Why are port and operations QNames in the jmx ON?

2008-09-05 Thread Hadrian Zbarcea 

Hi,

I noticed that in the ObjectNames for PerformanceCounter the port and  
operation are modeled as QName instead of String.  I believe that is  
totally unnecessary and only makes the ONs long and harder to read.   
Is there a reason for the port and operation to be qnames?  The  
service is a QName and defines the namespace already.


If you agree with this change I can contribute a patch.


org 
.apache 
.cxf:bus.id=cxf945122723,type=Performance.Counter.Server,service="{http://apache.org/hello_world_soap_http 
}SOAPService",port="{http://apache.org/hello_world_soap_http}SoapPort";


Thanks
Hadrian

Re: Why are port and operations QNames in the jmx ON?

2008-09-05 Thread Daniel Kulp 
On Friday 05 September 2008 10:25:34 am Hadrian Zbarcea wrote:
> Hi,
>
> I noticed that in the ObjectNames for PerformanceCounter the port and
> operation are modeled as QName instead of String.  I believe that is
> totally unnecessary and only makes the ONs long and harder to read.
> Is there a reason for the port and operation to be qnames?  The
> service is a QName and defines the namespace already.
>
> If you agree with this change I can contribute a patch.

Sure.  Contribute a patch.   :-)


>
> org
> .apache
> .cxf:bus.id=cxf945122723,type=Performance.Counter.Server,service="{http://a
>pache.org/hello_world_soap_http
> }SOAPService",port="{http://apache.org/hello_world_soap_http}SoapPort";
>
> Thanks
> Hadrian



-- 
Daniel Kulp
[EMAIL PROTECTED]
http://www.dankulp.com/blog

WS-RM issue

2008-09-05 Thread Bharath Ganesh 
https://issues.apache.org/jira/browse/CXF-1156

Look like WS-RM makes more sense for aysncronous invocation and one way
invocation (quite similar to JMS reliability), rather than a standard
request-response pattern.
So the above issue wont be a very valid case. Am I right?

Re: WS-RM issue

2008-09-05 Thread Eoghan Glynn 


Dunno if I'd agree with this JIRA, if I've understood it correctly.

For a request-response MEP, WS-RM can be configured so that a "202 
Accepted" response is immediately sent back to the client (possibly 
including an eager ACK) and then whenever it becomes available the real 
response is sent over a separate server->client connection.


/Eoghan


Bharath Ganesh wrote:

https://issues.apache.org/jira/browse/CXF-1156

Look like WS-RM makes more sense for aysncronous invocation and one way
invocation (quite similar to JMS reliability), rather than a standard
request-response pattern.
So the above issue wont be a very valid case. Am I right?




IONA Technologies PLC (registered in Ireland)
Registered Number: 171387
Registered Address: The IONA Building, Shelbourne Road, Dublin 4, Ireland

Re: WS-RM issue

2008-09-05 Thread Bharath Ganesh 
So that is kind of an asynchronous invocation. Not a synchronous
request-response pattern.That again means the JIRA is not a very valid use
case. If convinced, I shall close the JIRA.


On Sat, Sep 6, 2008 at 12:42 AM, Eoghan Glynn <[EMAIL PROTECTED]> wrote:

>
> Dunno if I'd agree with this JIRA, if I've understood it correctly.
>
> For a request-response MEP, WS-RM can be configured so that a "202
> Accepted" response is immediately sent back to the client (possibly
> including an eager ACK) and then whenever it becomes available the real
> response is sent over a separate server->client connection.
>
> /Eoghan
>
>
>
> Bharath Ganesh wrote:
>
>> https://issues.apache.org/jira/browse/CXF-1156
>>
>> Look like WS-RM makes more sense for aysncronous invocation and one way
>> invocation (quite similar to JMS reliability), rather than a standard
>> request-response pattern.
>> So the above issue wont be a very valid case. Am I right?
>>
>>
> 
> IONA Technologies PLC (registered in Ireland)
> Registered Number: 171387
> Registered Address: The IONA Building, Shelbourne Road, Dublin 4, Ireland
>

Re: Why are port and operations QNames in the jmx ON?

2008-09-05 Thread Willem Jiang 

Hi Hadrian,

The port and operation could have different target namespace against the 
the service.
Maybe we could remove the namespace of port and operation in the 
ObjectName if they are same with the service's target namespace.


Regards,

Willem
Hadrian Zbarcea wrote:

Hi,

I noticed that in the ObjectNames for PerformanceCounter the port and 
operation are modeled as QName instead of String.  I believe that is 
totally unnecessary and only makes the ONs long and harder to read.  
Is there a reason for the port and operation to be qnames?  The 
service is a QName and defines the namespace already.


If you agree with this change I can contribute a patch.


org.apache.cxf:bus.id=cxf945122723,type=Performance.Counter.Server,service="{http://apache.org/hello_world_soap_http}SOAPService",port="{http://apache.org/hello_world_soap_http}SoapPort"; 



Thanks
Hadrian