We’re - in Digital-Archi Japan (DA) - excited to share a project we’ve been developing: the DA RhinoCompute Web App. Our goal was to build on the power of Rhino.Compute, its parallel computing capability to create a more dynamic and robust experience for web-based computational design, addressing some of the common limitations we’ve encountered.
This app is different from previous tools you might have seen, focusing on two major areas:
1. Instant Geometry Edit / Updates One of our primary objectives was to allow for a truly interactive workflow. Users can now modify and swap input geometries (points, curves, breps) on-the-fly and see the results update almost instantly in the browser. This creates a much faster and more intuitive design feedback loop, removing the need to constantly re-upload files or restart the process.
2. Enhanced Stability with Active Memory Management To support more complex and long-running sessions, we’ve integrated a dedicated memory management system on the backend. It actively monitors and optimizes how resources are used by the Rhino.Compute server. This means the application can handle more intricate models and heavier computations with greater stability, reducing server-side crashes and slowdowns.
Our Tech & Prototyping Journey
The user interface is organized into four distinct tabs, creating a streamlined workflow that separates geometry input, parameter control, 3D visualization, and data output.
For those interested in the technical side:
-
We rapidly developed our initial MVP using the Gradio framework, which was fantastic for testing our core ideas.
-
Our backend has since matured into some different applications, and we lean heavily on Pydantic for all data modeling. Every piece of data—from a slider value in the UI to complex geometry sent to the server—is defined by a strict schema. This acts as a powerful data contract, ensuring that malformed data is caught and rejected at the API level with a clear error message, long before it has a chance to cause a cryptic failure in a Grasshopper script. This approach has been a game-changer for debugging and has dramatically improved the app’s overall stability.
We believe this approach opens up new possibilities for sharing complex Grasshopper definitions with colleagues or clients who aren’t GH experts, providing them with a powerful, interactive, and reliable web tool.
We’re still in active development but wanted to share our progress with the community. We would love to hear your thoughts, feedback, or any features you’d consider essential for an application like this.
Thanks for reading!
Best, The DA Team DigitalArchi | Realizing a revolution in architecture.