search

SED-ML / sed-ml

Simulation Experiment Description Markup Language (SED-ML)
http://sed-ml.org
5 stars 2 forks source link

Clarify use of anyURI vs URN for model languages and data formats #181

Closed jonrkarr closed 3 years ago

jonrkarr commented 3 years ago

2.1.1.8 says model languages and data formats should be specified by anyURIs. Should these have type URN?

luciansmith commented 3 years ago

Looking things up, it looks like the answer is no--URI is the generic term, and while we do constrain ourselves to the 'urn:...' format for some things (like the 'language' attribute') we don't use that constraint for other things. We could theoretically explicitly limit the language attribute specifically to URNs, but since we're actually more constrained than that, I don't think we need to have a new 'URN' type.

luciansmith commented 3 years ago

https://danielmiessler.com/study/difference-between-uri-url/

jonrkarr commented 3 years ago

Consistent values of these attributes is important for portability of SED-ML among tools. Because URNs are a special case of URIs (a regular expression pattern), this additional constraint is noted in the text. The schema could also reflect this additional constraint. This would in turn clarify the UML diagrams (which are derived from the schema) and help programs that interpret the schema validate documents.

luciansmith commented 3 years ago

Done! I hadn't quite understood the request originally; I assumed you wanted anyURI replaced wholesale with URN. It's now required to be a URN for model/language and dataDescription/format.