W3C home > Mailing lists > Public > public-rdf-dawg-comments@w3.org > May 2012

RE: Editorial suggestion for CREATE operation

From: Polleres, Axel <axel.polleres@siemens.com>
Date: Mon, 14 May 2012 08:36:33 +0200
To: David Booth <david@dbooth.org>
CC: public-rdf-dawg-comments <public-rdf-dawg-comments@w3.org>
Message-ID: <9DA51FFE5E84464082D7A089342DEEE80139B0715D28@ATVIES9917WMSX.ww300.siemens.net>
Dear David,

As for the paragraph you quote:

> However, I just noticed the "Note" below that section, which 
> says that any CREATE may be viewed as immediately followed by a DROP:
> [[
> Since (non graph-aware) Graph Stores may remove graphs that 
> are left empty, for such Graph Stores any CreateOperation may 
> be viewed as implicitly immediately followed by a 
> DropOperation (see next subsection), or simply as an 
> operation with no effect.
> ]]

Please note that it says explicitly here that 

"(non graph-aware) Graph Stores may remove graphs 
 *that are left empty* "

Thus, existing contents (i.e. non-empty graphs) 
wouldn't be possibly affected.

Please let us know whether this response 
Addresses/clarifies your concern.

Best regards,
Axel

 
-- 
Dr. Axel Polleres 
Siemens AG Österreich 
Corporate Technology Central Eastern Europe Research & Technologies 
CT T CEE 
 
Tel.: +43 (0) 51707-36983 
Mobile: +43 (0) 664 88550859
Fax: +43 (0) 51707-56682 mailto:axel.polleres@siemens.com 
 

> -----Original Message-----
> From: David Booth [mailto:david@dbooth.org] 
> Sent: Friday, 11 May 2012 4:30 PM
> To: public-rdf-dawg-comments
> Subject: Re: Editorial suggestion for CREATE operation
> 
> I'm looking further at this, and now I am more confused.  I 
> am trying to verify that when a CREATE SILENT operation is 
> performed on an existing named graph, the content of that 
> named graph is not changed.  This is how I previously 
> understood it to work and how I think it *should* work.
> And when I read the formal semantics of CREATE, I see:
> http://www.w3.org/TR/sparql11-update/#defn_createOperation
> [[
>   OpCreate(GS, iri) = GS union {(iri, {})} if iri not in 
> graphNames(GS); otherwise, OpCreate(GS, iri) = GS ]]
> 
> If I am reading the above correctly, in this case iri *would* 
> be in graphNames(GS), so the "otherwise" clause would apply 
> and the graph store would be unchanged, i.e., the content of 
> an *existing* named graph should be unchanged.
> 
> However, I just noticed the "Note" below that section, which 
> says that any CREATE may be viewed as immediately followed by a DROP:
> [[
> Since (non graph-aware) Graph Stores may remove graphs that 
> are left empty, for such Graph Stores any CreateOperation may 
> be viewed as implicitly immediately followed by a 
> DropOperation (see next subsection), or simply as an 
> operation with no effect.
> ]]
> 
> But a DROP operation would delete the content of an existing 
> named graph, so if I am understanding all this correctly, 
> there seems to be a conflict between this "Note" and the 
> formal semantics.
> 
> Clarification please?  I hope it is the "Note" that is in error.
> 
> Thanks!
> David
> 
> 
> 
> On Wed, 2012-04-11 at 12:17 -0400, David Booth wrote:
> > At http://www.w3.org/TR/sparql11-update/#create
> > I suggest adding the sentence: "If the graph already exists 
> then its 
> > content is unchanged."
> > 
> > 
> 
> --
> David Booth, Ph.D.
> http://dbooth.org/
> 
> Opinions expressed herein are those of the author and do not 
> necessarily reflect those of his employer.
> 
> 
> 
Received on Monday, 14 May 2012 06:37:03 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 14 May 2012 06:37:04 GMT