Darling Project Allows Use Of AI In Development To Boost Support For MacOS Applications On Linux

The Darling project, which develops a Wine-like layer for running macOS applications on Linux, has confirmed the possibility of using AI tools in development. Darling is now officially allowed to accept code created through AI, but subject to certain requirements. The decision was made in the context of a lack of developers and resources for the project, as well as against the backdrop of a constant increase in the number of proprietary Apple frameworks (while in macOS 10.14 there were 273 public and 878 private frameworks, in macOS 27 there were 464 and 2754, respectively).

It is assumed that the new rules will prevent stagnation of development and fragmentation of the community, and will also provide an opportunity to join forces with developers forks that actively use AI. Similar forks have advanced and demonstrated high development speed, for example, the fork darling-lol achieved the launch of a macOS build of the game League of Legends, and VibeDarling achieved success in supporting the ARM64 architecture and running many original applications macOS.

In the next two months, they intend to integrate developments from all external forks into the main Darling composition, using AI tools to merge changes. It is also planned to implement official support for i386, x86-64 and ARM64 (aarch64) architectures and to provide the ability to run ARM64 builds of applications on x86-64 systems using the Dynarmic recompiler. Long-term plans include stabilizing the compatibility layer with AppKit and creating our own graphical utilities for managing the Darling environment (for example, a Console.app-style log viewer), which will also act as reference test applications.

/Reports, release notes, official announcements.