- From: Bernard Vatant <bernard.vatant@mondeca.com>
- Date: Mon, 3 Dec 2012 10:49:26 +0100
- To: public-vocabs@w3.org
- Message-ID: <CAK4ZFVEsTQkGB4+JstC_eHRG9OyWUMTCCYAUDuWrK-BjAjyj0Q@mail.gmail.com>
Hi Dan Great to see the things moving forward to clarify those points we have discussed for quite a while :) Semantics of domainIncludes and rangeIncludes are quite clearly explained "for humans" in the wiki page, but seems to me some formal clarification should be added, even if it's lighter than RDFS. 1. Subsumption and inheritance : Let's assume the hierarchy of schema.orgtypes/classes will continue to use rdfs:subClassOf until some logic guru forbids it. This entails that schema.org classes are RDFS classes by RDFS declared semantics. Are domainIncludes and rangeIncludes inherited by subclasses, IOW do the following rules hold ? Seems implicit in the structure of schema.org pages. (p schema:domainIncludes D1) AND (D2 rdfs:subClassOf D1) => ((p schema:domainIncludes D2) (p schema:rangeIncludes R1) AND (R2 rdfs:subClassOf R1) => (p schema:rangeIncludes R2) 2. In the current state of schema.org pages, the rangeIncludes assertions are declared for each class using the property (IOW for value of domainIncludes), but actually the values of rangeIncludes are global if we trust the RDFa declarations, which means they are the same for all classes in the domainIncludes, there is no such thing as local restriction for rangeIncludes. And it seems that in the pages If those rules hold, it would be good to see them expressed formally somewhere. And BTW when do you think the RDFa page is about to use the new property names? Bernard 2012/11/30 Dan Brickley <danbri@google.com> > Hi folks > > 3. Class/Property > > There are several cases (including the above-mentioned Actions draft) > where it is useful within schema.org to have a first class type > representing the notion of 'Class', and of 'Property'. This is rather > meta and while it is not something designed for mainstream webmasters > to encounter, it will help with structuring and documenting the > vocabulary. > > I have written up a proposal for adding these (and aliasing them to > rdfs:Class, rdf:Property) at > http://www.w3.org/wiki/WebSchemas/SchemaDotOrgMetaSchema ...alongside > a proposal to use schema.org/domainIncludes and > schema.org/rangeIncludes in our RDFa representation of the schema. > > > Comments on any / all of the above are welcomed; ideally in the > WebSchemas area of the W3C wiki or here on public-vocabs. If you reply > by mail please adjust the Subject line to match your topic... > > cheers, > > Dan > > -- *Bernard Vatant * Vocabularies & Data Engineering Tel : + 33 (0)9 71 48 84 59 Skype : bernard.vatant Blog : the wheel and the hub <http://blog.hubjects.com/> -------------------------------------------------------- *Mondeca** ** * 3 cité Nollez 75018 Paris, France www.mondeca.com Follow us on Twitter : @mondecanews <http://twitter.com/#%21/mondecanews>
Received on Monday, 3 December 2012 09:50:21 UTC