Sparkling: A New Framework for Lynx
Featuring
What Sparkling Adds to Lynx
Sparkling is a framework created within TikTok’s cross-platform infrastructure and now open sourced for the community. While Lynx focuses on rendering and runtime, Sparkling provides the application layer around it: navigation, native capabilities, and project scaffolding.
In practice, Lynx is the engine and sparkling is the framework that makes it usable at scale.
A Multi-Page Model by Design
One of the most important takeaways was Sparkling’s multi-page architecture. Each page runs in its own native container, closer to a browser model than a single-runtime app. Navigation is scheme-based, similar to deep linking.
This structure enables strong boundaries between features and teams, which reflects how large organizations like TikTok build and maintain complex products.
Native Capabilities Through Sparkling Methods
We also explored Sparkling Methods - a structured bridge between JavaScript and native. Developers define TypeScript interfaces, run code generation, and implement platform-specific logic with clearly defined contracts.
On top of that, Sparkling ships with built-in modules like navigation and storage, forming a practical foundation for real applications.
Why It Matters
Sparkling shows how Lynx can be organized for enterprise environments. The layered approach - engine, infrastructure, app - mirrors microfrontend thinking in the web world and makes scaling teams and features more predictable.
The project is still evolving, but this live session clarified its direction and architectural intent.
Watch the full recording to see us bootstrap a Sparkling app, configure navigation, and implement a custom method step by step.
