[ROADMAP] Language binding

[gg582]

Support Range

• Rust (Already Implemented, Need Hardening)
• Go
• Java
• Zig
• D
• F#

How should we do?

Plan A. C ABI and FFI - Recommended

Pros

• Fast development: high code reuse without modifying the core C codebase
• Automation friendly: bindings can be generated using tools like bindgen, cgo, Java FFM, and P/Invoke
• Natural fit for systems languages: Rust and Zig call C ABI natively with zero overhead

Cons

• FFI call overhead in runtime managed languages: Go goroutine stack switching and JVM boundary checks introduce latency
• Manual lifecycle management: developers must manually bridge C pointers to target language GC or ownership models
• Deployment complexity: dynamically loaded shared libraries require strict path and environment management in production

Plan B. Native C API Extensions

Pros

• Tight runtime integration: binds directly to host GC lifecycles, exceptions, and internal object layouts
• Lowest overhead for managed environments such as direct JNI access

Cons

• Extreme development overhead: requires learning and writing separate glue code for each language runtime
• Incompatible with several target languages: Rust, Zig, and F# do not use traditional C extensions and rely directly on C ABI
• Build pipeline fragmentation: maintaining compilation matrices and toolchains across multiple languages and targets creates severe technical debt