See also: IRC log
No modifications to the agenda
<JacekK> last minutes: http://www.w3.org/2002/ws/sawsdl/minutes/20070123
Last Minutes approved
<scribe> ACTION: Eric to upgrade the SPDL page for SAWSDL readers and then work things out with the Usage Guide [PENDING] [recorded in http://www.w3.org/2002/ws/sawsdl/minutes/20070130#action01]
<scribe> ACTION: JacekK to move the test suite to the W3C site [PENDING] [recorded in http://www.w3.org/2002/ws/sawsdl/minutes/20070130#action02]
<scribe> ACTION: JohnM to provide an example for annotating operation inputs and/or outputs [DONE] [recorded in http://www.w3.org/2002/ws/sawsdl/minutes/20070130#action03]
<scribe> ACTION: JacekK to reply to Mary about the limitation for only global element declaration and type definition for schema mappings [PENDING] [recorded in http://www.w3.org/2002/ws/sawsdl/minutes/20070130#action04]
Jack: CR published last friday
... start thinking about recommendation pres release
... both documents have been published
... any discussions about the recommendation pres release to be done through the member-only mailing list
... usage guide remains as a working draft for there are no normative specs
... it will end up being a group note
<JacekK> CR issues list open: http://www.w3.org/2002/ws/sawsdl/issues/CR-20070126.html
Proposal by ajith: http://lists.w3.org/Archives/Public/public-ws-semann/2007Jan/0021
Ajith: in the scope of operations we could add annotations at the input/output level
Jack: what else would this annotations provide us?
... how would implementations use these annotations?
John: allowing pre and post conditions could provide us the same functionality
<Zakim> laurenth, you wanted to say that an automatic system can decide to drop reques acknowledgements but not order ones
Laurent: in automated composition knowing the types of replies would be very useful
Jack: pre and post conditions would help us here
... inputs and outputs can be correlated easily
... then you can identify the operation
<Zakim> laurenth, you wanted to say that model pre post conditions do not address the same problem
laurenth: at the semantic level, coming back to the WSDL to retrieve the semantics does not seem reasonable
Jack: the operation has this information
John: we could perhaps keep this placeholder for pre and post conditions
Jack: don't think this would do it, e.g. there could be operations with no output
... there are two related issues: CR issue 3 and Issue 33
Rama: we should add a section on annotating inputs and outputs
Joel: may not be able to add as it's borderline wrt charter
Jack: not really convinced about the need for annotating inputs/outputs
<Zakim> laurenth, you wanted to say : ok for adding a sentence in the spec to say that inputs and outputs is possible
laurenth: add a sentence about the possibility to add these annotations without specifying the motivation
Jack: we need a motivation
<Zakim> JacekK, you wanted to talk about the tech problem with input/output for prec/eff
JacekK: using input and outputs annotations for pre and post conditions would probably be inapropriate
... the issue falls down to an ontological modelling issue
ajith: shall we enforce any specific interpretation?
laurenth: still think it would be useful, we can't ignore that operations have inputs and outputs
Jack: modelReference can be used there, do we need a specific mention to inputs and outputs?
ajith: no addition to the spec seems to be necessary
RESOLUTION: issue CR3 resolved with no modification to the spec
jack: not refer in the abstract to WSDL 2.0
... make clear the spec is not just for wsdl 2.0
RESOLUTION: issue CR 4 resolved as suggested
Rama: all the things discussed have been implemented
... effects to be added yet
JacekK: we will discuss this on the next call, please send email when updates are done