W3C home > Mailing lists > Public > public-rdf-dawg@w3.org > July to September 2005

Re: Andy's review of the the protocol doc (part II)

From: Kendall Clark <kendall@monkeyfist.com>
Date: Mon, 8 Aug 2005 11:18:30 -0400
To: "Seaborne, Andy" <andy.seaborne@hp.com>
Cc: RDF Data Access Working Group <public-rdf-dawg@w3.org>
Message-ID: <20050808151830.GC17401@monkeyfist.com>

On Mon, Aug 08, 2005 at 12:24:41PM +0100, Seaborne, Andy wrote:
> 
> From "B. HTTP Bindings"
> v1.56
> 
> ==== "a. SELECT with no RDF dataset"
> 
> ==>
> "a. SELECT with service-provided dataset"
> 
> There is no RDF dataset description provided in the protocol request of 
> query language string but there is a dataset - it's fixed by the service.

Ah, yes, of course. Excellent catch Andy. Makes me feel dumb. :>

> Results need to use the new results namespace

ACK.

> ==== "b. SELECT with simple RDF dataset"
> 
> Maybe explain what's happening with default-graph-uri and maybe make it a 
> URL to a different host.
> 
> Results need to use the new results namespace
> Results need <results distinct="false" ordered="false">

ACK and ACK.

A question about the attributes on <results> -- do these need to be added to
all of the examples which don't explicitly add ordering and don't explicitly
request distinct results? That is, there's nothing special about this
request that I can see...?

> ==== "c. CONSTRUCT with simple RDF dataset"
> 
> Using [] in the query might be clearer.
> 
> CONSTRUCT { [] foaf:name ?who; ...

ACK.

> ----
> 
> Something is wrong with the results: may be it's the labels because there 
> are 6 different subjects but there should be 6 triples, with 3 subjects.
> 
> I guess the bnode labels are wrong
> e.g. "r1_r1_somebody" vs "r1_r1_somebody"

Those are the same.

> Using RDF/XML that grouped foaf:name and foaf:made toegther would be 
> clearer.

I'm just going to swap out a different CONSTRUCT example altogether. I don't
much care for this one anyway. ;>

> ====  "d. ASK with simple RDF dataset"
> ASK { ?book dc:creator ?who }
> is a possible query

Yes. I don't understand yr point.

> Mention its an identified default graph.

ACK.

> Old result format : should have the <results>
> Old results namespace

ACK and ACK.

> (skipping red text'ed examples)
> 
> ==== "C. SOAP Bindings"
> 
> Feel free to shorten the SOAP trace.

ACK -- haven't worked on this yet, only pasted it into the doc from yr
email. :>

Cheers, 
Kendall
Received on Monday, 8 August 2005 15:20:46 GMT

This archive was generated by hypermail 2.3.1 : Tuesday, 26 March 2013 16:15:24 GMT