W3C home > Mailing lists > Public > public-rdf-dawg-comments@w3.org > December 2004

Re: bnodes/existentials in Construct

From: Seaborne, Andy <andy.seaborne@hp.com>
Date: Wed, 15 Dec 2004 09:13:45 +0000
Message-ID: <41C00049.1040901@hp.com>
To: Sandro Hawke <sandro@w3.org>
Cc: public-rdf-dawg-comments@w3.org


The working group has decided about CONSTRUCT yet but bNodes do not appear 
in the graph patterns themselves.

The CONSTRUCT result form does not assert the graph formed back into the 
original dataset so triples from a CONSTRUCT result do not effect the 
matching of the query pattern iteself.


Sandro Hawke wrote:
> Are bnodes allowed in the Construct clause?  If so, that's a lot like
> existentially quantified terms in the consequent of a rule, and we're in
> Horn logic instead of the more conventional datalog.
>     -- sandro
Received on Wednesday, 15 December 2004 09:14:17 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 20:52:05 UTC