ryanfranz
commented
3 years ago The category and subcategory ideas are only used in creating directories within the GTLibrary (geotypical model library). As long as there is a replacement or rework of the directory structure, then these definitions are not needed.
Currently in the CDB 1.1 feature dictionary XML, features are organized into groupings of category and subcategory, which are primarily used to generate the first and second character of the 5-letter FACC-style code for feature types.
Aside from being used to define these 5-letter codes, these two sets of groupings appear to be unused in actual CDB vector content: all feature and attribute organization and constraints are done by dataset, and as such dataset seems to be the most appropriate definition to map to the proposed "Theme" concept.
Is there a reason to preserve category and subcategory as independent definitions in the metadata? These groupings have long been obsolete and replaced in upstream data models such as GGDM and NAS, which use a different set of thematic groupings entirely that are only loosely related to the 5-letter codes.
In the absence of any such requirement, I propose omitting these definitions in CDB X and using the "Dataset" definitions to populate the Theme table.