|
From: 周 子益 <z12...@ou...> - 2025-10-29 02:21:58
|
Hi all, I've been following the Unicode discussion with interest, and I appreciate the valid concerns raised about maintenance burden. I'd like to propose a middle ground: an **optional** Unicode module that: ## Key Principles 1. **Opt-in by default** - no impact on existing users 2. **Minimal scope** - only what the compiler truly needs 3. **Stable dependencies** - using proven libraries like utf8proc ## What it WOULD include: - Basic UTF-8/16 decoding for source files - Unicode identifier validation (per C standard) - Optional normalization warnings (disabled by default) ## What it WOULD NOT include: - Homoglyph detection (leaving this to linters) - Font-dependent logic - Mandatory normalization ## Implementation Approach: - Compile-time flags control feature inclusion - Zero runtime cost when disabled - Clean module separation for maintenance This approach lets us support modern development environments while respecting the "if it ain't broke, don't fix it" principle for existing users. What are your thoughts on this balanced approach? Best regards, zhouziyi |