+ for the solution proposed by asankha, That is to have an optional scope attribute so that we can get rid of putting all our local message properties to Axis2 MC which I dont see as a good practice. It is also nice to specify the corelation as a scope....
Thanks, Ruwan. On 12/15/06, Asankha C. Perera <[EMAIL PROTECTED]> wrote:
We use the local message properties, esp for the set-property mediator and the getProperty() XPath function etc.. I propose that we modify the set-property mediator as follows: <set-property name="string" (value="literal" | expression="xpath") [scope=axis2|correlate]/> Then we could use the same for correlations of messages as well.. asankha Paul Fremantle wrote: > I think I prefer #1. If we aren't using the distinction between the > Synapse MC property and the Axis2 MC then maybe we should get rid of > it? > > Paul > > On 12/14/06, Asankha C. Perera <[EMAIL PROTECTED]> wrote: >> >> Tijs >> >> 0.90 is now released. We could do another point release in January >> and/or >> prepare for 1.0. However I think quick release cycles would be good >> for us - >> and hence your fix should become available soon. >> >> From the proposed solutions, I prefer option 2... and if you would >> like to >> do this enhancement, you are most welcome >> >> asankha >> >> >> [EMAIL PROTECTED] wrote: >> Paul, >> >> I think optiion 1 would be the most effective one. >> If this is not possible, option 2 is a good alternative. >> Do you think this can be resolved before release version 0.90? >> It would be a bit disappointing if the WS-RM support is not fully >> working in >> version 0.90. >> I have some time to help out if this is helpful. >> >> Tijs >> >> >> >> van: "Paul Fremantle" <[EMAIL PROTECTED]> >> datum: 2006/12/14 Thu PM 06:21:39 MET >> aan: [email protected] >> onderwerp: Re: WS-RM TerminateSequence message >> >> Tijs >> >> I figured out what's going on. Although our MessageContext is a >> mapping onto the underlying Axis2 MC, we have our own properties too. >> So when we set-property, that doesn't get reflected in the underlying >> Axis2 MC, so it doesn't affect Sandesha2. >> >> Three simple options I can see: >> >> 1) get rid of our properties and just have the axis2 mc properties.. >> If a user wants to avoid overwriting an underlying axis2 property, >> just make sure this has a unique name. I don't think we actually use >> this feature but maybe I'm wrong. >> 2) add a set-axis2-property which modifies the underlying axis2 >> property. >> 3) add a specific lastmessage mediator. >> >> Thoughts everyone? >> >> Paul >> >> On 12/14/06, [EMAIL PROTECTED] <[EMAIL PROTECTED]> wrote: >> >> >> Paul, >> >> Has there been some progression with the WS-Reliable Messaging last >> message >> for the termination of the sequence? >> >> Thanks, >> >> Tijs >> >> >> Tijs >> >> You are quite right. The set-property isn't working. I'm just >> debugging it. >> The property should be going right through to the sandesha module >> which is why you can't find the code in Synapse. >> >> >> Paul >> >> On 12/8/06, [EMAIL PROTECTED] <[EMAIL PROTECTED]> wrote: >> >> >> Paul, >> >> I have tried to use the set-property element, but I don't see any >> change in >> the WS-RM >> >> handling. >> >> >> Where exactly does the set-property needs to be set in the Synapse flow >> configuration >> >> file? >> >> >> >> --------------------------------------------------------------------- >> To unsubscribe, e-mail: >> [EMAIL PROTECTED] >> For additional commands, e-mail: [EMAIL PROTECTED] >> >> >> >> -- >> Paul Fremantle >> VP/Technology, WSO2 and OASIS WS-RX TC Co-chair >> >> http://bloglines.com/blog/paulfremantle >> [EMAIL PROTECTED] >> >> "Oxygenating the Web Service Platform", www.wso2.com >> >> --------------------------------------------------------------------- >> To unsubscribe, e-mail: >> [EMAIL PROTECTED] >> For additional commands, e-mail: [EMAIL PROTECTED] >> >> >> >> >> --------------------------------------------------------------------- >> To unsubscribe, e-mail: >> [EMAIL PROTECTED] >> For additional commands, e-mail: [EMAIL PROTECTED] >> >> >> >> >> --------------------------------------------------------------------- >> To unsubscribe, e-mail: >> [EMAIL PROTECTED] For additional >> commands, e-mail: [EMAIL PROTECTED] > > --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
