Papers

Higher-order patterns in replicated data types

In this paper we argue that the essence of CRDTs is not about being “operation-based” or “state-based”. Instead, CRDTs have a more abstract meaning: they are ways to derive a value from a set of known operations. These derivations follow a few simple patterns which can be reused to succinctly define the individual behaviors of a large catalog of CRDTs. This means that when writing complex CRDTs it might be worth creating a little programming language which supports those patterns as first class primitives and implements them efficiently.

Adriaan Leijnse, Paulo Sérgio Almeida, and Carlos Baquero. 2019. Higher-Order Patterns in Replicated Data Types. In Proceedings of the 6th Workshop on Principles and Practice of Consistency for Distributed Data (PaPoC ’19). Association for Computing Machinery, New York, NY, USA, Article 5, 1–6. https://doi.org/10.1145/3301419.3323971

Links:

• Official version with free download
• Preprint
• BibTex