It’s fascinating to see so many different parties converging on Datalog for reactive apps & UI. - There are several such talks at https://www.hytradboi.com/ ht

It’s fascinating to see so many different parties converging on Datalog for reactive apps & UI. - There are several such talks at https://www.hytradboi.com/ https://www.hytradboi.com/ (happening this Friday) - Roam Research and its clones Athens, Logseq, use Datascript / ClojureScript https://github.com/tonsky/datascript https://github.com/tonsky/datascript - differential-datalog isn’t an end-to-end system, but is highly optimized for quick reactivity https://github.com/vmware/differential-datalog https://github.com/vmware/differential-datalog - Datalog UI is a Typescript port of some of differential-datalog’s ideas https://datalogui.dev/ https://datalogui.dev/ I think an under appreciated aspect of Datalog is as a pure specification language for relationships. Schema languages like GraphQL specify a relation between types exists, but not how it is computed. Prisma and other ORM DSLs like ActiveRecord do a better job specifying how to compute a relation from SQL tables, but support only a few strict computation types that can be modeled using SQL JOIN on columns. Datalog lets you purely specify arbitrarily complex logical relationships between data types. Even if you were to only use Datalog as a rules engine at test time, it could still be quite powerful for validating logic implemented in different languages on the same data model, to say keep your Typescript, Kotlin and Swift logic in sync.