It looks like the use case is specifying types for dataflow operators (aka endpoints for dataflow pipes) [0] and I surmise composition should be super easy. I w

It looks like the use case is specifying types for dataflow operators (aka endpoints for dataflow pipes) [0] and I surmise composition should be super easy. I was surprised not to see any mention of XML or XML Schema as prior art, especially with their discussion of schema registries. Edit: Oh, the point of reference is Kafka [1] [0] https://zed.brimdata.io/docs/language/overview/ https://zed.brimdata.io/docs/language/overview/ [1] https://docs.confluent.io/platform/current/schema-registry/index.html https://docs.confluent.io/platform/current/schema-registry/i...