[jboss-user] [JBossWS] - Re: WS-Eventing - sending notification

2007-01-08 Thread palin
"[EMAIL PROTECTED]" wrote : I did create a JIRA for this: | http://jira.jboss.com/jira/browse/JBWS-1388 Hi Heiko, first of all sorry for the long silence, but we (Stefano and me) have been quite busy. Soon we'll work on the ws-eventing clustering issues; in the mean time, speaking of this issu

[jboss-user] [JBossWS] - Re: WS-Eventing - sending notification

2006-11-16 Thread [EMAIL PROTECTED]
I did create a JIRA for this: http://jira.jboss.com/jira/browse/JBWS-1388 View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=3986659#3986659 Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=3986659 _

[jboss-user] [JBossWS] - Re: WS-Eventing - sending notification

2006-11-16 Thread [EMAIL PROTECTED]
Well, let's do it then. I would simply call it DispatchException. View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=3986658#3986658 Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=3986658

[jboss-user] [JBossWS] - Re: WS-Eventing - sending notification

2006-10-30 Thread artem_spector
Just a short clarification from my side. I mean the wsdl deployed on the server defines the contract between the event source and the event consumer. So yes, the validation should be done against the schema on the server. View the original post : http://www.jboss.com/index.html?module=bb&op=vi

[jboss-user] [JBossWS] - Re: WS-Eventing - sending notification

2006-10-26 Thread maeste
One more question: I suppose you are thinking the dispatch method must throw an exception? Spec doesn't define one for this case, have you any prefered name or hirearchy? If you get some times take a look at my post on design forum because the modifications proposed there could be done in the sa

[jboss-user] [JBossWS] - Re: WS-Eventing - sending notification

2006-10-26 Thread maeste
Hi Heiko, I agree with you and Alessio and me could work on this stuff. But we need just a clarification: we you thinking to validate against the schema defined in wsdl by the server. Do you agree? I'm asking about because probably I didn't understand what artem means with: "but aparently it mi

[jboss-user] [JBossWS] - Re: WS-Eventing - sending notification

2006-10-26 Thread [EMAIL PROTECTED]
I think artems suggestion is perfectly valid. The SubscriptionManager knows about the schema and could actually validate the event messages. At least this was my original intention. But as palin already realized, there's two things to be done: The MetaDataBuilder needs to extract the schema (t

[jboss-user] [JBossWS] - Re: WS-Eventing - sending notification

2006-10-26 Thread maeste
Hi artem, I can't understand what you mean? Specification of ws-eventing doesn't constrain notifications because (specification said) any message MAY be a notification. But in real world producer and receiver have to constrain the message accepted. And it is an application level know how. Of c

[jboss-user] [JBossWS] - Re: WS-Eventing - sending notification

2006-10-26 Thread artem_spector
Yes, I know the notification schema, but aparently it might be different from the instance loaded by the subscription manager. Therefore the "real" validation can be done only by the engine - or by the application if the engine provides an access to that schema (by URI) thank you, Artem View t

[jboss-user] [JBossWS] - Re: WS-Eventing - sending notification

2006-10-25 Thread palin
Hi Heiko, I've just taken a look at the code: it seems to me that the patch at the SubscriptionManager's dispatch method could be done, validating the received element against the right notification schema from eventSourceMapping. However I noticed that notification schema is currently always se

[jboss-user] [JBossWS] - Re: WS-Eventing - sending notification

2006-10-25 Thread [EMAIL PROTECTED]
No, it is just not implemented ;) Actually the EventDispatcher should validate the message before dispatching. Feel free to open a JIRA feature request so people can volunteer. View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=3980670#3980670 Reply to the post :

[jboss-user] [JBossWS] - Re: WS-Eventing - sending notification

2006-10-25 Thread artem_spector
Yes, it is simple. But is it consistent? Currently my notification producer builds messages which are not validated against the wsdl, and the endpoint accepts the messages without validate them against the wsdl. Moreover, if I'd wish to perform such validation on either side, I have no tools to

[jboss-user] [JBossWS] - Re: WS-Eventing - sending notification

2006-10-25 Thread [EMAIL PROTECTED]
I see what you are trying to archieve. The way it is designed, things are kept really simple and nofitifcation producer doesn't have a dependency on web services at all. View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=3980655#3980655 Reply to the post : http

[jboss-user] [JBossWS] - Re: WS-Eventing - sending notification

2006-10-12 Thread palin
Unfortunately as far as I know/understand what you would like is not provided by jbossws because not described in any specifications (jsr181,jsr109,...) bye Alessio Soldano http://www.javalinux.it/blogs View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=3977834#397

[jboss-user] [JBossWS] - Re: WS-Eventing - sending notification

2006-10-12 Thread artem_spector
Alessio, thank you for your explanation. These exactly are the things which seem strange to me. "palin" wrote : I think you can create that xml string the way you prefer, it doesn't matter how you do it as long as what you get is in compliance with the schema you provided in you wsdl; for examp

[jboss-user] [JBossWS] - Re: WS-Eventing - sending notification

2006-10-12 Thread palin
"artem_spector" wrote : Something is unclear to me after reading the user guide: | | 1. Notification payload: | The example shows the payload creation via parsing some xml string - is there a better way to do it? I think you can create that xml string the way you prefer, it doesn't mat