The API provides for returning lists of the following:
Needs
Circumstances
Service Types
E.g. https://host:port/o/ServiceDirectoryService/v2/taxonomies/?vocabulary=esdServiceTypes
There seems to be some ambiguity as to whether these taxonomy queries should return ALL values available (based on standard ESD data & mappings) or simply those values as actually used within a directory (in other words, only values assigned to a live service found in the directory would be returned).
Anybody have any guidance/thoughts/opinions on this?
This will depend on which side you are coming from i.e. collecting/maintaining data or consuming it.
Remember that the ORUK dataset only includes service-type. It is the LGA taxonomy model that links Service-type to needs and circumstances.
If you are an application collecting service data and will want to know which service-type to apply then you will need to pull all service-types in and use some logic to make it easier for the end-user to identify the correct service-type. There is no reason for the collecting application to pull in needs and circumstances.
If you are a consuming application then you are only ever going to request the service-types that relate to the needs and circumstances that you offer as a means to find services. You might pull all needs and circumstances in and offer everything to the end-user or you may choose to only pull in a subset which are appropriate to the context of the end-user (This is what Lancashire & South Cumbria do) The difficulty here is that you might have needs and circumstances that won’t return any service offerings. You need to trust/understand the model which links needs and circumstances to the service-types which then will pull the services in.