Halfspace experimental IDE for solid modeling with distance fields

Halfspace is an experimental IDE for solid modeling that uses distance fields. It is best experienced on a computer, as mobile Safari has some WebGPU issues and pan/tilt/zoom interactions are not yet designed for multitouch. Halfspace serves as a showcase app for the Fidget kernel, which handles rasterization and meshing. Within the GUI, images are rasterized in real-time, and models can be exported as images or triangle meshes.

This write-up assumes familiarity with implicit surfaces, but provides background information to help readers understand the topic.

Implicit surfaces can be challenging to work with, as they require combining mathematical variables to create desired shapes. To simplify this process, various abstractions such as high-level languages and libraries are often used. In Halfspace's case, the developers have taken a dual approach: they include a growing standard library of shapes and transformations, while also making it easy for users to build complex models incrementally.

This approach allows users to split complex models into smaller, individually parameterized and visualized pieces.

Solid modeling refers to the creation of objects with a clear inside and outside. This property is crucial for creating objects that can be physically realized, which is important for Halfspace's creators, as they have a background in CAD/CAM software (with an emphasis on 3D printing). There are various methods to achieve solid modeling, but Halfspace opts for a GUI wrapped around the Fidget geometry kernel.

Users can design models using pre-defined primitives and hand-written scripts, and then export them as images or triangle meshes.

Instead of relying solely on constructive solid geometry (CSG) to build shapes and combine them using logical operations, Halfspace emphasizes the use of distance fields. This approach allows for easier diagnosis of issues, such as discontinuities in the boundary of shapes. By tweaking the gradient in distance fields, the developers aim to improve various algorithms, including mesh quality.

Currently, Halfspace is not suitable for load-bearing capacity, but it allows the developers to flex their software architecture skills and improve the underlying distance fields.