On Sat, May 2, 2009 at 6:22 PM, Raymond Feng <[email protected]> wrote: > Good catch. I was about to turn the FaultExceptionMapper into a declared > utility instead of creating it in the activator code. > > We can safely remove the Mediator lookup in JavaRuntimeModeulActivator. I > just committed a change for that. > > Thanks, > Raymond > -------------------------------------------------- > From: "Simon Laws" <[email protected]> > Sent: Saturday, May 02, 2009 9:34 AM > To: <[email protected]> > Subject: Re: svn commit: r770532 - in > /tuscany/branches/sca-java-1.x/modules: > core-databinding/src/main/java/org/apache/tuscany/sca/core/databinding/module/ > core-databinding/src/main/java/org/apache/tuscany/sca/core/databinding/wire/ > databinding/src/main/java > >> >>> >> >> There's a line added in this commit... >> >> + >> registry.getExtensionPoint(UtilityExtensionPoint.class).addUtility(faultExceptionMapper); >> >> Mediator mediator = >> >> registry.getExtensionPoint(UtilityExtensionPoint.class).getUtility(Mediator.class); >> >> which makes the fault exception mapper available when the Mediator >> initializes. The Mediator is also retrieved in the >> JavaRuntimeModeulActivator. Am going to try commenting it out in >> JavaRuntimeModeulActivator and leave this one in as retrieving the >> Mediator in two activators gives unpredictable results depending on >> which one starts first. >> >> Not sure it there is a reason that the JavaRuntimeModeulActivator gets >> the Mediaor. It doesn't seem to use it directly. I guess I'll soon >> find out. >> >> Simon > >
Thx Raymond
