Re: Comments on http://www.w3.org/TR/2008/WD-rdfa-syntax-20080221

Hi Johannes,

Thanks for your comments. Some of the points you raise are easily
addressed and only involve minor changes to the document, but they
would require the general approval of the task-force. I've outlined
the possible solutions below, for discussion by the group.

>  5.3. Chaining
>
>  The first code example
>
>  <div about="http://dbpedia.org/resource/Albert_Einstein">
>    ...
>    <img rel="foaf:depiction"
>  src="http://en.wikipedia.org/wiki/Image:Albert_Einstein_Head.jpg" />
>
>  seems to indicate that a triple
>
>  <http://dbpedia.org/resource/Albert_Einstein>
>    foaf:depiction
>    <http://en.wikipedia.org/wiki/Image:Albert_Einstein_Head.jpg> .
>
>  should be created. But according to section 5.5 step 5 and section
>  6.1.1.3 the rel attribute would have to be on the div element, not the
>  img element. The example as is IMHO creates an incomplete triple
>
>  <http://en.wikipedia.org/wiki/Image:Albert_Einstein_Head.jpg>
>    foaf:depiction
>    ? .

You are correct. That mark-up should have been removed, since it
predates the final decision we made to have img/@src represent the
subject of statements, rather than the object. (Alternatively, the
mark-up could be 'corrected' by changing it so that @rev is used
instead of @rel.)


>  5.5. Sequence
>
>  * step 5:
>  "Note that final value of the [current object resource] will either be
>  null (from initialization), a full URI or a bnode."
>  According to the algorithm, [current object resource] is never set to a
>  bnode.

You are correct. This also should have been removed, since as a result
of comments from Ed and Diego, the 'create a bnode' step was moved to
step 8.


>  * After step 8 follows step 10: Probably the numbering is incorrect.

That is correct...a closing <li> tag is missing.


>  * step 10 (9?):
>  "as a [plain literal] if:
>    * @content is present;
>    * or all children of the [current element] are text nodes;
>    * or there are no child nodes;
>    * or the body of the [current element] does have non-text child
>      nodes but @datatype is present, with an empty value."
>  According to this the example in 6.3.1.2
>
>  <span property="cal:dtstart" content="20070916T1600-0500"
>        datatype="xsd:datetime">
>    September 16th at 4pm
>  </span>.
>
>  would not create a typed literal (@content is present). Or can the
>  following bullets (XML literal, typed literal) "overwrite" the result
>  from the first bullet (plain literal)?

Well spotted! There is no notion of 'overwriting', and the simplest
solution is to make the minor change of moving the third step (that
checks for typed literals) to become the first.


>  6.1.1.5.2. Using an implicit object
>
>  second example:
>  <div about="http://dbpedia.org/resource/Baruch_Spinoza"
>  rel="dbp:influenced">
>    <div>
>      <span property="foaf:name">Albert Einstein</span>
>      <span property="dbp:dateOfBirth" datatype="xsd:date">1879-03-14</span>
>    </div>
>  </div>
>
>  Because the inner div element has no instanceof attribute, according to
>  section 5.5 step 4 [new subject] is set to the [evaluation context]'s
>  [parent object], which is <http://dbpedia.org/resource/Baruch_Spinoza>,
>  because the [current object resource] for the outer div was null.
>
>  The incomplete triple
>    <http://dbpedia.org/resource/Baruch_Spinoza>
>      dbp:influenced ? .
>  never gets completed. And because no bnode was created, the foaf:name
>  and dbp:dateOfBirth properties apply to
>  <http://dbpedia.org/resource/Baruch_Spinoza>.

I don't think there is a problem here. The [parent object] should have
been set at the previous level to a bnode, through the action of step
8. This should be passed to the empty <div>, and then passed again to
each of the two spans due to [skip element] being set to 'true'. Once
you get to the two spans, the [parent object] becomes the subject for
the statements, so the bnode gets used then.


>  Similar things happen with the third and forth examples in 6.1.1.5.2 and
>  the sixth example in 6.2:
>
>  <div about="http://dbpedia.org/resource/Baruch_Spinoza"
>  rel="dbp:influenced">
>    <span property="foaf:name">Albert Einstein</span>
>    <span property="dbp:dateOfBirth" datatype="xsd:date">1879-03-14</span>
>    <div rel="dbp:citizenship">
>      <span about="http://dbpedia.org/resource/Germany" />
>      <span about="http://dbpedia.org/resource/United_States" />
>    </div>
>  </div>

I think this is ok, too.


>  9.3. @rel/@rev attribute values
>
>  * While there is a "last" value listed, there is no "first".
>  * What about "top"?

They seem good additions to me, but I'll let the group comment. :)


>  A. Other XML Languages
>
>  "If a language includes @xml:base [XMLBASE], an RDFa parser for that
>  host language must process it, and use its value to set [base]."
>  Is this a requirement? The section is marked "informative".

Yes...informative language always seems to cause a problem. :) I think
that's also a discussion the group needs to have.

Regards, and thanks again for your input.

Mark

-- 
  Mark Birbeck

  mark.birbeck@x-port.net | +44 (0) 20 7689 9232
  http://www.x-port.net | http://internet-apps.blogspot.com

  x-port.net Ltd. is registered in England and Wales, number 03730711
  The registered office is at:

    2nd Floor
    Titchfield House
    69-85 Tabernacle Street
    London
    EC2A 4RR

Received on Tuesday, 26 February 2008 14:51:23 UTC