Re: ISSUE-37 WAS:Proposal for containers

On 2 Feb 2013, at 11:28, "Wilde, Erik" <Erik.Wilde@emc.com> wrote:

> hello henry.
> 
> On 2013-02-01 15:57 , "Henry Story" <henry.story@bblfish.net> wrote:
>> I agree. I was just putting forward what I took to be Erik Wilde's views
>> there. I think he is saying, that atom functions like this
>> (removing the owl:sameAs indirection ) after POSTing a binary:
>> 
>> <> a ldp:Container;
>>  rdf:entry  <entry3> .
>> 
>> <entry3>   a :Entry;
>>     :author <http://joe.example/#i>;
>>     :id "http://atom.example/entry3";
>>     :title "mouse";
>>     :updated "2013-01-13T19:10:02Z"^^xsd:dateTime;
>>     :content <mouse.gif> .
>> 
>> But I don't think that this is in fact the only interpretation
>> one can have of Atom. I think one could have an interpretation
>> of Atom that would bring it closer to LDP.
> 
> i think that this representation is accurate, but i am not quite sure how
> else you would be able to say what's happening in atompub. what's not
> represented here is the fact that the mouse is managed by the server. in
> atompub this would be exposed by an edit-media link that would allow
> clients to edit the mouse, when following that link. for clients not
> interested in editing (i.e., ignoring the edit-media link), this would
> look like just another aggregated entry (i.e., they follow the link to the
> mouse to GET the mouse), and that would be good enough for what they need
> to do.


We want to choose which resource the rdfs:member relation in an
LDPC can point at. Should the object of that relation be the <entry3> or
should it be <mouse.gif> ? One can make arguements for both sides.

I wrote up in more detail what taking <mouse.gif> as the object of 
the rdfs:member relation would be in the longer mail 
"POSTing binary data - the Atom Use Case" 
http://lists.w3.org/Archives/Public/public-ldp-wg/2013Feb/0010.html
that addresses ISSUE-15 .

Depending on what you think the meaning of the :edit 
and :edit-media relation is, you could have

 <entry3> :src <mouse.gif> ;
          :edit <edit/entry3> ;
          :edit-media <edit/mouse.gif> .

or

 <mouse.gif> :src <mouse.gif> ;
             :edit <edit/entry3> ;
             :edit-media <edit/mouse.gif> .

Clearly in the second case the :src becomes redundant,
and all one needs is:

 <mouse.gif> :edit <edit/entry3> ;
             :edit-media <edit/mouse.gif> .

which is the kind of argument that would favor <mouse.gif>
as the subject of the entry - especially after a POST of
a binary.

Anyway this is tricky as one has to interpret a syntax without
an explicit model. One would need to look at how it is used
on the internet, get empirical data to corroborate different theories,
and then use that as part of a process for creating the official atom 
ontology for which one could then produce a GRDDL. I tried that in 2006. 
Since then Atom has had much wider usage, so one can perhaps come
to different conclusions, with hindsight. ( Also I did not formalise
the atompub notions ). 

	Henry

> cheers,
> 
> dret.
> 

Social Web Architect
http://bblfish.net/