SPARQL HTTP Update terminology (was: Re: ISSUE-49 (are graphs information resources?))

> However, section 5.4 of HTTP Update [1] says,
>
>> The HTTP GET method SHOULD be used to retrieve a graph representation of the networked RDF knowledge identified by the Request-URI.
>>
>> This is equivalent to the following SPARQL query:
>>
>> CONSTRUCT { ?s ?p ?o } WHERE { GRAPH<graph_uri>  { ?s ?p ?o } }

I'm a bit confused here as to the intent in SPARQL HTTP Update: the 
section "Graph Identification" just talks graph identification, with two 
forms, direct and indirect.   It does not talk about graph stores or RDF 
datasets nor define terminology for the first part of an indirect 
identification:

http://host/thing?graph=http://www.example.org/other/graph

Elsewhere in the doc itself, there is indirect reference to graph stores 
and RDF datasets through the use of SPARQL Update illustrations as above.

	Andy