On 07/10/2014 11:58 AM, Robert Sanderson wrote:
>
>
>
>
>         Line 1: There is a resource without a global identifier which is an
>         Instance.
>
>
>     The implication of using a blank node is not "without a global
>     identifier" but "without a global identifier I care to supply here".
>
>
> Well, okay, but that just makes it even more ridiculous:
>
> There is a resource without a global identifier that I care to supply
> here (line 1), and here's its global identifier (line 2-4).
>
> QED? :)

Yes. This is standard in RDF.

Bear in mind that raw RDF data formats are aimed primarily at machines.

When handling datasets orders of magnitude that the available RAM, there 
are several useful processing modes which make radically different 
assumptions than data processing systems we might be used to. In these 
kinds of context what might seem trivial doesn't turn out to be and 
conversely, some things that seem hard become significantly easier. See 
for example https://en.wikipedia.org/wiki/Map-reduce 
https://en.wikipedia.org/wiki/Data_stream_mining etc.

cheers
stuart