Why CDD?
CDD is not just a pattern — it's a mindset.
It reimagines system architecture from the ground up by placing communication at the center of design.
The Problem with Traditional Architecture
Traditional layered or feature-based architectures often focus on data ownership, object modeling, or component hierarchy.
But in complex systems, these boundaries quickly blur—modules grow tightly coupled, and cross-cutting concerns become hard to manage.
Most importantly, they don't reflect how the system actually behaves.
CDD's Core Idea
CDD starts with the question:
“What are the core messages that flow through this system?”
By identifying what is communicated, who needs to know, and how they react, you begin to uncover the true shape of your system—one built on interactions, not objects.
This leads to:
- 📦 Modular design – Each module is responsible for how it listens and reacts to specific messages.
- 🧭 Predictability – You can trace system behavior just by following the communication flow.
- 🎯 Decoupling by intent – Instead of calling methods or sharing state directly, modules observe and react to meaningful events.
CDD vs Event-Driven Architecture
While similar in surface, CDD is not just event-driven.
| Event-Driven | Communication-Driven (CDD) | |
|---|---|---|
| Focus | Emitting and handling events | Designing around communication patterns |
| Implementation-first | Usually yes | Often no—starts with semantics |
| Communication semantics | Implicit or ad-hoc | Central and explicit |
| Suitable for | Backend, pipelines | Fullstack, cross-role design |
| System understanding | Hard to infer from code | Clear from message contracts |
Why It Matters
Modern applications aren't just built by developers — they're designed by teams, including designers, product managers, and sometimes even AI agents.
CDD enables:
- 🧠 Shared understanding across roles
- 🔌 Pluggable and testable modules
- 📐 Architectures that reflect real system behavior
But more importantly, CDD prepares us for the future.
As collaboration between humans and AI becomes increasingly common, our systems must be interactive and understandable — not just to us, but to machines that work alongside us.
This is the foundation of Vibe Coding:
Programming through shared context, intent, and interaction — across people and intelligent agents.
Communication isn't a side-effect — it's the architecture.