Re: Interesting experience: my foaf in RDFa

Dan Brickley wrote:
> 
>> However, one thing _really_ bugs me: Our nice (X)HTML+RDFa documents are
>> ignored by 'Semantic Web search engines' as ptsw [2] or sindice [3].
>> People, move on!
> 
> Can I put in a vote for W3C to spend some effort on updating the RDF
> Validator (http://www.w3.org/RDF/Validator)? I reckon that, alongside
> the syntax checker at validator.w3.org, will go a long way to pushing
> these developments out into wider usage.
> 

I am not sure what the current state of the problems are. My foaf.html
file validates!

> The next thing I try in RDFa will probably be expressing (a subset?) of
> the FOAF schema in the FOAF spec. Not really sure what that should look
> like, but ... would be nice for that doc to have a "valid html" icon in it!
> 

See above

>>> Do not crucify me: maybe we should have (as some sort of a very edge
>>> case advanced feature) some sort of an import mechanism. Something like
>>>
>>> <link rel="rdfa-import" href="...."/>
>>
>> As I see the benefit and the use I do support your proposal, though
>> the concrete solution you propose seems a bit awkward to me. Mark? Any
>> thoughts?
> 
> Isn't this re-inventing iframes? How should an RDFa parser pointed at
> dan.html deal with an iframe that includes photos.html ?
> 

Hm. I did not think of anything that complicated. The only thing I had
in mind is to point at an existing RDF/XML file that would be parsed and
merged with the resulting graph. That important file would have no
bearing on the HTML content at all (that is why I thought of using a
simple <link> mechanism). Even if we do something here, I would try to
keep the complexity to the bare minimum here...

Ivan



> cheers,
> 
> Dan
> 
>> Cheers,
>>     Michael
>>
>> [1] http://sw-app.org/mic.xhtml
>> [2] http://pingthesemanticweb.com
>> [3] http://sindice.com
>>
>> ----------------------------------------------------------
>>  Michael Hausenblas, MSc.
>>  Institute of Information Systems & Information Management
>>  JOANNEUM RESEARCH Forschungsgesellschaft mbH
>>  Steyrergasse 17, A-8010 Graz, AUSTRIA
>> ----------------------------------------------------------
>>> -----Original Message-----
>>> From: public-rdf-in-xhtml-tf-request@w3.org
>>> [mailto:public-rdf-in-xhtml-tf-request@w3.org] On Behalf Of Ivan Herman
>>> Sent: Friday, August 24, 2007 12:39 PM
>>> To: W3C RDFa task force
>>> Subject: Interesting experience: my foaf in RDFa
>>>
>>> I decided to experiment with RDFa with a slightly more complex example.
>>> I used to have my foaf file[1] only in RDF/XML, ie, edit it, update
>>> it, etc, in that format. I now created an XHTML/RDFa version of it[2]
>>> and from now on my intention to update [2] and generate [1]
>>> automatically (as an extra bonus my non-informational URI[3]
>>> redirects to either the html or the rdf file, depending on the HTTP
>>> request).
>>>
>>> My foaf file is fairly complex, and it mixes a good number of
>>> namespaces (I always used it as some sort of an experiment to express
>>> different things). Putting it into RDFa was a bit of a challenge here
>>> and there, but it was worth it. Some experiences:
>>>
>>> - It works!  This is a really important point: I did not find any
>>> construction in my previous foaf file that I could not express somehow.
>>> And that is really important; it is a one-time evidence that we do
>>> have something good here.
>>>
>>> - It is an authoring challenge (this is not unlike microformats). Of
>>> course, editing the file in a screen editor is possible but makes it
>>> a bit difficult to follow and is error prone. The problem with a
>>> WSWYG editor like Amaya is that editing the attributes is sometimes a
>>> bit complicated. What you need is an editor that makes it easy to
>>> switch between WYSWYG and source view but most of them mean moving to
>>> another window and thereby loosing context. My best experience is
>>> with Adobe GoLive which allows (in the WYSWG window) to have a pop up
>>> window on a specific element with all the xml source editable. That
>>> helps a lot...
>>>
>>> - It can be a bit convoluted sometimes. No real surprise there: if
>>> you look at [2] you can see that I tried to incorporate some sioc
>>> statements and turning that into RDFa was a bit complicated. Well,
>>> the sioc statements themselves are convoluted....
>>>
>>> This means that some of the RDF constructs become a bit unnatural in
>>> HTML, if you want to add some humanly readable text to it. Which
>>> raises a practical question. What if the author wants to keep some of
>>> what he/she wants to express in RDF/XML and would like to 'bind' it
>>> to the RDF extracted from HTML? My solution was to add a <link>
>>> statement with rel="rdfs:seeAlso", but this relies on the RDF
>>> environment to understand and interpret that.
>>>
>>> Do not crucify me: maybe we should have (as some sort of a very edge
>>> case advanced feature) some sort of an import mechanism. Something like
>>>
>>> <link rel="rdfa-import" href="...."/>
>>>
>>> which means that the end result should be the merge of the value of
>>> @href and the extracted RDF. I am not sure... just raising the
>>> possibility here...
>>>
>>>
>>> Ivan
>>>
>>>
>>>
>>> [1] http://www.ivan-herman.net/foaf.rdf
>>> [2] http://www.ivan-herman.net/foaf.html
>>> [3] http://www.ivan-herman.net/Ivan_Herman
>>> -- 
>>>
>>> Ivan Herman, W3C Semantic Web Activity Lead
>>> Home: http://www.w3.org/People/Ivan/
>>> PGP Key: http://www.ivan-herman.net/pgpkey.html
>>> FOAF: http://www.ivan-herman.net/foaf.rdf
> 

-- 

Ivan Herman, W3C Semantic Web Activity Lead
Home: http://www.w3.org/People/Ivan/
PGP Key: http://www.ivan-herman.net/pgpkey.html
FOAF: http://www.ivan-herman.net/foaf.rdf