RE: [wsrp-wsia] Issue #51 - Naming scheme

[Monica Martin <mmartin@certivo.net>]

Wouldn't this also be an opportunity to make a comment or suggestion
back on WSDL under W3C?

Thanks.

-----Original Message-----
From: Rich Thompson [mailto:richt2@us.ibm.com]
Sent: Tuesday, October 08, 2002 6:04 AM
To: wsrp-wsia@lists.oasis-open.org
Subject: Re: [wsrp-wsia] Issue #51 - Naming scheme







I agree this modification scopes different bindings for a factor better.
Andre Kramer has also privately replied that using '.' as a delimiter
leads
to some mangled names in the .Net tools, but that using '_' for the same
role doesn't.  This would lead to:

   SPEC_VERSION_FACTOR_WSDLTYPE_TYPESPECIFIC


The fact that SSL isn't indicated until one gets to the soap:address
element in the port definition is a WSDL choice. We should live with it
rather than inventing something else and then having to deal with it
getting out of sync with the soap:address specification.



 

                      Richard

                      Cieply/Germany/IB        To:
wsrp-wsia@lists.oasis-open.org                          
                      M@IBMDE                  cc:

                                               Subject:  Re: [wsrp-wsia]
Issue #51 - Naming scheme               
                      10/08/2002 04:11

                      AM

 

 





Hi Rich,

I would choose a different order for the bindings like:
WSRP.v1.ServiceDescription.Binding.SOAP.
This seems to be more consistent with WSRP.v1.Markup.PortType.
Furthermore it is more efficient for the search.
I think one search use case could be to find _all_ ServiceDescription
bindings no matter what transport they use and then choose the one that
is
compatible with the one the consumer supports.
Therefor I would propose a slightly changed scheme:

SPEC.VERSION.FACTOR.TYPE.TYPESPECIFIC

where type is PortType|Binding and type specific is SOAP, DIME, MIME,
whatever we may have...

One additional question arises for me here:
How should we reflect secure/unsecure bindings like HTTP vs. HTTPS.
The binding definition in the WSDL would remain the same for both. The
only
distinction between secure/unsecure bindings is the protocol identifier
in
the
location attribute of the ../wsdl/soap:address element of the port
definition.

Richard




|---------+---------------------------->
|         |           Rich             |
|         |           Thompson/Watson/I|
|         |           BM@IBMUS         |
|         |                            |
|         |           10/07/2002 07:20 |
|         |           PM               |
|         |                            |
|---------+---------------------------->
  >
------------------------------------------------------------------------
------------------------------------------------------------------------
--|

  |
|
  |       To:       wsrp-wsia@lists.oasis-open.org
|
  |       cc:
|
  |       Subject:  [wsrp-wsia] Issue #51 - Naming scheme
|
  |
|
  |
|
  >
------------------------------------------------------------------------
------------------------------------------------------------------------
--|









While this issue explicitly raised how the naming scheme should be
chosen
for use when publishing to UDDI, it is important to consider the naming
scheme for the factors/bindings. Draft v0.7 uses this scheme:

          SpecIindicator '.' VersionNumber '.' FactorName '.'
(DistinguishingFactor '.') WSDLname

as in:
 WSRP.v1.Markup.PortType
 WSRP.v1.ServiceDescription.SOAP.Binding

To me this provides the scoping in the right order
(spec/version/factor/WSDLDescriptionPortion).

Proposed resolution: Accept v0.7 naming scheme



----------------------------------------------------------------
To subscribe or unsubscribe from this elist use the subscription
manager: <http://lists.oasis-open.org/ob/adm.pl>





----------------------------------------------------------------
To subscribe or unsubscribe from this elist use the subscription
manager: <http://lists.oasis-open.org/ob/adm.pl>





----------------------------------------------------------------
To subscribe or unsubscribe from this elist use the subscription
manager: <http://lists.oasis-open.org/ob/adm.pl>