Re: ORG: proposed Last Call draft for review

One more editorial change:

The usage note for org:Role states "The normal SKOS lexical properties should be used when labelling the Role." However, the two examples of org:Role use rdfs:label to label the Role. Since org:Role subclasses skos:Concept, and given the usage note above, I figure the examples should use skos:prefLabel.

On 2012-10-12, at 5:19 AM, Dave Reynolds wrote:

> Hi James,
> 
> On 12/10/12 04:28, James McKinney wrote:
>> Returning to the original subject of this thread, here are a few small things I noticed in http://dvcs.w3.org/hg/gld/raw-file/default/org/index.html I don't know if there's a better way for me to report these proposed changes. If there is, please point me in the right direction. Lots of small fixes and a few bigger questions.
> 
> Thanks. I'll look at the editorial changes which I get a chance, hopefully before Thursday.
> 
> Responses to the non-editorial questions in line ...
> 
>> Is the diagram meant to include all classes and properties?
> 
> No. A complete diagram gets to be too messy to be useful. This is intended to guide people in understanding and using the main parts of of the ontology.
> 
>> FYI, there is no inverse for org:organization, org:role or org:basedAt. Other similar properties have inverses. I'm not a fan of inverse relations, just pointing it out.
> 
> True. Mostly those are covered in the style note where it says "omitting attribute-like relations" - that doesn't really explain org:basedAt but whatever :)
> 
>> "Indicates a VCard (using the http://www.w3.org/TR/vcard-rdf/ vocabulary)"
>> Use the "[VCARD]" notation.
>> 
>> Why do basedAt and location have a domain of foaf:Person? We only ever use foaf:Agent everywhere else.
> 
> The thinking there was that only people have physical locations like that. Non-person-agents are things like committees which don't have a location.
> 
> That could be generalized without harm since it wouldn't break any deployed usage if people feel that's a limitation that should be addressed.
> 
>> I'm a little confused by the org:location property. Why do we need it? It's only mentioned once in the document (for its definition).
> 
> It was included because one of the use cases was to represent typical company organization charts and those often include an internal location within a building such as a room number. For example, in HP there was the notion of a "mailstop" which was essentially a coordinate for a cubicle within a big building given by nearest pillar on a square grid.  However, that's not part of the delivery address and so doesn't fit on the vCard, its purely internal to help you physically find the person.
> 
> I don't think it is *needed* in any strong sense and certainly isn't a big feature, which is why it doesn't have any further discussion. However, it does no harm and can be useful for at least some situations.
> 
>> General thought: Is it a problem that org:Organization subclasses foaf:Agent? Doesn't this mean that an organization (which is a foaf:Agent) can be a member of an organization?
> 
> That's not a problem but a requirement :)
> 
> For example an org:OrganizationalCollaboration is an organization whose members are other organizations.
> 
> Within UK government there are cases where a Post that you think of as being held by one person is in fact held by a committee or other collective which act as if they were one agent. So in that case you need the committee to appear as a member of the Organization in which the Post exists.
> 
> Dave
> 
> 

Received on Sunday, 14 October 2012 17:46:23 UTC