In OPC UA, it is allowed to add additional components to an InstanceDeclaration and to use a subtype for the InstanceDeclaration.

That can lead to the following situation

BType
|

• TextVariable = "Mondo" ("Mondo" is the default value)

CType : BType
|

• StringVariable = "Ciao"

AType
|

• B : BType
  |
  • StringVariable = "Hello"
  • TextVariable = "World"

When I create an instance of AType and use the CType for B, what is the default value?
And what is happening or allowed, when the StringVariable of CType is different from the one of the InstanceDeclaration B (different DataType, etc.)
I have not found any statements in the spec about this case.
My assumption: The default value of the InstanceDeclaration B would apply.
I also would assume, that CType.StringVariable or B.StringVariable could restrict their counterpart (e.g. one uses a subtype of the datatype of the other), and the instance would need to follow both and therefore has to use the subtype.
If both use contradicting types, then the instance of B cannot be of Type CType (even if the StringVariable would be optional in B or C).
Would be good to define this in the specification.
(so far I never had a real scenario where this had happened, but obviously it can be constructed)