Description
A lack of semantically consistent versions of EA models (e.g. v1.1, v1.2, etc.) or poor versioning in general.
Context
Derived from: “API Versioning”
The version is the first hint at different behaviors or changed functionality.
Consequences
In case of updates to non-semantically versioned models, users may face breaking changes. For example, the structure is changed fundamentally, such that the majority of procedures has to be performed in a new way.
Solution
EA models need to be semantically versioned to allow easier communication and to decide whether the communication needs to be adapted to a new contract.
Example
release v1 of a model, and work on v3 afterwards
Tags
microservices, soa, business, application, technology, ambiguity
Related smells
- Deficient Names (relates)