On Mon, Sep 20, 2010 at 6:46 AM, Simon Laws <[email protected]> wrote: > Well I guess we need to be a bit more precise about what the likely > impact is. IIUC from the commit the model is that > the JDKInvoker and implementation invoker implement the transition > between the local (to them) component implementation and Tuscany's > internal representation of holders. Between these and the binding > there are a set of interceptors on both reference and binding sides. > So potentially any interceptor, and the binding runtimes themselves > of course, could be impacted by the presence of holders. >
Yes, and this is what is worrisome about the current design. > > Doesn't even one parameter as a holder imply two return values? One > for the holder and one for the return (which would also be a holder > according to JAXWS). > Yes, but the 1.x support only worked with void return types and a single Holder. Basically, the return value became the Holder rather than whatever was returned from the method invocation. Brent
