Learn from the past to understand the present
This article distinguishes architecture styles from patterns, explores three primitive patterns, compares technical vs domain partitioning, details the fallacies of distributed computing, and explains how Conway's Law and Team Topologies align team structure with architecture.
This article introduces architecture quantum as the boundary for architectural characteristics, explaining why system-level assumptions fail in microservices. It details four coupling types — semantic, implementation, static, and dynamic — and defines architecture quantum by four properties: independent deployability, high functional cohesion, low external static coupling, and synchronous communication, guiding architectural style selection.
This article details the design evolution of gRPC-Rust's client API, explaining why Tonic's API was replaced, how arena allocation and async builder patterns enable high-performance unary and streaming calls, and why the channel API uses associated types over RPITIT to support flexible interceptors with four distinct variants for metadata access and retry logic.
This article details three measurement approaches for architecture characteristics—operational metrics like p99 latency, structural metrics such as cyclomatic complexity, and process metrics for agility—and shows how fitness functions like ArchUnit automate governance by codifying architectural rules.
This article explains three sources of architectural characteristics—domain concerns, project requirements, and implicit domain knowledge—and outlines a three-step process to extract explicit characteristics from requirements, supplement implicit ones from domain knowledge, and collaboratively select the top three driving characteristics that guide architectural decisions.
This article categorizes architectural characteristics (non-functional requirements) into four groups—operational, structural, cloud-native, and cross-cutting—detailing key attributes like availability, scalability, extensibility, security, and legal compliance to help architects build a systematic understanding of quality attributes that drive architectural decisions.
This article explains that architecture and design form a continuous spectrum distinguished by three dimensions—significance of trade-offs, investment level, and strategic vs tactical scope—helping engineers clarify decision impact and make wiser technical choices.
This article introduces software architecture through its etymology and the software crisis, then defines four interconnected dimensions—architecture characteristics, logical components, architecture style, and architecture decisions—that together form a complete architectural picture, using examples like database selection and e-commerce components to illustrate trade-offs.
This article details Spring AI's observability architecture built on Micrometer Observation, covering seven built-in observation points for models, vector stores, and chat clients, the four-component collaboration pattern, and OpenTelemetry integration via micrometer-tracing-bridge-otel with Maven and configuration examples.
This article analyzes OpenAI's Harness Engineering blog post, which advocates rigorous system design—context engineering, architecture constraints via CI/CD gates, and entropy combat—over reliance on better models to solve AI code maintainability, while noting high implementation costs and overlooked functional verification, contrasting with Anthropic's generator-evaluator approach.
