Hallo Ant

I found the problem. And I reckon it is either a bug or not yet 
implemented.

Your test case defines the topic through the jndi.properties. If you 
comment the 'jndi topic declaration' and change the 'create' attribute to 
'always' in the 'destination' declaration, a queue instead of a topic gets 
generated. I have attached an extended version of your test case, which 
aims proofs this (patch attached).

Regards
Alex






"ant elder" <[EMAIL PROTECTED]> 
Gesendet von: [EMAIL PROTECTED]
18/11/2008 19:21
Bitte antworten an
[EMAIL PROTECTED]


An
[EMAIL PROTECTED]
Kopie
[email protected]
Thema
Re: Re: binding.jms & destination type 'topic'






You were correct, the warning you see is out dated and a red herring,  the 
code producing the warning looks like it was added during development 
before the type attribute was implemented and after it was implemented the 
warning didn't get correctly removed, so the warning still comes out even 
though the type attribute is processed correctly and using topics should 
work. Sorry for that confusion.

I've added a test to show topics - 
https://svn.apache.org/repos/asf/tuscany/branches/sca-java-1.x/itest/jms-topics/


Can you compare that test code to the code you have to see if there are 
any obvious differences that could be causing a problem? If you want to 
send a zip of your code I'd be happy to have a look for whats going wrong.

   ...ant
 
On Tue, Nov 18, 2008 at 3:55 PM, <
[EMAIL PROTECTED]> wrote:

Hi Ant 

What I did: 
- Got a snapshot of the tuscany sources (not today, already a couple of 
days back). 
- Loaded everything into Eclipse,  including 
        - sample-helloworld-service-jms 
        - sample-helloworld-reference-jms 
- once I had established that these two communicate in their vanilla 
state, I applied some changes: 
        - disabled the embedded ActiveMQ instance, providing my own 
        - changed the HelloWorldService.getGreetings to return void 
        - marked it with org.osoa.sca.annotations.OneWay 
        - added 'type="topic"' to all '<destination.../>' definitions in 
the composite declarations 
        - commented everywhere the declaration of the '<response.../>' 
declaration 

Now if I run either one of the projects, I get messages of the kind: 
18-Nov-2008 16:40:51 
org.apache.tuscany.sca.binding.jms.impl.JMSBindingProcessor 
WARNING: JMS Binding doesn't process destination type yet 

Additionally, looking into ActiveMQ via JMX reveals that a 'queue' is 
being created instead of a 'topic'. 

Now I had a quick look into the tuscany code as well, but not too deeply. 
The JMSBindingProcessor actually seems to parse the 'type' argument, but 
still complains about it (see above log message). I did not look for the 
place where the 'JMS destination' is created, and at a first glance simply 
assumed, "hmm, maybe the warning is simply outdated". Anyhow, looking into 
ActiveMQ somehow proofs that a queue is being created. 

Beside that there are no exceptions. Everything seems to be working (since 
so far I only run once service consumer ...). 

Regards
Alex 

PS: I could provide a zip with the two projects if you require, just takes 
some time on my side to simply verify the above again (did not expect 
anyone to contest the 'maybe' missing functionality question).



"ant elder" <[EMAIL PROTECTED]> 
18/11/2008 16:27 


Bitte antworten an
[EMAIL PROTECTED]


An
[email protected] 
Kopie
[EMAIL PROTECTED] 
Thema
Re: binding.jms & destination type 'topic'










On Mon, Nov 17, 2008 at 9:53 AM, <
[EMAIL PROTECTED]> wrote: 

Hi All 

I have tried to use the binding.jms via topics (even with the current head 
version). However, this seems not to be implemented yet. My question now 
is: Are there any plans to implement this in the very very near future, or 
would this require a bigger effort that it has not been tackeled yet? 

Regards 
Alex Chatziparaskewas 

I thought this should work, what are the problems you see when using 
topics, could you give a bit more detail, stack traces or error messages?

  ...ant 

Attachment: patch.patch
Description: Binary data

Reply via email to