Include OWL files. Possibly fixes #102

[cmungall]

See

• Review JSON-LD and RDF serialisations #102

Note if we merge this and make a w3id PURL for it we should probably include more inline caveats, see:

https://linkml.io/linkml/generators/owl.html

I think many in the OBO community would be confused by this "ontology" of SSSOM

[afk314]

Nice. One more request, can we not duplicate definitions of properties that don't belong to the SSSOM namespaces? In this file I see resources from DCT, OWL, PROV, and RDF-S merged in. I think it would be cleaner if these were imported. I'm happy to do this and commit but not certain if this is being generated from somewhere else.

[matentzn]

Hm, this is a very complex request, as all of these files are derived files using a framework entirely independent from SSSOM - (https://linkml.io/). If you wanted to articulate the need of separating imported entities more cleanly from internal ones, you should do it on their repo.

However, I doubt you will have much success - we have been pushing front and center to get rid of owl:imports, due to its weaknesses when it comes to tooling (network dependencies, special tools needed to load) and versioning (our imports change frequently, and it has become impossible to figure out exactly how "an ontology was at timepoint x". As this is very far beyond this issue here, you can, if you want, see this discussion here for further details on that matter:

INCATools/ontology-development-kit#194

[afk314]

OK, I understand. I suppose I haven't thought deeply enough about the topic to outright disagree with that! They are a problem to manage but I don't rely on dereferenced ontologies ever and I guess just accept all the pain. Thanks for the answer.