Styles

Summary and comparison of all 8 styles; decision guide

Broker vs mediator topologies; DDD: three event types (notification, ECST, domain event); private vs public events; distributed big ball of mud anti-pattern; data liberation

Technical partitioning; simplest monolith; domain layer isolation as the DDD prerequisite; Smart UI anti-pattern; sinkhole anti-pattern; closed/open layers

Core system + plug-ins; single quantum; customisability-focused

Fine-grained domain services; maximum quanta; highest agility/cost; DDD: deep module heuristic, BC vs microservice relationship, subdomain as safe granularity, OHS/ACL for deeper services

Domain-partitioned single quantum; appropriate coupling via fitness function enforcement; intermediate migration step

Orchestration-driven SOA; technically partitioned; legacy enterprise pattern

Filter-and-pipe; technically partitioned; ETL/data processing; EIP treats Pipes and Filters as a messaging composition pattern

Hexagonal/onion/clean architecture; DIP-inverted dependencies; business logic at centre; DDD fit: domain model and event-sourced domain model; architectural slices principle

4–12 coarse domain services; pragmatic middle ground; supports ACID

In-memory replicated cache; extreme performance/scalability; high cost