W3C home > Mailing lists > Public > public-grddl-wg@w3.org > December 2006

RE: The Steven King Example becomes the Jimmy Hendricks/MusicBrainz example

From: Dan Connolly <connolly@w3.org>
Date: Fri, 08 Dec 2006 17:45:43 -0600
To: public-grddl-wg <public-grddl-wg@w3.org>
Message-Id: <1165621543.4211.321.camel@dirk>

On Tue, 2006-12-05 at 21:45 -0500, Chimezie Ogbuji wrote:
> On Tue, 5 Dec 2006, Dan Connolly wrote:
> > I'm struggling to figure out a natural place to put that
> > URI in the input document of the corresponding test.
> > http://www.w3.org/2001/sw/grddl-wg/td/titleauthor.html
> 
> Perhaps this is a nitpick, but it seems 
> preferable (to me) to use a BNode to refer to the concept of the song - with a link 
> identifying a web page (READ: *Location*) that the song is the primary 

We talked about this in the telcon; Harry took the ball...

ACTION: HH to respond to Chime, clarifying song/album and discussing
bNode vs URI

I think Chime and I just worked it out in #swig, Harry.

Short version:

<chimezie> if it's only in the test suite, then there really isn't any
issue

more context:

<DanC> Hi Chime. I think we set a record for
actions-processed-per-attendeed in this week's GRDDL WG telcon.
<chimezie> excellent!
* iand waves to chimezie
<DanC> 11 new + 3 cont + 6 done + 3 withdrawn = 23 actions / 5
participants.
<chimezie> wow
<DanC> 4.6 actions/attendee
<chimezie> if only our goverment was as efficient
<DanC> ;-)
<DanC> did you see that the musicbrainz uri is about an album, not a
song?
<DanC> I'm still wondering why, if the publisher gives the album a nice
URI, you'd rather use a bnode?
<chimezie> no i didn't..  did you end up using <link rel='self' to
hardcode the concept uri?
<DanC> I used rel="alternate"
<chimezie> well ...
<DanC> I'm not wild about the HTML encoding, but I'm pretty happy to use
a URI rather than a bnode in the RDF/XML.
<chimezie> just the fact that it's a URI and link semantics *seem*
mostly about locations not identifiers
<DanC> umm... what could I do to get you to drop the whole
location/identifier issue?
<DanC> the IETF and the W3C have dropped it.
--> Yahovah (n=yhwh@c-68-84-255-87.hsd1.fl.comcast.net) has joined #swig
<chimezie> it's mostly a nitpick, like I said.  There just doesn't seem
to be a *natural* way to embed a URI using XHTML  vocabulary.
* chimezie burrows through XHTML specs for anything on rel='alternate'
<DanC> ok... I share the concern about the HTML being a little goofy. I
think the HTML only shows up in the test suite, not the intro to the
spec
* DanC confirms that 'alternate' doesn't occur in
http://www.w3.org/2004/01/rdxh/spec Revision 1.173  2006/12/06 00:05:26
<chimezie> if it's only in the test suite, then there really isn't any
issue
<-- swh has quit ("This computer has gone to sleep")
<-- vant has quit (Read error: 113 (No route to host))
<-- mikeliebhold has quit (Read error: 110 (Connection timed out))
<DanC> logger, pointer?
<logger> See http://chatlogs.planetrdf.com/swig/2006-12-08#T23-44-38


> topic of - than to hardcode a Uniform Resource *Identifier* with a mechanism  (the link 
> element) which is primary used for Uniform Resource *Locators*.
> 
> However, if we insist on using a <link> to refer to the identifier then
> an alternative (used often in atom feeds) worth noting is an empty 
> relative reference with a link[@rel= 'self'] element and an explicit xml:base for the source document:
> 
> <html xmlns="http://www.w3.org/1999/xhtml"
>        xmlns:grddl='http://www.w3.org/2003/g/data-view#'
>        grddl:transformation="glean_title.xsl http://www.w3.org/2001/sw/grddl-wg/td/getAuthor.xsl"
>        xml:base = 'http://musicbrainz.org/mm-2.1/album/6b050dcf-7ab1-456d-9e1b-c3c41c18eed2'>
> 
>    <head>
>      <title>Are You Experienced?</title>
>      <link rel="self" href="" />
>    </head>
> 
> </html>
> 
> > MusicBrains keeps the same information in HTML and RDF,
> > but they don't use content negotiation on the same URI;
> > they use totally distinct URIs. And they don't seem
> > to make any links between the HTML and RDF versions.
> 
> Semantic Media Wiki does the same thing actually: two URI's - one for the 
> 'concept' the other for the article.  However, there is no link in the 
> article to the concept identifier.  There is a link to an RDF export 
> (which includes the URI), however.
> 
> -- Chimezie
-- 
Dan Connolly, W3C http://www.w3.org/People/Connolly/
D3C2 887B 0F92 6005 C541  0875 0F91 96DE 6E52 C29E
Received on Friday, 8 December 2006 23:45:46 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 8 January 2008 14:11:47 GMT