As is generally recognized, developing dynamic component models with multiple levels of subcomponents can produce models that break easily when you later try to change one of their components. Does anyone have a set of rules to follow in the construction of multi-level DC models that will help us to avoid such breakages?
From my limited experience, the most critical formulas are those that involve position attributes. Here are some things that may influence the stability of a component under change:
-
How are the axes of each subcomponent located and oriented?
-
Are the various attributes referenced in a formula constrained or not?
-
Is the sub-component unique or not?
-
Are the parents of the subcomponent unique?