I would like to get you comments on product schema modelling.
Let's assume I have a very simple product definition with 3 properties of interest
a,
b and
c. Properties
a and
b are interesting in buying context whereas property
c is important to client but only for operational purposes (e.g. it is an IP address that was assigned during provisioning).
I imagine 3 alternative approaches to model that:
- There is a single ProductSpecification for which 3 characteristics are defined. I am using this product specification as a reference in orders (TMF 622) and in inventory (TMF637).
- There are two interrelated product specifications. One referenced for ordering (contains only a and b), second used in inventory, Second ones contains all 3 properties (or maybe I am not even including property b as it is not relevant after order - e.g. it is only used to infer value of c).
- There is single ProductSpecification which is referenced from orders and inventory. However it only models a and b.
For me for a very long time the only viable option was 1, but now I am not sure. To build a more realistic example lets imagine we do not deal with 3 properties but 10s or 100s where significant number of these are operational. And within this context here are my questions:
1. Is approach 2 a valid one?
2. When using approach 1 how can I indicate that a given property is operational one? In other words, how to mark these not relevant during pre-order phases?
3. How can you learn the full data model ("schema") for operational representation of the product in approach 3?
Thank you in advance for all comments / hints.
Bartosz Michalik
------------------------------
Bartosz Michalik
Amartus
------------------------------