WSDL 1.1 doesn't *require* that a schema have a targetNamespace. The BP does, but that's a best practice, not a spec requirement. WSDL 2.0 requires that schemas have a targetNamespace, but that doesn't change WSDL 1.1.
The better solution is to throw a warning if the targetNamespace isn't there --let XMLBeans throw the error.
Anne
On 3/27/06, Dennis Sosnoski
<[EMAIL PROTECTED]> wrote:
That just BP - which everybody likes, but AFAIK Axis2 is not enforcing.
- Dennis
Davanum Srinivas wrote:
>Sanjiva,
>
>Is it this one?
>
>http://www.ws-i.org/Profiles/BasicProfile-1.0-2004-04-16.html#refinement16552904
>
>-- dims
>
>On 3/26/06, Dennis Sosnoski <[EMAIL PROTECTED]> wrote:
>
>
>>Sanjiva Weerawarana wrote:
>>
>>
>>
>>>On Mon, 2006-03-27 at 14:59 +1200, Dennis Sosnoski wrote:
>>>
>>>
>>>
>>>
>>>>The resolution of http://issues.apache.org/jira/browse/AXIS2-67 was
>>>>to
>>>>add a check that the schema definition within a WSDL defines a
>>>>targetNamespace. What is the justification for this? AFAIK it's
>>>>
>>>>
>>>>
>>>>
>>>The justification is that its required by WSDL. WSDL 1.1 wasn't clear
>>>but it has since been clarified.
>>>
>>>
>>>
>>>
>>So the requirement was added after WSDL 1.1 - is it spelled out
>>somewhere, or just an informal interpretation? I'm happy enough to
>>accept it either way, as long as other frameworks such as .NET require
>>schema namespaces, but it'd be nice to have something to point to by way
>>of justification.
>>
>>The current code throws an NPE if a schema doesn't have a
>>targetNamespace rather than reporting anything useful - that's what led
>>me into this. I'll fix that.
>>
>> - Dennis
>>
>>
>>
>>
>
>
>--
>Davanum Srinivas : http://wso2.com/blogs/
>
>
>
