SAWSDL Comments

Dear all,

In November the SAWSDL Working Group held a face to face meeting in
which we discussed the comments submitted by the members of the SEE
Technical Committee regarding the current version of the SAWSDL
specification. This email intends to communicate the resolutions that
were adopted in order to address some of the issues raised.

First, the Semantic Execution Environment Technical Committee presented
their concerns with respect to the restriction to using modelReference
to annotate services. The rationale behind this request is, we believe,
the need to support adding global annotations to services is a similar
fashion as Capabilities are used in WSMO. The SAWSDL Working Group
believes that this is mainly a conceptual mismatch between both
approaches to annotating Web Services, which can be easily handled by
means of the current specification. It has therefore been agreed to keep
the specification unchanged since it intends to be a generic basis on
which more complex frameworks should build upon and adapting it to fit
the specificities of all the existing approaches would lead to an
unnecessarily complex and confusing specification. Instead the working
group suggests the following solution which is based on the fact that
all modelReferences apply. When it comes to annotating WSDL 2.0, the
same effect can be obtained by simply annotating the interface since
services only have one interface. For WSDL 1.1 the solution is similar
although more verbose since it involves annotating every portType
defining a service.

A second concern raised by the SEE Technical Committee regards the
mechanism for annotating the behavioural aspects of Web Services. In
this respect, the SAWSDL Working Group would like to point out that it
was not chartered for producing this type of annotations [1]. Still,
modelReference could be applied for annotating the behavioural aspects
by simply pointing to framework-specific definitions such as
Pre-conditions, Post-conditions, Effects, etc. Finally, we would like to
stress the fact that the specification does not uniquely mean Web
Services Choreography by "behavioural aspects". Doing so would restrict
the specification and could actually lead to confusion since there
already exists a W3C Working Group which is precisely addressing Web
Services Choreography. Instead by "behavioural aspects" the SAWSDL
Working Group understands any behavioural aspect which may include, but
is not limited to, Choreography.

The SAWSDL Working Group would like to thank the effort devoted by the
SEE Technical Committee to review the specification. Further resolutions
will be communicated as soon as they are adopted. Meanwhile, we are
looking forward to hearing from your reactions regarding the decisions
herein.

Many thanks,

Carlos Pedrinaci, on behalf of the SAWSDL Working Group.


[1] http://www.w3.org/2005/10/sa-ws-charter.html

Received on Monday, 27 November 2006 15:02:23 UTC