Re: off-topic question on cursors

Axel,

You don't need cursors to stream results nor any enhancement of the 
protocol.  HTTP is a TCP connection so you can flow the results from 
query processor to client application and TCP does the flow control.  I 
do this for both SPARQL XML Results and for JSON results and while the 
client is parsing and processing the front of the stream, the tail is 
still being generated.

Cursors address a number of issues - client control of the flow (alt. 
use TCP buffers filling up) and the ability to access the results in a 
different order.

	Andy

On 11/01/2010 8:06 PM, Orri Erling wrote:
>
>
> Axel
>
> Virtuoso supports cursors over SPARQL result sets over SQL CLI's like JDBC
> and ODBC.  There are extension data types for typed literals, distinguishing
> between strings and iri's etc.  Then there is a Jena and Sesame compatible
> wrapper that has cursors such as these support.
>
> I do not know of a cursor implementation over HTTP.  OWL QL specifies
> something of the sort, though.
> een
>
> Orri
>
>
> -----Original Message-----
> From: public-rdf-dawg-request@w3.org [mailto:public-rdf-dawg-request@w3.org]
> On Behalf Of Axel Polleres
> Sent: Saturday, January 09, 2010 2:21 AM
> To: SPARQL Working Group
> Subject: off-topic question on cursors
>
> <chair-hat-off>
> an off-topic (for sure off-charter) question:
>
> does any implementation support some form of enhancement of the protocol by
> something like cursors
> to stream query results one by one or in chunks? If so, how does such
> support look
> in current implementations?
>
> thanks still for answers, if any,
> Axel
> </chair-hat-off>
>
>
>
>
> ______________________________________________________________________
> This email has been scanned by the MessageLabs Email Security System.
> For more information please visit http://www.messagelabs.com/email
> ______________________________________________________________________

Received on Monday, 11 January 2010 20:25:10 UTC