WoWee vs. Benilla: The Reality of Rebuilding a WoW Client in 2026

If you pull up the AI Overview for open-source World of Warcraft clients right now, it will spit out a static comparison of GitHub stars and languages. That is useless to you as a developer. The real question isn't which repo has more lines of code; it's which architectural philosophy is going to save you from a debugging nightmare three months from now.

Having pulled the raw repo metadata, language breakdowns, and license structures for both Benilla and WoWee, the reality is that these two projects represent a strict fork in the road for open-source emulator development in 2026.

---

The Data Deception: Why Benilla's 123 Stars Beat WoWee's 583

The reality is that raw GitHub metrics heavily favor legacy C++ projects, but they actively obscure modern developer velocity. At first glance, Kelsidavis's WoWee looks like the obvious choice. It has 583 stars and 70 forks compared to samwhosung's Benilla (123 stars, 27 forks).

But context is everything. WoWee was created in February 2026. Benilla was created roughly 12 days ago. Generating 123 stars in under two weeks for a highly niche Rust/Bevy MMO client indicates an incredibly aggressive demand for modern-stack reimplementations that WoWee's slower, steady C++ growth doesn't reflect.

Furthermore, the codebase size tells the true story of scope. WoWee is sitting at 218,016 GitHub size units—a massive monolith incorporating Python tooling, GLSL shaders, Lua scripting, and cross-platform shell scripts. Benilla is 9,433 units of pure Rust and WGSL. If you are looking to contribute or fork a project, reading and understanding 9KB of modern Rust is infinitely more viable than untangling a 218KB C++ polyglot engine.

The Architecture Divide: Bevy's Modernism vs. C++'s Legacy Plumbing

Data indicates that the choice between these two repos boils down entirely to your tolerance for manual memory management versus ecosystem safety.

The WoWee Approach (C++ / Vulkan / SDL2)

WoWee is a true "engine experiment." It targets Classic, TBC, and WotLK. This means the developers are building抽象 layers for multiple client versions, handling vanilla 1.12.1 packet structures alongside WotLK terrain rendering. If your goal is to build a comprehensive, multi-expansion custom client—and you are deeply experienced in C++ and Vulkan—this is the only repo that matters. However, you will be fighting traditional C++ demons: memory leaks, undefined behavior, and complex build chains involving Python and Lua.

The Benilla Approach (Rust / Bevy)

Benilla has a single, laser-focused goal: World of Warcraft 1.12.1. By leveraging the Bevy engine, the project outsources the grueling low-level rendering and ECS (Entity Component System) plumbing. The reliance on WGSL (WebGPU Shading Language) instead of GLSL also future-proofs the rendering pipeline. What you lose in expansion support, you gain in compile-time safety. Rust's borrow checker forces the network packet parsing and state management to be inherently thread-safe. You will spend less time hunting down segmentation faults and more time actually implementing game logic.

The Hidden Trap: Licensing and Community Health

The most critical operational difference between these two projects isn't technical; it's legal. A non-standard license on an open-source game engine is a critical liability for any serious contributor.

Benilla utilizes the Apache-2.0 license. It is clean, standard, and explicitly protects contributors from patent trolls while allowing commercial use. WoWee is listed under a generic "Other" license. In the emulation scene, ambiguous licensing often means the author has included custom clauses restricting certain types of use, or hasn't bothered to formalize it at all. If you intend to fork WoWee to build a commercial project, or even just contribute heavily, that "Other" tag is a massive red flag that requires a manual audit of the repo's LICENSE file before you write a single line of code.