Future work¶

Formal CAP theorem resiliency analysis.
HTTP network hops upper limits, idempotency
Comparative study of Graph databases, support more graph technology.
DGraph trumps ATM IMO but there are neat JVM parings of GQL/GraphDB as mentioned. Lacinia, Crux, Apollo, Relay, etc.=
What is manual vs. automated?
Support for native Corda Types as Graphable: Token SDK, Account SDK, Party, etc
Concrete typing comparisons on JVM v Go/Graph
UUID v Strings for graph ids, Dates
Add persistence before rawUpdates, queue off back of flow.
Introduce raw maps state i.e GraphMapState
Shared ownership of “master” schema. Frontend and backend schema segments.
Upgrade paths analysis with examples
Add metering. Want to avoid deep, taxing user traversal queries.
Fully automated “low partition” same machine of Corda and Graph.
Go process inside Corda? Crazy talk, probably not!
Strategy for UI development with samples. Server first UIs
Schema metadata-driven programming
Full-text search, analytics, subscriptions
Fully-fledged lambda js server in the cloud with braid.js client, with analysis

Code¶

CGraphReconciler -> Corda-Graph reconciliation process w/ availability potential
CGraphSchemaGenerator -> Schema generation tool for user CorDapp states
DGraphServiceFactory -> Start DGraph/Go lang from Corda →
DGraphGRPCClient - from JVM to Go Graph process (DQL CorDapp API)

Feel free to raise and MR or contact me directly over at eric.mcevoy@r3.com