In Byzantine settings, Conflict-free Replicated Data Types (CRDTs) provide strong eventual consistency without coordination, but classical approaches assume benign participants. Typically, convergence is enforced through agreement on update validity, often relying on identity-based filtering. However, such methods struggle in post-compromise scenarios where a previously correct participant turns malicious: retroactive exclusion of its updates may break causal dependencies and invalidate subsequent computations.
This paper decouples identity-based trust from content-based trust and introduces a fine-grained trust model that combines both dimensions. Building on deterministic reconstruction, our approach allows replicas to preserve previously accepted updates while enabling selective inclusion or exclusion based on both the originating identity (e.g., public keys) and the semantics of individual updates.
Trust decisions can incorporate application-level policies, enabling precise control over the impact of each update on the system state. Our approach preserves causal consistency and enables robust and flexible handling of both Byzantine and faulty behavior in decentralized CRDT systems.
Blogger's Review: This paper presents an innovative trust model in the Byzantine fault tolerance and CRDT domain, addressing the shortcomings of traditional methods in the presence of malicious participants. By decoupling identity from content trust, it significantly enhances the flexibility and robustness of the system, making it a worthy study for developers to explore and implement.