I think that obscuring the person in a blank node in a resource that
abstracts its way around the person is Herbert Van de Sompel's "one
bridge too far". http://twitter.com/hvdsomp/status/15777598515
People want to talk about people, not abstractions of people.
On Tue, Jun 15, 2010 at 10:10 AM, Karen Coyle <[log in to unmask]> wrote:
> In my (librarian-ish) mind, the confusion goes away if we state that we have
> resource that is ABOUT a person/book/subject. We then describe the resource
> and its topic, something like:
> <rdf:Description rdf:about="http://openlibrary.org/authors/OL20A">
> <foaf:name>Michael Madhusudan Dutt</foaf:name>
> <rdfs:label>Dutt at Poet Seers</rdfs:label>
> In this case, adding dcterms:modified at the rdf:Description level isn't a
> It seems to me that the problem comes about when trying to assert that the
> URI represents the person that is the topic of the page, rather than
> accepting (maybe that's the wrong word) that what is on the web is a
> representation, one step away from the real world, and that we are
> manipulating representations that MAY stand for real world things. There
> will be many such representations for any truly real thing, and we'll be
> spending the rest of our time determining which ones assert information
> about the same thing. :-)
> Quoting Jonathan Rochkind <[log in to unmask]>:
>> Um, but now you're ignoring the original question, which was about
>> keeping assertions about the _administrative metadata record_ seperate
>> from assertions about the _actual thing_. ('actual thing' being a
>> book, or an author in this case).
>> Which is the URI http://openlibrary.org/authors/OL2632437A for then,
>> under the below plan?
>> At first Ed suggested that
>> http://openlibrary.org/authors/OL2632437A.rdf woudl be for the metadata
>> and http://openlibrary.org/authors/OL2632437A would be for the book.
>> Then I brought up that ".rdf" is just one representation of the
>> metadata record, is that an identifier for just the RDF representation,
>> or is it an identifier for the "metadata record"?
>> And now we've come around to saying, oh, well, yeah, then just use
>> http://openlibrary.org/authors/OL2632437A . But then we're back to
>> the confusion over whether assertions involving that URI are for the
>> actual book (or the actual author/person in this case), or for the
>> metadata record about him/her/it.
>> My conclusion: There is no obvious answer here, probably no matter
>> what OL/Karen chooses to do, it's going to confuse some people.
>>> On Tue, Jun 15, 2010 at 4:24 AM, Felix Ostrowski <[log in to unmask]>
>>>> http://openlibrary.org/authors/OL2632437A/about could be used as
>>>> the generic metadata URI serving the representation that is requested
>>>> via content negotiation. Assertions about the metadata can then be made
>>>> using that URI:
>>>> @prefix dc: <http://purl.org/dc/terms/> .
>>>> @prefix foaf: <http://xmlns.com/foaf/0.1/> .
>>>> dc:modified "2010-06-15" ;
>>>> foaf:primaryTopic <http://openlibrary.org/authors/OL2632437A> .
>>> This is an excellent suggestion. And I don't think it necessary
>>> precludes also making a json document available at a non-content
>>> negotiated URL. Having URLs for the specific format is often handy so
>>> that people can look at them in their browser without having to
>>> twiddle with Accept headers.
> Karen Coyle
> [log in to unmask] http://kcoyle.net
> ph: 1-510-540-7596
> m: 1-510-435-8234
> skype: kcoylenet