dvent 0.1.0

Simple, immutable, functional models for domain driven design with event sourcing

Dvent is a minimal set of simple, immutable, and functional models and tools intended to support Domain Driven Design (DDD) and function as the source-of-truth data model in a CQRS / Event sourcing architecture pattern.

Installation

pip install dvent

Dvent is

A library

A set of tools from which to build rich solutions, not a solution which you may easily apply to an existing problem. It aims to guide your development of models that are robust, simple, scalable, easy to reason about, and easy to test. While opinionated about the manner in which you model event orchestration it is still up to you to implement that model.

Minimal

Only include what is necessary and eschew that which is not; allow for easy expansion and composition

Simple

Aim for simplicity over ease of use; this generally manifests in very explicit code with consistent naming, loose-coupling, composition, and no “magic”

Immutable

With help from the excellent pyrsistent library all data structures and patterns follow and reinforce immutable best-practices; so all state transitions require reassignment and all entities (Events, Commands, Aggregates) are hashable.

Functional

Prefer functional patterns to classic OOP while retaining a style that is approachable for developers familiar with more classic Python idioms.

Fully Specified and Tested

Dvent is developed with DDD itself and features living documentation in the form of Gherkin features. Long-term plans include implementation of multiple languages against the same specification.

Beta Software

While Dvent is currently in use in an enterprise-level setting, it is still provided as Beta software until further scaling and consistency guarantees are tested.

Dvent is not

A framework

Dvent implements no “real-world” or practical data persistence and because it knows nothing about your problem domain it does not attempt to provide a solution. Instead it aims to guide your development of models that are robust, simple, scalable, easy to reason about, and easy to test. For something more fully-featured and “out-of-the-box” take a look at the excellently documented eventsourcing project which implements many of the same (and more) patterns.

Usage

Coming Soon - for now check out the sample notebook(s)!

Exploring

Dvent includes integration with the Jupyter project via a docker image. To use the notebook(s) you’ll need docker version 17.09.0+ and can simply run

DVENT_JUPYTER_TYOKEN=<password-or-token-value> docker-compose up -d dvent-notebook

Note: this will download and build the images which may require a good internet connection

Then in your browser visit your personal notebook, enter your password/token from above, browse to the “notebooks” folder, and open any you like.

Testing

All tests are currently done via behave and gherkin feature files. To run the test suite you can use docker via docker-compose run --rm dvent-test

Why make “Dvent”?

I was leading a team at Discogs building a new “greenfield” project which needed a basic set of event-sourcing models which were immutable, unopinionated about implementation details, and enabled basic functional patterns. I couldn’t find anything that quite fit the need, so I made one over an intense weekend and we’ve been using it without any major modifications for over 15 months as of release 0.1.0 (May 2018).

It’s very plausible that without Discogs I would never have written Dvent and even so not considered it as vialble enterprise-level software. I consider it the birthplace and indirect sponsor of the project.

Why the name “Dvent”?

An extreme portmanteau of:

• Domain Driven Design

• Discogs

• Event

Obilgatory naming things is hard reference ;)

Special Thanks

• Discogs and my team members (you know who you are)!

• The excellent pyrsistent library which makes working with data structures in Python almost as joyful as in Clojure

• Greg Young for his many helpful talks, posts, etc. which inform much of this library’s patterns

• Rich Hickey and the broader Clojure community for the inspiration in building practical, immutable, functional solutions