Darling(macOS兼容层)项目团队于2026年第四季度发布了最新开发进展,正式采纳了针对AI辅助代码贡献的新规则,以应对与Apple之间巨大的工程能力差距 1。根据新政策,人类贡献者必须对提交的每一行代码承担责任,并能够清晰解释其具体功能 1。该项目强调遵循“清洁室”原则进行功能重新实现,严格禁止直接使用反汇编代码或泄露的源代码,以确保法律合规性 1。
在技术层面,Darling已实现初步的ARM64架构支持,旨在适应Apple自研芯片的发展趋势 1。这一进展的背景是macOS系统复杂度的显著提升,例如macOS 27包含464个公共框架和2754个私有框架,远多于macOS 10.14版本 1。社区中已出现利用AI大量修改的分支案例,如VibeDarling,该分支成功运行了《英雄联盟》等应用程序 1。
对于未来规划,Darling团队计划通过Agent驱动的合并逻辑,将多个基于AI开发的分支工作合并回主仓库 1。长期来看,项目致力于稳定AppKit支持,并增加配置实用程序和日志查看器等原生GUI工具 1。
The development team behind Darling, a macOS compatibility layer, released its fourth-quarter 2026 update, officially adopting an AI-assisted contribution policy to address the engineering capability gap between the project and Apple 1. Under this new policy, contributors are required to take responsibility for their code and adhere strictly to clean-room principles 1. The guidelines mandate that human contributors must be able to explain every line of code, including those generated by AI tools 1. Additionally, the project explicitly prohibits the use of disassembled code or leaked source materials to ensure legal compliance 1.
Despite these constraints, Darling has achieved preliminary ARM64 support to align with Apple’s transition to self-designed chips 1. The project highlights the scale of the challenge it faces: macOS 27 contains 464 public frameworks and 2,754 private frameworks, significantly more than in macOS 10.14 1. Community efforts have already demonstrated progress using this approach, with branches such as "VibeDarling," which relied heavily on AI modifications, successfully running applications like League of Legends 1.
Looking ahead, the Darling team plans to merge various AI-developed branches back into the main repository using agent-driven merging logic 1. Long-term objectives include stabilizing AppKit support and adding native GUI utilities, such as configuration assistants and log viewers 1.
评论
还没有评论,欢迎留下第一条。