[oslc-core] Comment on Vocabulary Description Vocabulary

[Arthur Ryman]

Ian,

rdfs:label is supposed to be a human-readable string that you could 
substitute for the URI. You might use it to label fields in a form, 
columns in a table, or nodes and arcs in a graph. 

The oslc:inverseLabel property was motivated by the use case of a query 
builder that navigated from resource to resource. Hence, we need a label 
that flips the sense of rdfs:label. There was no discussion of roles.

I hope that people are not writing queries that depend on certain values 
of rdfs:label, and that these values are retrieved "at the last minute" 
when presenting data. This would let us improve the labels without 
breaking anyone.

Regards, 
___________________________________________________________________________ 

Arthur Ryman 

DE, Chief Architect, Reporting &
Portfolio and Strategy Management
IBM Software, Rational 

Toronto Lab | +1-905-413-3077 (office) | +1-416-939-5063 (mobile) 





From:   Ian Green1 <ian.green at uk.ibm.com>
To:     oslc-core at open-services.net, 
Date:   11/29/2013 08:18 AM
Subject:        [oslc-core] Comment on Vocabulary Description Vocabulary
Sent by:        "Oslc-Core" <oslc-core-bounces at open-services.net>



The annotation vocabulary [1] defines oslc:inverseLabel and gives some 
examples of its use. 

It seems to me that rdfs:label and oslc:inverseLabel ought not to be 
duals.  The RDFS spec has R rdfs:label L to mean "L is a human readable 
label for R".  I don't think our usage of rdfs:label reflects this 
intended usage - rather we're using the rdfs:label of the predicate to 
designate the role the subject plays in relation to the object of the 
link.  The idea of oslc:inverseLabel is to have something to designate the 
role the object plays in relation to the subject. 

The example on the wiki is to do with query building.  The suggestion is 
the user would see "validates" in the UI from the perspective of the 
testcase, and they would see "validated by" from the perspective of the 
requirement.  But in neither of these cases is the user examining the name 
of the resource whose URI is oslc_rm:validatedBy.  That resource has the 
label "validatedBy".  The query builder use case is about labelling the 
role of the testcase in relation to the requirement (or vice versa).  This 
is conceptually different from the human-readable name of the predicate. 

Was this distinction discussed when the draft was being drawn up 
(apologies if so)?  Does this distinction matter? 

I looked through the RM 2.0 vocabulary and it is the case that all of the 
rdfs:labels "make sense" as a role name;  I looked at DC and most of those 
also follow that pattern, but not all do.  For example, dcterms:relation 
has the label "Relation", with the definition   "A related resource". I'd 
expect the role name for such a link to be "Related To". 

Any comments? 


[1] http://open-services.net/wiki/core/Vocabulary-Annotation-Vocabulary/ 

best wishes,
   -ian

ian.green at uk.ibm.com (Ian Green1/UK/IBM at IBMGB)
IBM Rational
Unless stated otherwise above:
IBM United Kingdom Limited - Registered in England and Wales with number 
741598. 
Registered office: PO Box 41, North Harbour, Portsmouth, Hampshire PO6 3AU
_______________________________________________
Oslc-Core mailing list
Oslc-Core at open-services.net
http://open-services.net/mailman/listinfo/oslc-core_open-services.net