What Is LinkedQL ​

LinkedQL is a modern SQL interface that drops into any JavaScript application or agentic workflows with advanced capabilities.

At its simplest, it gives you a stable way to talk to data:

  • query()
  • stream()
  • transaction()
  • wal.subscribe()

But the same gives you a much broader system:

  • live queries as a first-class runtime capability
  • powerful SQL language shorthands for relationships, JSON, and more
  • cross-runtime quering via EdgeClient (e.g. client -> server querying)
  • an embeddable local SQL engine named FlashQL, that itself enables...
  • federation, materialization, and bidirectional sync
  • version safety and point-in-time replay
  • and more

Why LinkedQL ​

If you've built a realworld system before, you most certainly stacked up tools and capability layers on top of SQL:

  • ORMs for relationships
  • schema mappers and a migration system
  • subscription servers like GraphQL for live updates
  • custom local caches and sync layers for offline-first apps
  • and probably more

The idea with LinkedQL is to be what SQL can be: a self-aware, reactive system that can directly answer modern application flows.

LinkedQL's direction is to absorb more of SQL's compensatory layers into the language.

The Three Big Areas ​

Common Client Contract ​

Across runtimes, LinkedQL tries to keep the core application contract stable:

  • query normally with query()
  • do live queries with query({ live: true })
  • do pull-based queries with stream()
  • do explicit transactional work with transaction()
  • subscribe to table-level commits with wal.subscribe()

See: Query Interface

Language Capabilities ​

LinkedQL extends SQL with application-oriented language features such as:

These are about expressing richer intent in the query itself instead of pushing that intent into external glue code.

Runtime Capabilities ​

LinkedQL also extends the runtime side of the database contract:

These collapse existing tall tooling stacks, capability layers, and large scale glue code into a single application-ready primitive.

FlashQL's Role ​

FlashQL is an answer to an everyday question: how do I run SQL locally from the browser/edge runtime/current process? How do I optionally back the local instance up with an upstream database, plus: optionally get both instances in sync?

It gives you:

  • a local SQL engine, with pluggable persistence backends
  • federation, materialization, and sync with a strong conflict handling model
  • historical data and point-in-time replays

See:

Where to Start ​

MIT Licensed

© Oxford Harrison