You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Reframe tiko-di's public-facing story so the orchestrator + event-model identity is what humans and agents absorb first. The current framing organizes everything as a Spring Boot feature comparison, which primes a "Spring minus features" reading and has led LLM agents to fail simple tasks by searching for tiko equivalents of Spring annotations instead of using the orchestrator model.
Target audience is the greenfield "start a new service" path, not migration. A migration audience would re-introduce the 1:1 Spring frame this reframing exists to remove.
Guiding principles
These gate every sub-issue.
Lead with what tiko IS. Spring may appear later as illustration, never as the organizing axis.
Open design questions: deliberately small — extensions to the existing event model only (async generalization, scheduling-as-tick-event, retry-as-event-loop). Avro / Schema Registry, Kotlin interop, full observability stack are tracked separately and not in this bucket — see "Out of scope".
Cookbook, not wrapper. tiko owns the seam (@Produces), never the surface (a library's full API). Module status is reserved for capabilities that add tiko-native primitives — e.g. tiko-kafka earns module status because it adds @KafkaSource / @KafkaSink topology validation, not because it wraps the Kafka client.
Goal
Reframe tiko-di's public-facing story so the orchestrator + event-model identity is what humans and agents absorb first. The current framing organizes everything as a Spring Boot feature comparison, which primes a "Spring minus features" reading and has led LLM agents to fail simple tasks by searching for tiko equivalents of Spring annotations instead of using the orchestrator model.
Target audience is the greenfield "start a new service" path, not migration. A migration audience would re-introduce the 1:1 Spring frame this reframing exists to remove.
Guiding principles
These gate every sub-issue.
@Produces), never the surface (a library's full API). Module status is reserved for capabilities that add tiko-native primitives — e.g.tiko-kafkaearns module status because it adds@KafkaSource/@KafkaSinktopology validation, not because it wraps the Kafka client.Sub-issues
Execution order — same as listed:
docs: sweep banned vocabulary and delete tiko-di-gap-analysis.md #268— closed as no-op (the gap-analysis doc never landed in the repo; remaining banned-vocab hits are legitimate technical English absorbed by docs: rewrite README around the three-bucket orchestrator model #267)Out of scope