Manyana

Manyana is a demonstration project that explores a new approach to version control using Conflict-Free Replicated Data Types (CRDTs). Developed by Bram Cohen, the tool aims to address long-standing issues in version control systems, particularly around conflict resolution and merge reliability. Unlike traditional systems where merges can fail and conflicts can be opaque, Manyana leverages CRDTs to ensure that merges always succeed, with conflicts surfaced only when changes overlap, resulting in more informative and structured conflict markers.

A key feature of Manyana is its conflict presentation: instead of displaying two undifferentiated blocks of code during a merge, it highlights the specific actions taken on each side, such as deletions or insertions, and clearly indicates which contributor made each change. This makes it easier to understand the nature of conflicts and how they arose. The underlying CRDT model guarantees eventual consistency, so merges are deterministic regardless of the order in which branches are combined, and line ordering becomes permanent once determined. The tool also introduces a new way to handle rebasing, allowing for replaying commits onto a new base without losing historical information, facilitated by a "primary ancestor" annotation in the directed acyclic graph (DAG).

Manyana is implemented as approximately 470 lines of Python code that operate on individual files. It is not a full-featured version control system but rather a proof of concept intended to demonstrate that CRDT-based version control can address difficult user experience challenges. Features such as cherry-picking and local undo are not yet implemented, though the project's README outlines potential designs for these capabilities.

The code for Manyana is released into the public domain. The project is accompanied by a detailed design document explaining its approach and vision for the future of version control. Manyana is intended for developers and researchers interested in the evolution of version control methodologies and the practical application of CRDTs in collaborative software development.