Getting Started with Nim: Tauri-like Nimino and Zed-like Nimculus

Nim is an interesting language that combines the strengths of system programming languages like C and Pascal with the ease of writing high-level languages such as Python. While Rust focuses on safety, Nim offers flexibility—you can “downshift to low gear only when necessary.” In typical development, you’ll be fine with high-level code, but if you need low-level operations or interoperability with C ABIs, you can smoothly transition to using pointers and manual memory management. This is particularly advantageous when directly handling OS/GUI APIs like Win32, GTK, or WebKitGTK.

Thanks to this flexibility, compiling Nim to native code is straightforward, enabling high performance while keeping distribution sizes small. Furthermore, by leveraging macros and template features, you can build safe APIs on top of low-level FFI. The Python-like syntax has a low initial learning curve, accelerating prototyping and experimentation. Since Nim links C and C++ libraries more directly, I’m looking forward to comparing its performance with Rust.

Recently, I’ve started two projects in Nim: “Nimino,” a lightweight cross-platform WebView desktop app framework, and “Nimculus,” a GPU-native code editor. Unlike Electron, which bundles Chromium entirely, Nimino leverages the OS’s native WebView (WebView2 on Windows, WebKitGTK on Linux). This design philosophy is influenced by Tauri, with initial targets being Windows, Native Linux, and WSL—macOS is planned for future expansion. Intended uses include internal tools, admin panels, local AI tools, and developer utilities.

The second project, Nimculus, is a GPU-native code editor built using Nim and NimNUI, drawing inspiration from Zed. The initial target platform is Apple Silicon-based macOS, implementing Metal rendering, Cocoa integration, tree-sitter parsing, and a Piece Table-based editor core. This is also in early stages.

At first glance, these projects seem to head in different directions, but they share a common validation goal: determining how practical a GUI/desktop foundation can be built with Nim. We aim to assess Nim’s strengths and weaknesses across OS APIs, WebViews, GPU rendering, packaging, and the overall development experience.

Nimino leverages native WebView capabilities as its foundation, while Nimculus pushes rendering performance in a GPU-native editor. However, neither is a simple port or clone of Tauri or Zed; instead, they reinterpret those design philosophies to align with Nim’s language features.

For example, Nimino utilizes OS-side technologies like WebView2 and WebKitGTK while writing host logic in Nim for cross-platform app development. Conversely, Nimculus implements Metal rendering and a Piece Table-based editor core in Nim to harness GPU power for a fast, responsive editing experience.

Though the directions differ, we intend to share insights between them. Knowledge gained from OS integration, packaging, and lifecycle management in Nimino can benefit Nimculus, while input handling, rendering, and editor core design learnings from Nimculus should contribute to a broader understanding of Nim’s GUI ecosystem.

Details on these projects and deeper dives into Nim’s design philosophy are available on the dedicated “Nim Architecture Guide” site. You’re welcome to check out their respective repositories if interested:

• Nim Architecture Guide
• Nimino Repository
• Nimculus Repository