The possibility of Fornjot as an alternative CAD kernel for CadQuery

[huskier]

CadQuery is an intuitive, easy-to-use Python module for building parametric 3D CAD models. At present, the underlying CAD kernel for CadQuery is OpenCascade. Multiple CAD kernels supporting was considered in the initial design of CadQuery. However, the open source CAD kernel is rare. To experiment with adaptation to other CAD kernels, I've created a minimal CadQuery. IMO, another kernel may benefit the development of CadQuery.

To make the adaption, we firstly need to create a Python binding for Fornjot. I wonder whether somebody has done the Python binding thing? Through the web search, I have found no work on the binding.

On top of the above mentioned Python binding, Fornjot should support the following CAD types:
Vertex
Edge
Wire
Face
Shell
Solid
CompSolid
Compound

A STEP exporting function should be there for output.

@hannobraun I'd like to see that the CadQuery could work with Fornjot, and I'd like to make this happen, even if only a simplest prototype at the first beginning.

[hannobraun]

Hey @huskier, thank you for your interest in Fornjot!

I think using Fornjot as a kernel for CadQuery is an exciting possibility, but I worry that the project is not ready for that. Specifically, the geometry that Fornjot can represent is very limited, which is something I'm working on fixing.

To make the adaption, we firstly need to create a Python binding for Fornjot. I wonder whether somebody has done the Python binding thing? Through the web search, I have found no work on the binding.

To my knowledge, no Python binding exists for Fornjot.

On top of the above mentioned Python binding, Fornjot should support the following CAD types:
Vertex
Edge
Wire
Face
Shell
Solid
CompSolid
Compound

Fornjot supports most of those. It doesn't represent wires, and I don't know what CompSolid and Compound are.

A STEP exporting function should be there for output.

This is not supported. And I don't think it makes sense to work on that until #2118 is done.

---

Since this doesn't track actionable work that needs to be done within this repository (except for STEP support, which is on the feature wishlist, I'm converting this to a discussion.

[huskier]

Thank you for your explanation. I will keep tracing the status of the Fornjog development. If I have found something I can do, I would like to jump into the work.

[jdegenstein]

I am a contributor on the build123d project which is related to CadQuery and also depends on OCCT. I have been following Fornjot for awhile and I am also interested in the eventual potential for CQ+Fornjot and build123d+Fornjot.

Regarding bindings, it might be worthwhile to investigate projects like PyO3. Given hannobraun's response, I don't know if this effort is worthwhile yet. I definitely would not want to put an additional drain on the resources here.

[hannobraun]

Thank you for the info, @jdegenstein!

Given hannobraun's response, I don't know if this effort is worthwhile yet.

It might be a worthwhile project to start experimenting with Python bindings, and provide feedback. I have absolutely no idea how PyO3 works, for example, and for all I know I might do something easily avoided, that prevents PyO3 from working for this project.

I do definitely think it is too early to create Python bindings with the expectation of having anything that is useful for CadQuery or build123d. Fornjot's functionality is simply too basic for that, and that won't even start to change until #2118 (or something like it) is implemented and stable enough to be built on top of.

I definitely would not want to put an additional drain on the resources here.

To be very clear, any Python bindings would be a 3rd-party effort and not be maintained by me. There's simply no way I can do it, given current levels of funding.