Config-as-architecture
When XML, YAML, and low-code graphs replace design decisions — until the config becomes the unmaintainable system.
Config-as-architecture is the enterprise tradition of hiding design in files too large to code-review. Spring XML, ESB routes, Terraform stacks, BPMN diagrams, low-code graphs — all can be legitimate tools, and all can become a shadow codebase owned by nobody who understands git blame.
The pattern thrives where buying is easier than hiring. A consultant can deliver ten thousand lines of XML that "wire services together" faster than a team can write explicit modules with tests. For a while, change requests flow through an integration team that speaks WSDL. Then the original authors leave, the tooling rots, and every change is archaeology with an IDE that validates syntax but not intent.
UML-everywhere and waterfall-big-design were cousins of this failure: diagrams and documents treated as the product instead of running software. DevOps partially fixed packaging but sometimes recreated the problem in YAML — infrastructure as architecture theater without service ownership.
Healthy config is thin: it declares environment differences and leaves business logic in code with tests. Unhealthy config is where the if-statements went to retire — and where outages hide until someone toggles a flag nobody documented.
Exhibits
Practice · 1998–2014
XML-as-architecture
When every wiring decision lived in 4,000-line XML files. Configuration became a second programming language nobody wanted.
Practice · 1997–2012
UML Everywhere
Diagrams as deliverables. Useful as a sketch tool; catastrophic when treated as source of truth.
Framework · 1999–now
J2EE / EE frameworks
EJB-heavy stacks taught a generation that XML config was architecture. Spring and lighter runtimes absorbed the survivors.
Framework · 2003–now
Spring
Java's practical operating system for apps. DI and Boot made enterprise Java shippable again.
Platform · 2014–now
Terraform
Declarative infra with state files everyone learns to fear. The closest thing to versioned datacenters — then BSL and OpenTofu.
Practice · 2018–2025
Low-code will replace eng
Drag-and-drop apps for the enterprise backlog. Useful for internal tools; fantasy as a full replacement for engineering — a pitch later absorbed by AI demos.
Architecture · 2003–2015
Enterprise SOA
Reuse via shared services — in theory. In practice: central buses and political ownership fights.
$ ESBs, governance boards, and "service portfolios" burned seven-figure programs. Many orgs got distributed monoliths with worse latency and a permanent integration team.
Platform · 2011–2023
Jenkins
Plugin palace CI that defined pipeline culture — and Groovy nightmares. Still running somewhere in your org; greenfield moved on.