W3C home > Mailing lists > Public > public-rdf-in-xhtml-tf@w3.org > January 2008

Re: Do the following examples generate any triples?

From: Ivan Herman <ivan@w3.org>
Date: Wed, 09 Jan 2008 11:53:33 +0100
Message-ID: <4784A7AD.5040008@w3.org>
To: Mark Birbeck <mark.birbeck@formsPlayer.com>
Cc: Manu Sporny <msporny@digitalbazaar.com>, RDFa <public-rdf-in-xhtml-tf@w3.org>


Mark Birbeck wrote:
>> In all models
>>
>> <> foaf:knows <#me> .
>>
>> is generated if there is an implicit @about="" on the <body> element.
>> But I am not sure that is the case, in which case that triple is not
>> generated either (sorry Ben, we seem to disagree on that).
> 
> Ok...we need to check this. I think we say that the document URI
> itself is the default for @about (which may be set by <base>), but
> we'll need to double-check. But since we have agreed that this is what
> we _want_, let's just assume for now that this _is_ the triple
> generated.
> 

I agree. And I think we did have a discussion in the path about this, 
when you proposed to have an implicit @about="" on the <head> element 
that I misunderstood and I thought you meant to have this implicitly on 
the <html> element. I am not sure where we are with that. I still 
believe that having an implicit @about="" on <html> would make the trick.

(Well... Shane was not sure whether we should allow any RDFa attributes 
on <html>; the current DTD does not. I asked Steven last Monday about 
that, he said that some old IE version's DOM implementation made it 
difficult if not impossible to access attributes on the <html> element, 
hence the restriction. We may have to test this separately.)

> 
>> I first believed that in the model of Mark the
>>
>> <#me> foaf:knows <#ben> .
>>
>> is also generated regardless of whether the previous triple is
>> generated. I then had to re-read the syntax document to find out, via
>> the processing steps, that this is not so. (@rel has a higher priority
>> in setting [new subject], ie, it will be set to null, and that is not
>> send down to <span>)
>>
>> Which by itself is also significant (thanks, Manu!): if there is an
>> uncertainty on this for us, too, this shows the extra complexity induced
>> by the usage of @resource (sorry Mark:-)...
> 
> I'm not sure if it's completely logical to say that a proof of some
> kind of extra complexity in the rules is that you had to read them!
> But then I'm no mathematician. ;)
> 

:-) None of the end users will read that part of the document, and to 
answer this (edge) case one has to. And that is what bothers me. Anyway...


> 
>>> -----------------------------------------------------------------
>>> <div href="#me">
>>>    <span rel="foaf:knows" href="#ben">
>>> </div>
>>> -----------------------------------------------------------------
>>>
>> In the model of Mark @href sets the [new subject] in <div>, which is
>> then turned into [current subject] that is sent down to <span>. Ie, the
>>
>> <#me> foaf:knows <#ben> .
>>
>> will be generated.
> 
> Yes. Ben seems ok with this. Are you, Ivan?
> 

Well... I am not sure, and I am not convinced. I am absolutely content 
to use @about in such a case and not @href; I am concerned by the 
appearance of unexpected triples again.

(Note that this may be considered as a mild edge case, too: after all, 
@href can appear on <a> only in XHTML1)

> 
>>> -----------------------------------------------------------------
>>> <div resource="#me" rel="foaf:knows">
>>>    <div resource="#ben">
>>>       <a rel="foaf:knows" href="#mark">Mark</a>
>>>    </div>
>>> </div>
>>> -----------------------------------------------------------------
>>>
>>> -- manu
>>>
>> <#me> foaf:knows <#ben> .
>>
>> will be generated for the same reason.
> 
> Not in my model...no. There is actually no connection between the
> outer statement and the inner one. It could just as easily be written:
> 

Oops, I am sorry, I was wrong. In my analysis of your model, one would get

<> foaf:knows <#me> .
<#ben> foaf:knows <#mark> .

which is what you say below. Sorry about that.


>   <div resource="#me" rel="foaf:knows">
>   </div>
> 
>   <div resource="#ben">
>     <a rel="foaf:knows" href="#mark">Mark</a>
>   </div>
> 
> The first statement gives:
> 
>   <> foaf:knows <#me> .
> 
> as usual, and the second gives:
> 
>   <#ben> foaf:knows <#mark> .
> 
> Ben indicated that he is happy with the second statements generating
> said triples; are you?
> 

Well, this is exactly the same question as the one above because we aim 
at a symmetry of @href and @resource.

Mark, in one of your previous mail you try to separate this case  from 
the behaviour of @href/@resource with hanging rels and which was the 
core of our discussions the past few days. I am not sure that is a good 
idea; if we do _not_ agree on your model with hanging rels, but we would 
agree on this usage of @href/@resource, we may then end up with some 
sort of a spagetti specification which introduces all kinds of extra 
cases when @resource is around but without @rels, etc. I am not sure 
that is wise... That also adds to me discomfort to the previous question 
you raised.

Ivan


> Regards,
> 
> Mark
> 

-- 

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


Received on Wednesday, 9 January 2008 10:53:27 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Wednesday, 9 January 2008 10:53:28 GMT