifconfig ping wget udhcpc; do
    ln -sf busybox $cmd
done
cd ../..

# If using dynamically linked binaries, copy required libraries
ldd rootfs/bin/busybox  # Check dependencies
ldd rootfs/bin/git      # Check dependencies
ldd rootfs/bin/sudo     # Check dependencies
# Copy libraries to rootfs/lib or rootfs/lib64 as needed

Testing Network Connectivity

Once booted, you can test the network:

# Check interface status
ip addr show

# Test DNS resolution
ping -c 3 google.com

# Test direct IP connectivity
ping -c 3 8.8.8.8

# Download a file
wget http://example.com/file

Future Roadmap

• Qt6/QML full screen GUI
• Wayland compositor integration
• C++ CLI tools (spark command)
• Package management via spark CLI
• WiFi configuration via spark CLI
• Advanced network configuration

Contributing

Contributions are welcome! This is an early-stage project focused on:

1. Maintaining minimal footprint
2. Clean, readable code
3. Proper documentation

License

See LICENSE file for details.

Notes

This is an MVP implementation. The system currently provides:

• Basic init system with network initialization
• Shell environment
• Build infrastructure
• Image creation tooling
• Wired networking configuration

To create a fully bootable system, you'll also need:

• Linux kernel binary (vmlinuz)
• Essential system binaries: busybox, git, sudo
• Required libraries (if using dynamically linked binaries)
• Bootloader installation (handled by scripts)

Minimum System Requirements:

• Kernel: Linux kernel with networking support
• Init: Custom SparkOS init (included)
• Shell: Busybox with networking utilities (udhcpc, ip/ifconfig, ping, wget)
• VCS: Git (for installing spark CLI)
• Security: Sudo (for privilege elevation)

After bootstrap:

1. Use wired network to clone spark CLI via git
2. Use spark CLI to configure WiFi and other system features
3. Install additional packages as needed via spark CLI