OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.

 


Help: OASIS Mailing Lists Help | MarkMail Help

office-metadata message

[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]


Subject: Re: [office-metadata] Multiple content nodes representing on RDF subject


Hi Svante,

On Dec 22, 2006, at 7:26 PM, Svante Schubert wrote:

> Image there is your name "Bruce D'Arcus" in a text:p. This string 
> "Bruce D'Arcus" is a literal and when only parts are being used, it 
> has in general not the same semantic as in full length.

I keep asking you to be specific and to avoid using the too-brood word 
"semantic."

What -- exactly -- are you trying to say here? Your subject is what? 
The paragraph? Or me as it's referenced in the paragraph? The 
distinction matters.

> In content.xml:
>
> <text:p xml:id="_foo1">Bruce D'Arcus</text:p>
>
> In metadata:
>
> <RDF:RDF xmlns:RDF="http://www.w3.org/1999/02/22-rdf-syntax-ns#";
>         xmlns:test="http://www.meta-subcomitee.com/dummy/rdf#";>
> <rdf:Description rdf:about="#Bruce">
>  <test:name rdf:resource="content.xml#_foo1"/>
> </rdf:Description>
> </rdf:RDF>

OK, please pay attention to how I explain this, because this is what 
I'm asking of you:

You wish to say that the subject of the paragraph is the person 
identified with the (local) URI <#Bruce>. You further wish to say that 
the test:name value for that subject is "Bruce D'Arcus."

Subject = <#Bruce>
Predicate = test:name
Object = "Bruce D'Arcus"

The way you are encoding the RDF there is totally non-standard. You 
will never see, for example, name or title properties encoded in that 
way in the wild, and it would be completely unreasonable of this SC to 
insist that every developer out there follow some new rules we invent 
in order to model (as opposed to encode) their data.

The ONLY reason you are doing this is because you don't want -- for 
some unspecified reason -- to do either:

1)

<text:p meta:about="#Bruce">
	<text:span 
meta:property="http://www.meta-subcomitee.com/dummy/rdf#name";>Bruce 
D'Arcus</text:span>
</text:span>

... or even:

2)

<text:p xml:id="#Bruce">Bruce D'Arcus</text:p>

<RDF:RDF xmlns:RDF="http://www.w3.org/1999/02/22-rdf-syntax-ns#";
         xmlns:test="http://www.meta-subcomitee.com/dummy/rdf#";>
<rdf:Description rdf:about="#Bruce">
  <test:name>Bruce D'Arcus</test:name>
</rdf:Description>
</rdf:RDF>

>>> Imagine a name, quote, which looses his semantic when it's string is 
>>> reduced.  There should be a nice example where the literal has to be 
>>> split, someone help me out on this?
>>
> Now I am looking for an example as earlier stated.
> An example of a silly user could be, if he inserts a table somewhere 
> inside your name.
>
> In content.xml:
>
> <text:p xml:id="_foo1">Bru</text:p>
> <table:table>...</table:table>
> <text:p xml:id="_foo2"> ce D'Arcus</text:p>
>
> By an indirection the semantic won't be lost.
>
> <content xml:id="_fooA">
> <ref idref="_foo1"/>
> <ref idref="_foo2"/>
> </content>

I dunno, maybe something like:

<text:p meta:about="#Bruce">Bru</text:p>
<table:table>...</table:table>
<text:p meta:about="#Bruce"> ce D'Arcus</text:p>

>> You keep using the word "semantic" which is effectively meaningless. 
>> Again, what's the subject that you want to describe, and what 
>> statement(s) do you want to make about it?  If you cannot be specific 
>> about this, it's impossible to address your issue.
>>
>> If you must, post the entire example you have in mind.
> Hopefully this could help you.

Yes, thank you.

Bruce



[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]