Avoid the use of `sh:IRI` for contact point

[tpluscode]

The constraints on schema:contactPoint require the value to be an IRI

cube-link/validation/standalone-cube-constraint.ttl

Lines 46 to 49 in 2b69ad5

	sh:path schema:contactPoint ;
	sh:minCount 1 ;
	sh:nodeKind sh:IRI ;
	sh:message "cube:Cube needs an schema:contactpoint"

Some other properties too, are constrained like that. I propose to relax this to allow blank nodes. Possibly as an alternative, where it's either IRI or a blank with deep description using sh:node. For the standalone cube I expect schema:name and schema:email at the least

[giacomociti]

why would you require name and email only when the contact point is blank?

[tpluscode]

I don't understand. What else do you think is necessary?

[giacomociti]

my understanding is that name and email are required on a contact point only when the contact point is blank. Why they are not needed also when the contact point is an IRI?

[tpluscode]

Ah, gotcha. I somehow limited the logic to IRI meaning remote resource.

[giacomociti]

now I get the rationale and it makes sense to me. In a sense, choosing blanks is often a way to express that the data 'belongs' to the referencing object (as in concise bounded descriptions which include blanks and instead stops at IRIs)

[tpluscode]

TBH, the case of contact point may not even make sense as a remote resource in the first place but let's not got there now