W3C home > Mailing lists > Public > public-ws-semann@w3.org > March 2007

Re: SAWSDL implementation question

From: Jacek Kopecky <jacek.kopecky@deri.org>
Date: Mon, 12 Mar 2007 15:59:30 +0100
To: Rama Akkiraju <akkiraju@us.ibm.com>
Cc: SAWSDL WG <public-ws-semann@w3.org>
Message-Id: <1173711570.14602.41.camel@localhost>

Dear Rama, 

it seems that the implementation as you describe it would go towards our
implementation criterion 5: For every SAWSDL feature, at least one
specification that uses it for Semantic Web Services automation.
Your tool does some SWS automation, so the specification of how it does
it is at least implied, and most probably it's described in the
documentation of the tool.

So the answer is most probably yes, it would be useful if you do change
that namespace (if you can't extend the implementation scope). 

Best regards,
Jacek

On Thu, 2007-03-08 at 12:15 -0500, Rama Akkiraju wrote:
> 
> 
> Hi Jack,
> 
> We are discussing the possibility of adopting our IBM Semantic Matching
> System to support part of SAWSDL spec. A few things came up.
> 
> As you might be aware, our system currently supports basic annotations on
> WSDL inputs, outputs and XML schemas as per WSDL-S format for WSDL 1.1. If
> we were to change the namespaces we have from 'wssem' to 'sawsdl'in our
> system, would that count as an implementation enough for W3C? Would it
> contribute to advance the spec from candidate recommendation to
> recommendation? Please note that after we make the namespace change, the
> following things will NOT be supported by our system.
> 
>    There will be no support for WSDL2.2
>    For WSDL1.1, the following things will NOT be supported:
>    There will be no support for SchemaMappings (support as in any API for
>    parsing schema mappings)
>    We only deal with a single modelReference on an element/attribute.
>    Multiple modelReferences i.e., 'listOfAnyURI' in SAWSDL spec will not be
>    supported.
>    Annotations on faults will not be supported.
>    Annotations via modelReferences on operations (i.e., attrExtensions),
>    interfaces, and faults will not be supported.
>    Extraction of annotations on attributes has not been tested and cannot
>    be guaranteed.
> 
> Annotations via modelReference on simpletypes, complex types (not all
> possibilities discussed will be supported), and elements WILL be supported.
> We can then do Web Service interface matching, mapping, discovery and
> composition with or without annotations as you might recall from the demo I
> have shown you.
> 
> Please let me know if with the above omissions, if our system with the
> namespace changed  from 'wssem' to 'sawsdl' would still count toward an
> implementation for the WG. We will decide things accordingly.
> 
> Thank you
> 
> Regards
> Rama Akkiraju
> 
> 
> 
Received on Monday, 12 March 2007 14:59:46 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 19:58:46 UTC