W3C home > Mailing lists > Public > public-rdf-in-xhtml-tf@w3.org > August 2007

Re: title element, meta, h1 and rdfa

From: Ben Adida <ben@adida.net>
Date: Fri, 31 Aug 2007 15:25:29 -0700
Message-ID: <46D89559.8050102@adida.net>
To: Mark Birbeck <mark.birbeck@formsPlayer.com>
CC: Karl Dubost <karl@w3.org>, RDFa <public-rdf-in-xhtml-tf@w3.org>


I concur with Mark that priorities are prone to too much complexity: if
there are duplicate triples, then there are duplicate triples, it's not
the end of the world and it wouldn't be the first time :)

I'm loathe to start interpreting XHTML as RDFa in most cases, but I can
see the reason for <title>. That said, I don't see how we can make the
executive decision, in this task force, to map to dc:title. At best,
xh:title, as Mark mentions.

Oh and one more thing about Karl's markup: those should be plain
literals, not XMLLiterals. Our decision a few months ago was that if
there is no markup within the content, then it is a plain literal.

-Ben

Mark Birbeck wrote:
> Hi Karl,
> 
> I don't recall this being discussed in the task force, so it would be
> interesting to hear other views. However, it *was* discussed in the
> XHTML 2 Working Group when we first started to work on RDFa, and we
> concluded:
> 
>  * that you can't really have priorities for the triples, since that
> makes parsing very complex. Any
>    'priorities' or overlaid meanings need to come from an
> interpretation of the derived triples, and
>    not by trying to understand the triples in the parsing stage;
> 
>  * as with your comment, we also thought that <title> should express dc:title.
> 
> The h:title => dc:title part is a little controversial, and in
> particular it does favour one taxonomy over others. As a consequence
> it hasn't been introduced into the XHTML+RDFa 'dialect' that is being
> worked on in the task-force. However, seeing your examples, I'm now
> wondering whether having this 'shorthand' for dc:title is a good idea
> at all!
> 
> What we could do instead is to say that, yes, <title> is a shorthand,
> but that it generates a predicate in the XHTML namespace, rather than
> Dublin Core's:
> 
>   <> xh:title "The Old Man And The Sea - Literature" .
> 
> (Note that if there is no mark-up in the element content, we don't
> default to XML literals.)
> 
> With this triple it would be a simple matter to establish precedence
> when presented with the following triples from your example:
> 
>   <> dc:title "The Old Man And The Sea, Ernest Hemingway" .
>   <> dc:title "The Old Man And The Sea, Hemingway" .
>   <> xh:title "The Old Man And The Sea - Literature" .
> 
> The key point is that the precedence is worked out at the level of the
> generated triples, and not whilst in the process of parsing the
> initial mark-up, which is bound to be better.
> 
> Note by the way that strictly speaking the <title> is the title of the
> information resource sitting at the current URL, and therefore can't
> be the title of a book. So in reality this problem of 'priority'
> arises in situations like news stories, where the article title is in
> the first <h1> as well as in <title>. The odd thing here though is
> that you might actually want <h1> to take precedence.
> 
> The situation that would seem to be the most common would be something
> like the following, from today's Times:
> 
>   <html>
>     <head>
>       <title>
>         Empty seats at service show scars of Diana&rsquo;s life have
> not all healed
>       </title>
>     </head>
>     <body>
>       <h1 class="heading">
>         Empty seats at service show scars of Diana&rsquo;s life have
> not all healed
>       </h1>
>       ...
>     </body>
>   </html>
> 
> Reading @class="heading" as @property="dc:title", we see that the two
> titles are exactly the same.
> 
> However, The Sun, has the story like this:
> 
>   <html>
>     <head>
>       <title>
>         The Sun Online - Royals: Diana tribute gets underway
>       </title>
>     </head>
>     <body>
>       <h1 class="black32">
>         Diana tribute gets underway
>       </h1>
>       ...
>     </body>
>   </html>
> 
> It's quite a reasonable use of <title> since it will appear in search
> engines and in the title bar of the browser--but which would we say is
> the 'real' title here?
> 
> Regards,
> 
> Mark
> 
> On 31/08/2007, Karl Dubost <karl@w3.org> wrote:
>> Hi,
>>
>> I would like to know if there's a rule for getting the meaning of
>> this document.
>>
>> The XHTML document is located at
>> http://example.org/foo/bar
>>
>> The markup is
>>
>> <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML+RDFa 1.0//EN"
>>   "http://www.w3.org/MarkUp/DTD/xhtml-rdfa-1.dtd">
>> <html xmlns="http://www.w3.org/1999/xhtml" xml:lang="en">
>> <head>
>>      <meta name="dc:title" content="The Old Man And The Sea, Ernest
>> Hemingway" />
>>      <title>The Old Man And The Sea - Literature</title>
>> </head>
>> <body>
>> <h1>Book <span property="dc:title">The Old Man And The Sea,
>> Hemingway</span></h1>
>> </body>
>> </html>
>>
>>
>> Two solutions
>> <> dc:title "The Old Man And The Sea, Ernest Hemingway"^^XMLLiteral .
>> <> dc:title "The Old Man And The Sea, Hemingway"^^XMLLiteral .
>>
>> And another third one from the XHTML title element?
>>         (if title is considered to be dc:title. More on that later.)
>> <> dc:title "The Old Man And The Sea - Literature"^^XMLLiteral .
>>
>>
>> * Are there rules of precedence in RDFa?
>> * Is there a need to explain the semantics conformance of HTML-like
>> languages?
>>
>>
>>
>> --
>> Karl Dubost - http://www.w3.org/People/karl/
>> W3C Conformance Manager, QA Activity Lead
>>    QA Weblog - http://www.w3.org/QA/
>>       *** Be Strict To Be Cool ***
>>
>>
>>
>>
>>
> 
> 
Received on Friday, 31 August 2007 22:25:36 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 8 January 2008 14:15:09 GMT