Hi, cool +1!
1) yep, push should be already done 2) the server should already only render real new resources, if it's not happening, we should fix it on the server side 2017-10-08 16:35 GMT+02:00 Werner Punz <[email protected]>: > Am 08.10.17 um 16:17 schrieb Werner Punz: > > Hi I am almost done with my changes however I wanted to clarify some >> things: >> >> the jsf.push part, Leo wants do do that, right? If not then give me >> notice, it shouldn´t be to hard, the api seems rather straight forward. >> >> Secondly I noticed in the spec a special section and dug around in the >> mailing lists without coming to a decisive conclusion here: >> >> If an update element is found in the response with the identifier >> javax.faces.Resource: >> <update id="javax.faces.Resource"> >> <![CDATA[...]]> >> </update> >> append any element found in the CDATA contents which is absent in the >> document to the document's head section. >> >> Now up until now the behavior on our side was, that if just a resource >> element is coming in, it simply was appended. >> >> The question here is more along the lines, that who is performing the >> checks on those elements. >> >> The best bet here is probably to diff on the server for the resource >> appendix, where all the meta info is available, because from a client point >> of view >> this is just a big, character blob, which then would have to be pushed >> into a shadow dom element for browser interepretation. >> >> And from that point forward I do not have a 100% precise detection >> anymore whether a resource is new in the form( some refs might be the same >> but thats mostly what I have, with embedded CSS or javascript there is no >> way to detect for equality on my level anymore there always will be some >> case where it can fail) >> >> >> >> >> Ah I just checked the latest api code, the socket stuff seems to be in > already. > > Nice job. > > Cheers > > Werner > > >
