Also, I'm wondering if we might be able to do something Wih the no-op as well. For instance, can our XmlHttpResponseWriter detect whether content has been rendered before the response complete and? If it hasn't, send the noop response.
Don't know if it's possible but simply needing to set FacesContext.responseComplete() to do the noop would be less error prone. On Mar 13, 2011, at 9:18 AM, Scott O'Bryan <darkar...@gmail.com> wrote: > I like the idea of having three more simple API's although I think the > "redirect" is unnecessary.. That should be handled correctly by > ExtenalContext.redirect, even in PPR environments. > > On Mar 13, 2011, at 8:57 AM, Andy Schwartz <andy.g.schwa...@gmail.com> wrote: > >> Btw - another variant would be to drop the "Response" suffix from the method >> names. >> >> Andy >> >> On Mar 13, 2011, at 10:48 AM, Andy Schwartz <andy.g.schwa...@gmail.com> >> wrote: >> >>> Hi Bernd - >>> >>> I am a bit hesitant about the names as well. Three possible options: >>> >>> 1. >>> completeNoopResponse()/completeScriptResponse()/completeRedirectResponse(). >>> 2. renderNoopResponse()/renderScriptResponse()/renderRedirectResponse(). >>> 3. sendNoopResponse()/sendScriptResponse()/sendRedirectResponse(). >>> >>> Think #2 and #3 read better than #1. If we go with #2 or #3 we would >>> make it clear that these complete the response by documenting that >>> FaceContext.responseComplete() is called. >>> >>> Andy
