Re: [wsrp-interfaces] Re: [wsrp-webservice] [wsrp-webservices] WSDLcall today

[Michael Freedman <michael.freedman@oracle.com>]

Yes, that is the issue.  An example would be v2 supporting registration 
while v1 didn't not [or vice versa].  I am suggesting we be specific in 
not allowing the use of mixed versions as this will get even more 
complicated in the future now that we are adding optional calls in each 
port.
    -Mike-

Subbu Allamaraju wrote:

> Is the question about mixing ports (say, some from v1 and some from v2)?
>
> > across a producer that decided to publish distinct services per WSRP
> > port -- which I expect most consumers not deal with correctly.
>
> Any example?
>
> Subbu
>
> Michael Freedman wrote:
>
>> "The service definition contains both v1 & v2 ports. Personally, I 
>> would prefer a separate service for v1 and v2 ports."
>>
>>
>> One issue I have been meaning to post to the overall group is the 
>> question on whether the specification needs to say anything about how 
>> a consumer interacts with a service that contains mixed versions 
>> [ports]. I know we assume that consumers will choose to either only 
>> use the v1 or the v2 ports but should we have language that directs 
>> them [or strongly suggests they] to do this?  I ask in part because 
>> we have recently come across a producer that decided to publish 
>> distinct services per WSRP port -- which I expect most consumers not 
>> deal with correctly.
>>     -Mike-
>>
>>
>> Andre Kramer wrote:
>>
>>> The SC held a call to discuss the XSD schema and WSDL for draft 11.
>>>
>>> Attendees: Subbu, Clinton, Mike & Andre
>>>
>>> Andre walked through the .xsd & .wsdl files pointing out changes and 
>>> potential issues:
>>>
>>> The service definition contains both v1 & v2 ports. Personally, I 
>>> would prefer a separate service for v1 and v2 ports.
>>>
>>> Type extension is now used.
>>>
>>> String type restriction is not used (types declared but elements 
>>> still use string). This really is a pity but we are unlikely to get 
>>> all tools to work otherwise.
>>>
>>> Choice (union-like content model) is unused and may be unacceptable 
>>> in the lifetime scope of v2.
>>>
>>> Use of namespaces is simplified (mirroring what we had before in v1) 
>>> but people should check that they can handle (types from) both v2 & 
>>> v1 namespace if they need to.
>>>
>>> Testing (no issues reported):
>>>
>>> Andre and others have tried VS 2003.
>>>
>>> Clinton will try Axis 1.2 and VS 2005 (if he can find a copy).
>>>
>>> Subbu will test with internal tooling.
>>>
>>> [IBM will have some testing.]
>>>
>>> We request that Atul could test against the current JAX-RPC 
>>> reference implementation. Thanks.
>>>
>>> Other:
>>>
>>> We are unlikely to test MTOM/XOP any time soon.
>>>
>>> We should have one more call before the committee draft. Fine 
>>> checking of the WSDL / spec match still needs to be done. Andre 
>>> found some small inconsistencies but other (faults in particular) 
>>> are unlikely to have all been picked up.
>>>
>>> Regards,
>>>
>>> Andre
>>>
>>
>> --------------------------------------------------------------------- 
>> To unsubscribe from this mail list, you must leave the OASIS TC that 
>> generates this mail. You may a link to this group and all your TCs in 
>> OASIS at: 
>> https://www.oasis-open.org/apps/org/workgroup/portal/my_workgroups.php
>
>