Analysis of the Matrix Event Graph Replicated Data Type

Matrix is a new kind of decentralized, topic-based publish-subscribe middleware for communication and data storage that getting particularly popular as basis secure instant messaging. By comparison with traditional decentralized systems, replaces pure message passing replicated structure. This structure, which we extract call the Event Graph (MEG), depicts causal history messages. We show this MEG represents an interesting important type applications are based on histories events: First, prove Conflict-Free Replicated Data Type and, thus, provides Strong Eventual Consistency (SEC). With SEC being among best known achievable trade-offs in scope well-known CAP theorem, powerful consistency guarantee while available during network partition. Second, discuss implications byzantine attackers type's properties. note MEG, it does not strive consensus or strong consistency, can cope n > f environments participants, byzantine. Furthermore, analyze scalability: Using Markov chains, study number forward extremities over time observe almost optimal evolution. conjecture property inherent to underlying spatially inhomogeneous random walk. properties shown, promising element set structures applications, but distinct compared blockchains distributed ledger technologies.