ORDTs

Systems are synchronous. The world is distributed. Data and computations must be designed in a way that allows them to be replicated, and converge even if some they happen concurrently.

There has been a lot of research recently on commutative replicated data types. These types offer some strong eventual consistency guarantees, which can be very convenient in distributed systems. Most of the research is dedicated to finding new designs for common data structures, which often do not offer the expressiveness required by some business use-cases.

Operational Replicated Data Types help solve this issue by considering the operations that sites want to achieve in the distributed system. Rather than focusing on data, this type focuses on intent by asking sites to explicitly encode their operations. It then offers multiple utilities to manage this data.

The naming conventions, and important parts of this library are inspired by the paper Deep Hypertext with Embedded Revision Control Implemented in Regular Expressions from Victor Grishchenko.

## What's included ?

For the moment, not much. You'll find a basic implementation of an ORDT.Weft, which is nothing more than a vector clock. In the future, the following elements should be included :

• A fully featured ORDT.Weave with support for folding, merging and efficient operation insertion.
• Instances of Json.Decoder and Json.Encoder for ORDT.Weft and ORDT.Weave. This will be handy when sharing data across sites.
• (Optional) Some wrappers around ORDT.Weave to implement some common replicated data types, like sets, sequences or counters.

Should I contribute ?

If you find a bug (or fix one), sure ! If you want to change the API, ideas are welcome, but I tend to be a bit conservative and can't promise we will include your fancy-does-everything-API right away in this library.