Objectives and Scope of IPLD


We want to see more decentralized designs appearing in technology.

We believe specifications for data exchange, design languages that work in decentralized contexts, and content-addressing (hash-linking) of data are all a key part of enabling that.

IPLD is a body of work to move forward on all those fronts.


The goal of IPLD is to make it possible to build "the next git" in hours, instead of days. (Or weeks, instead of months: whatever the original timescale is, slide it down a whole order of magnitude.)

We aim to do this by taking a bunch of the practical problems you need to solve to build a decentralized / content-addressed system, and turn them into non-issues: either by making solutions so good there's no need to argue with them (CIDs); or, by making it into a pluggable system so that you can make one choice now, and defer worrying about it again until much later, and have agility "for free" whenever it becomes necessary to change (multicodecs, multihashes, etc).

On top of that, we start adding superpowers with more library layers: ADLs provide sharding and scaling; Schemas provide a design language and ways to talk about data migrations; and features like Selectors to describe graph walks unlock whole other flourishings of ecosystems, like graph-based data transfer systems, etc.



We want IPLD to provide many of the useful attributes of an SQL-style database for purely local/offline applications: IPLD Schemas and IPLD Selectors provide features comparable to DDLs and SQL, and set the stage for building migration systems. ADLs provide features comparable to database indexing.

At the same time, IPLD works on a concept of "blocks" which are much more comparable to messages: they're easy to send and receive over the network, and critically, can be regarded individually (there's no supposition of monolithicity, as SQL-style databases have). And the IPLD Data Model is fundamentally more similar to document stores than relational databases.

Our aim is that IPLD should straddle this divide between databases and message systems, and operate on either side of it gracefully. The same libraries and concepts and tooling should work well regardless of if you're building an offline app or a massively distributed system with remote actors coming and going and sending messages all the time.