Inside your managed beans you can use FacesContext context = FacesContext.getCurrentInstance(); context.getApplication().getVariableResolver().resolveVariable(context, name);
Where 'name' is the managed-bean name. IMO this feature is not well known but it seems to be quite handy! (I haven't used it yet but I will be in the next version of our project.) I'm not sure what happens if the bean doesn't exist yet but it sounds like you were going to have things set up. I believe VariableResolver may create a new instance of the managed bean for you if it hasn't been created yet but I'm not sure. Also, Craig McClanahan is doing some interesting stuff with his Struts Shale project. He's added extra lifecycle and processing capabilities to the JSF lifecycle. There are some interesting options for initialization there as well. HTH, sean On Thu, 27 Jan 2005 23:25:35 +0100, Werner Punz <[EMAIL PROTECTED]> wrote: > mfaine wrote: > > > > >What is the best way to pass around request scope managed beans? > > > >I have managed beans that needs to populated using values from other > >(previous) request scope managed beans. The beans have to be request scoped > >in order for the application to work correctly. > > > > > > > I have implemented a session scoped bean cache for such purposes... > Where beans can be dropped in and removed on purpose... > > (Currently as a size limited fifo cache, so that I dont run into memory > issues, a timed cache also would be feasable) > >
