ScalaDays 2025 - DDD workshop

Index

• Abstract
• Target Audience
• Prerequisites
• Requirements
• Agenda
• Full-stack DDD App

Abstract

One of the shortcomings of Domain Driven Design (DDD) is that it can be a lengthy process to define the bounded contexts (BCs) properly, and subsequently to transcribe the BCs into a code and to introduce them in your codebase.

This workshop aims to demonstrate how to minimize this inconvenience by relying on:

• The robustness of DDD in Scala
• The fast iteration loop provided by scripting in Scala-CLI
• The ease of transforming a Proof of Concept (PoC) into a Minimum Viable Product (MVP) with the current Scala tooling

You will learn:

• What the building blocks of DDD and Scala 3 are
• Which production-tested libraries help us do DDD in Scala 3
• How to go from a design to a safe and convenient code in a matter of hours

By the end of the workshop, you will have a full stack app with DDD implementations, ready for review.

Target Audience

If you are a Full-Stack Developer, a Software Engineer, a Data Engineer, a Migration Architect or a programming aficionado who would like to acquire, upgrade or master your skills in DDD, this workshop is for you.

Prerequisites

Prior knowledge of Scala is not necessary, although it is a plus – regardless of the flavour of Scala you code in.

Basic understanding of functional programming concepts like map, flapMap, error handling, type classes, etc. is required.

Requirements

Have installed on your computer:

• Scala-CLI (check additional docs here)
• IDE:
  • IntelliJ Idea with Scala plugin
  • VS Code with Metals
  • Or any other IDE that supports Scala

Agenda

Use Case: Spanish ID validator

• Business needs
• Front-end or back-end validation?
• Task requirements

Why DDD? Domain Driven Design premises

• Benefits of making illegal states unrepresentable
• Who are the domain experts?
• What is the ubiquitous language?
• What is a model?
• What is a bounded context?
• Quiz
• Bonus: DDD in the time of LLMs

Hands-on with Scala-CLI - PoC

• Identifying the nouns, behaviours, rules and invariants
• Codifying invariants
• Codifying the rules
• Testing the behaviour

Domain types

• Business Domain
• Types of Subdomains
• Comparing Subdomains
• Bounded Context vs Subdomains
• Quiz

Trade-offs

• Software complexity vs Domain accuracy
• Engineering expertise vs Business needs

Hands-on Full-stack App with Mill

• Basic Implementation - MVP
• Parse, don't validate - Error handling
• Scala 3 libraries to the rescue

DDD building blocks

• Value Objects
• Entities
• Aggregates
• Domain Events
• Domain Services
• Quiz

Hands-on Spanish ID as a Value Object

• Metaprogramming in Scala 3
• Compile-time validation
• Advance usage of Scala 3 libraries

Conclusion

• Trade-offs revisited
  • Software complexity vs Domain accuracy
  • Engineering expertise vs Business needs
• Benefits & Limitations
• Engineering staffing challenges

Full stack DDD ID validator

The app consist of two components that work slightly differently:

1. Front end
2. Back end
3. ID Validator

The front end relies on ScalaJS or JS while the back end is scala on the JVM.

Front end

Tyrian is an Elm-inspired ScalaJS library for frontend and game development.

Website: https://tyrian.indigoengine.io/

Alternative front-end is written using the JS framework react

Back end - Cask

Cask is a lightweight scala library similar to Python's Flask that belongs to the Haoyi Li stack. It spins a server with almost no ceremony, making it a perfect library for prototyping and figuring things out.

Website: https://com-lihaoyi.github.io/cask/

How to run it

Go to the dedicated readme.md files under the following paths:

• Tyrian
• React
• Cask

You need to run the cask server and at least one of the front-ends.