Separation of localization models


Dynamics 365 for Finance and Operations is now being licensed as Dynamics 365 Finance and Dynamics 365 Supply Chain Management. For more information about these licensing changes, see Dynamics 365 Licensing Update.

As part of the requirements for LCS solutions for localization and translation, if a localization solution for previous versions of Dynamics 365 Finance and Operations apps contains both regulatory and competitive features, localization ISV solution providers must split the solution into separate models for each feature type. This article provides information about this requirement.

Multinational customers must comply with regulatory requirements in all countries/regions where they deploy Finance and Operations apps. At the same time, these customers want to minimize the cost of code maintenance. Therefore, if a localization solution for previous versions contains both regulatory and competitive features, the solution must be split into separate models, so that customers can adopt and deploy the features that they require. An effort has been made to split the Application foundation and Application suite stack into multiple models.

The number of models is expected to grow over time. Splitting a monolithic code base provides many benefits, such as better scalability, manageability, and serviceability. The localization requirement to split a localization solution into more granular models builds on this effort. The goal is to provide the same benefits to multinational customers.

After you've classified features as either regulatory or competitive, as described in Classification of localization features, split the code for these features into at least two models, one model for the regulatory features and at least one model for the competitive features. If the competitive features can be split further (for example, into features that are related to Retail and features that are related to Fixed assets), it's a good idea to split them. However, further splitting isn't mandatory. For more information about how to split the stack into multiple models, see Model split.