IdeaWell

Project ideas from Hacker News discussions.

Open source CAD in the browser (Solvespace)

Original Article

Hacker News Discussion

📝 Discussion Summary (Click to expand)

1. FreeCAD’s growing maturity

"FreeCAD is becoming like Blender and InkSpace – incredibly robust and capable and equivalent in most cases to the commercial alternatives." – bhouston

2. UI and kernel limitations

"The UI is still very convoluted and very hard to learn... the geometry engine... sometimes downright fails to compute fillets or surface/surface intersections correctly." – ur-whale

3. Emerging alternatives and LLM‑driven CAD

"FreeCAD doesn't have the limitations of SolveSpace, and the UX is actually decent now. I moved to that." – IshKebab

🚀 Project Ideas

[WebCAD Studio]

Summary

  • A browser‑based parametric CAD editor that lets users design parts with full constraint solving, automatic fillet/chamfer generation, and one‑click STL/STEP export, eliminating the need for heavy desktop software.
  • Solves the pain point of “heavy desktop tools and lack of AI‑assisted geometry helpers” highlighted by HN commenters.

Details

Key Value
Target Audience Hobbyist makers, educators, small‑scale manufacturers who need quick, shareable CAD without installing software.
Core Feature Real‑time constraint solver with AI‑suggested geometry fixes and one‑click fillet/chamfer generation.
Tech Stack Rust compiled to WebAssembly for the kernel, React for UI, Three.js for viewport, OpenAPI for cloud sync.
Difficulty Medium
Monetization Revenue-ready: Subscription

Notes

  • HN users repeatedly mention “browser CAD would be amazing” and “FreeCAD UI is convoluted,” so a slick web UI would attract them.
  • Could spark discussion on integrating SolveSpace‑style solvers with modern UI paradigms.

[FreeCAD Extensions Hub]

Summary

  • A curated marketplace and installer for FreeCAD add‑ons that instantly provide missing capabilities such as advanced chamfers, enhanced rendering, and BIM support.
  • Addresses the frustration of “FreeCAD UI is still very convoluted and lacks certain features” expressed by several commenters.

Details

Key Value
Target Audience Existing FreeCAD users, professionals in woodworking and mechanical design, hobbyists seeking a smoother workflow.
Core Feature One‑click extension installer with automatic dependency resolution and optional premium extensions for advanced operations.
Tech Stack Python‑based FreeCAD API wrapper, Electron‑style distribution, SQLite for extension metadata.
Difficulty Low
Monetization Revenue-ready: Freemium

Notes

  • Commenters like ur‑whale and jstanley note the need for better UI and missing features—this directly satisfies that gap.
  • Could generate discussion on community‑driven extension models versus proprietary CAD ecosystems.

[LLM CAD Assistant]

Summary

  • A web service where users type natural‑language design requests (e.g., “Create a 20 mm chamfer on a 40 mm cylinder”) and receive a downloadable STEP model, leveraging an LLM‑driven parametric interpreter.
  • Tackles the “I can’t move beyond TinkerCAD” problem and the desire for AI help in CAD workflows.

Details

Key Value
Target Audience 3D printing hobbyists, educators, and non‑CAD users who want rapid prototyping without learning complex software.
Core Feature Natural‑language to parametric model conversion with export to STEP/STL and integration with OpenSCAD for scripting.
Tech Stack Python backend with GPT‑4‑style LLM, FastAPI, OpenCASCADE for geometry, Docker for scaling.
Difficulty High
Monetization Revenue-ready: Pay-per-model

Notes

  • HN discussion includes “why not use AI to write CAD code?” and “need a genius to replace Fusion 360,” indicating strong interest.
  • Could lead to debates on LLM limitations for geometry kernels and opportunities for open‑source kernel integration.

Read Later