(Feedback from Canada)

I like the way Sentinel  and Substantive Value Domains were included in the model! Having said that, I wonder why the takes value from association was restricted in both Substantive and Sentinel Value Domains to 0..1 at the Enumerated and Described Value Domain ends. Why not allowing a Sentinel/Substantive Value Domains to take values from multiple Described/Enumerated Value Domains? Value Domains are groupings of representations of Categories associated to some Conceptual Domain. Now that we have Substantive/Sentinel Value Domains connected to the Represented Variables, there is no reason (that I can see) for not allowing multiple Enumerated/Described Value Domains, when necessary, created just for management purposes. We could use Described and Enumerated Value Domains for managing representations more freely by allowing a Substantive Value Domain to take value from multiple Enumerated or Described Value Domains. Just a thought. 

  • No labels

3 Comments

  1. Jenny Linnerud

    22 Jan, 2019

    Given that the Enumerated Value Domain itself takes values from 1...* Codelists, 1 ..* Statistical Classifications do we still wish/need to change 0...1 to 0..*?

    Similiarly the Described Value Domain takes values from 0...* Substantive Value Domains.

  2. Flavio Rizzolo

    22 Jan, 2019

    I see your point re. Enumerated Value Domains, we could use Codelists and Statistical Classifications as management structures. However, the Described Value Domains doesn't take values from Substantive Value Domain, it's the other way around.

    Writing this answer I noticed that now Codelists and Statistical Classifications are required for defining an Enumerated Value Domain. That can't be right. First of all, because you could have Enumerated Value Domains without the support of either Codelists or Classifications. And second of all, because you can't require both a Codelist and a Classification, it's usually one or the other. I propose to change those cardinalities to 0..*


  3. Flavio Rizzolo

    24 Jan, 2019

    We propose to change the cardinalities of the "takes value from" relationships ending at Enumerated and Described Value domains from 0..1 to 0..*

    We propose to change the cardinalities of the "takes value from" relationships ending at Code List and Statistical Classification from 1..* to 0..*