You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
If we consider the combination of the {inmmediate_parent_name}.{branch_name} (e.g., Door.Window instead of the full Vehicle.Cabin.Door.Window), we still have duplicate entries:
I think it would be a smart move to correct that in the specification. More concretely:
If duplicate branches convey the exact same intended meaning, then merge them.
if duplicate branches have different intended meaning, then rename them to be more specific and to distinguish them from the others.
Add a rule that makes it mandatory for branches to have a unique name.
As you can see, there are not too many duplicates, and the proposed change is feasible.
We can start by fixing the uniqueness in the combination {inmmediate_parent_name}.{branch_name}. Once this is resolved, we can address the uniqueness of the branch name alone.
One big advantage I see is that we can decouple the dependencies of the nodes from the whole fqn path and the concepts become easier to maintain in the long run as: FeatureOfInterest and Property. For example, modeling the window position, will be about modeling the property position of the feature of interest Window.
For example, in GraphQL, types, enums, scalars are unique within the schema. This uniqueness makes it really easy to restructure the hierarchies and define mappings to other languages.
The text was updated successfully, but these errors were encountered:
During the refactoring of the GraphQL exporter, I noticed there are repeated names for branches in the specification.
Branch name only
For example, if we look at the branch name without the context path (i.e.,
Window
vs.Vehicle.Cabin.Door.Window
), we have the following duplicates:Immediate parent name + Branch name
If we consider the combination of the
{inmmediate_parent_name}.{branch_name}
(e.g.,Door.Window
instead of the fullVehicle.Cabin.Door.Window
), we still have duplicate entries:Proposal
I think it would be a smart move to correct that in the specification. More concretely:
As you can see, there are not too many duplicates, and the proposed change is feasible.
We can start by fixing the uniqueness in the combination
{inmmediate_parent_name}.{branch_name}
. Once this is resolved, we can address the uniqueness of thebranch
name alone.One big advantage I see is that we can decouple the dependencies of the nodes from the whole
fqn
path and the concepts become easier to maintain in the long run as:FeatureOfInterest
andProperty
. For example, modeling the window position, will be about modeling the propertyposition
of the feature of interestWindow
.For example, in GraphQL,
types
,enums
,scalars
are unique within the schema. This uniqueness makes it really easy to restructure the hierarchies and define mappings to other languages.The text was updated successfully, but these errors were encountered: