W3C home > Mailing lists > Public > public-rdf-in-xhtml-tf@w3.org > May 2008

RE: crazy ivan question? (sparql.org issues)

From: Seaborne, Andy <andy.seaborne@hp.com>
Date: Wed, 28 May 2008 15:06:04 +0000
To: Manu Sporny <msporny@digitalbazaar.com>, W3C RDFa task force <public-rdf-in-xhtml-tf@w3.org>
CC: Benjamin Nowack <bnowack@semsol.com>
Message-ID: <38CBA1F6A350B044AF785E63AAC3C677649B4678B8@G5W0276.americas.hpqcorp.net>


> -----Original Message-----
> From: Manu Sporny [mailto:msporny@digitalbazaar.com]
> Sent: 28 May 2008 15:57
> To: W3C RDFa task force
> Cc: Seaborne, Andy; Benjamin Nowack
> Subject: Re: crazy ivan question? (sparql.org issues)
>
> Ivan Herman wrote:
> > what does [ERROR] means when running the unit tests with crazy ivan? I
> > get this for tests #76,77,87,94 using sparqler...
>
> It means that either the RDFa web service or the SPARQL web service died
> in some horrible way (syntax error in the input, internal server error,
> busy, etc.)
>
> #76, #77, #87, and #94 fail on SPARQL.org because the SPARQL queries are
> too large for one of the proxy-ing services. This has been an issue for
> the past several months. Just switch SPARQL engines to Virtuoso for that
> test only and click on the "ERROR" text, which will re-run the unit test
> using the newly selected SPARQL engine.
>
> Andy - any chance that this could be fixed? I know you looked into it in
> February, but we're depending pretty heavily on sparql.org now that
> Virtuoso isn't working for our isBlank() queries.

I tracked this down (I think) to the fact that the web server used (Jetty) has limitations on handling GET requests (Google finds many such reports).  But it's only GET.

If you swap to using POST then you should be OK.  And given potential, if unlikely, caching, you want POST anyway.

There was a message from OpenLink recently (yesterday on jena-devel@lists.soureforge.nt) saying they will be implementing blank nodes labels (which is what stops them currently).  Then they will be able to support the whole of the Jena API as well as be SPARQL compliant for isBlank().

        Andy

>
> Benji, any chance that the isBlank() SPARQL query could be implemented
> correctly in Virtuoso?
>
> > My output for #94 is wrong, though and, note, the RDF/XML added to the
> > tests is wrong, too (it comes from an earlier incarnation of pyRdfa, I
> > believe...). Maybe that explains the 'Error' (is that an XML error?)
>
> Do you mean the SPARQL is wrong, or the RDF/XML is incorrect?
>
> The SPARQL isn't wrong, AFAIK. The RDF/XML is whatever is being output
> from the currently selected RDFa parser web service and could be wrong
> if your parser is generating incorrect RDF/XML.
>
> The N3 is being generated by PyRDFa... which is as accurate as PyRDFa is
> currently.
>
> -- manu
>
> --
> Manu Sporny
> President/CEO - Digital Bazaar, Inc.
> blog: DB Launches Medical Record Sales Service with Shepherd Medical
> http://blog.digitalbazaar.com/2008/02/24/health2trade/


Received on Wednesday, 28 May 2008 15:07:35 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Wednesday, 28 May 2008 15:07:35 GMT