- From: Jonathan Marsh <jmarsh@microsoft.com>
- Date: Tue, 30 May 2006 15:38:56 -0700
- To: <public-rdf-dawg-comments@w3.org>
- Cc: <www-ws-desc@w3.org>
- Message-ID: <37D0366A39A9044286B2783EB4C3C4E802C37BB2@RED-MSG-10.redmond.corp.microsoft.com>
Interop testing has revealed an oddity in the sparql-protocol-query.wsdl
file that we have in the WSDL 2.0 test suite.
The WSDL and the schemas associated with it generally use the
xs:import/@schemaLocation attribute to provide a hint as to where to
fetch the schema. Some of our implementations leverage this facility to
locate the schema components (as in the first xs:import below).
However, in one instance (the second xs:import of
sparql-protocol-query.wsdl), no schemaLocation is provided.
<types>
<!-- the SPARQL XML Results Format namespace -->
<xs:import
namespace="http://www.w3.org/2005/sparql-results#"
schemaLocation="result2.xsd" />
<!-- the W3C XML Schema types for SPARQL
Protocol -->
<xs:import
namespace="http://www.w3.org/2005/09/sparql-protocol-types/#" />
</types>
The schemaLocation attribute is of course optional, and we could mandate
additional mechanisms in our test framework to deal with it (e.g. add a
catalog file). However the single omission looks more like a mistake to
me than a deliberate design choice. Is the omission of schemaLocation
deliberate? If all other factors are equal, would you consent to adding
it?
[ Jonathan Marsh ][ jmarsh@microsoft.com
<mailto:jmarsh@microsoft.com> ][ http://spaces.msn.com/auburnmarshes
<http://spaces.msn.com/auburnmarshes> ]
Received on Tuesday, 30 May 2006 22:40:17 UTC