undr

undr is an open specification and reference implementation for encoding, validating, and exchanging a class of data objects in distributed systems. It aims to provide a simple, language-agnostic way to represent interconnected data items, with an emphasis on readability, versioning, and offline resilience. The project seeks to be lightweight enough for client applications while offering enough structure to support reliable data synchronization across networks and devices.

In its data model, an undr object comprises an identifier, a type, a set of attributes, and a collection of relationships to other objects. Optional metadata and a version field enable schema evolution and conflict detection. The specification defines a minimal JSON-based representation, with optional binary variants for performance-sensitive contexts. Validation can be performed against lightweight schemas or against a sparse vocabulary of common types, allowing developers to start simple and grow definitions as needed. The format is designed to support streaming updates and delta messages to minimize bandwidth requirements during synchronization.

History and scope: undr emerged from a collaborative effort in the open-source community to address data interchange in loosely coupled services and edge environments. Public drafts circulated in the late 2010s, with a reference implementation released for multiple languages. While not as widely adopted as established formats, undr has found niche use in experimental projects, peer-to-peer apps, and scenarios where a compact, schema-light exchange of interconnected objects is desirable.

Reception and status: reviewers note the clarity of the object model and the emphasis on evolution safeguards. Critics point to a modest ecosystem of tooling and libraries. Ongoing work focuses on expanding validators, improving interoperability with other data interchange standards, and growing community-maintained examples and tutorials.