Data Layers need precedence hierarchy defined

[croadfeldt]

In order to combine layers into a singular payload, we need to establish a hierarchy / algorithm / precedence mechanism for resolving data conflicts.

[kcattell]

Is there a need to have a "data definition" associated with layers? To provide the ability to combine, the need to rationalize the "sameness" of data types or instances is needed. Meaning and context [which would also be helpful for the AI examination and use age of a data layer] aids in minimizing confusion between different user or consumer contexts

[croadfeldt]

@kcattell - Agreed. There is a need for consistency and core functionality, possibly requiring prototyping examples of resources and associated layers. In an implementation of DCM, there will definitely need to be a definition of data in advance in order for the various components and parties involved to interoperate, eg; 3rd party layers to be applicable to a service provider layers. The architecture does not specify any data definitions but will need rules at least in order to facilitate the creation of re-usable and composable resource definitions via layers and policies.

An implementation however will need to expand on the rules and define common data types / resource definition templates as well as layer templates.

Current take away: More discussion is needed on the appropriate level of pre-defined data / resources / layers / policies in both the architecture and implementation. The goal being to be flexible and adaptable weighed against ease of deployment and integration.

Associated take away: It makes sense that a standards body could be identified or created in order to facilitate those standard definitions and rules into the core architecture itself, for the purpose of bootstrapping and ease of deployment and integration.