W3C home > Mailing lists > Public > public-swbp-wg@w3.org > November 2005

ALL: Request for SPARQL b-node problem use case

From: McBride, Brian <brian.mcbride@hp.com>
Date: Tue, 29 Nov 2005 07:17:23 -0000
Message-ID: <DE62D3D0BDEF184FBB5089C7D387C37455611B@sdcexc04.emea.cpqcorp.net>
To: <public-swbp-wg@w3.org>

David has highlighted a concern with SPARQL and we are asking all WG
members if they can suggest a use case illustrating the issue.

The issue concerns naming b-nodes.  Consider a query Q which cannot be
directly expressed in SPARQL, but must be split into two stages.  The
first stage Q' returns results which are processed by the querying
application, possibly including human review, and the nodes selected are
the subject of further queries.  If the nodes in question are b-nodes
there can be a problem because SPARQL does not provide a mechanism for
naming the b-nodes in question.  It may be that the b-nodes can be
identified by some of their property values.  The issue arises where
this is not the case.

David: is that an accurate characterization of your concern.

Brian
Received on Tuesday, 29 November 2005 07:18:43 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 8 January 2008 14:17:19 GMT