Re: [Graphs] Proposal: RDF Datasets

Antoine Zimmermann wrote:
> Le 08/03/2011 18:23, Ivan Herman a écrit :
>>
>> On Mar 8, 2011, at 18:15 , Antoine Zimmermann wrote:
>>
>>> Le 08/03/2011 17:09, Ivan Herman a écrit :
>>>>
>>>> [...] You and Antonie are arguing on the semantics of _datasets_; I 
>>>> am still not convinced that this discussion should happen in the 
>>>> first place!
>>>
>>>
>>> Well, we are arguing about the semantics of multigraph documents.
>>
>> I do not know what that means.
> 
> Something which contains the definition of several g-snaps, or g-boxes, 
> such as an N-Quads document, a TriG document or whatever which, 
> abstractly, maps to a finite set of g-snaps or g-boxes, or named g-snaps 
> or named g-boxes, depending how we interpret it. That is, we are 
> discussing how should the examples given in TF-Graphs-UC should be 
> interpreted.

Surely it could only be the current state of a named-g-box, or a g-snap, 
or is there a proposal to have all the previous contents of each g-box 
in the same serialization too?

There's a minor issue as well I guess, that given a doc which contains 
:g1 { .. }  :g2 { .. } you don't know whether it's to be interpreted as 
a snapshot of two named-g-box's, or as two g-snaps.

Also, I guess if one starts publishing rdf datasets on the web, or 
quads, then I'll be needing a 5-tuple store for those, and a 
serialization for those 5-tuples, which if I expose, you'll need a 
6-tuple store for (etc etc)

Just some minor thoughts.

Best,

Nathan

Received on Tuesday, 8 March 2011 18:15:54 UTC