Hi Greg, AFAIK there is no "bye bye" ;). ServletContainerInitializer are not in he whiteboard so you make them behave as you want and any servlet container has 2 impl of the servlet context, one for SCI and one for the runtime (where most registration methods throw exceptions). So it means pax-web can keep being whiteboard++ (ie whiteboard + SCI) by having these 2 (sadly custom) impl too and rewire it to the runtime properly. I fear you will need to do as for http felix impl - ie a filter chain dispatcher - but this is not crazy and solves all the issues you mentionned enabling to have all the desired features. Indeed, pax wouldnt 1-1 delegate to the servlet container but it would be http whiteboard compatible and still support JSF, JSP, etc..
Hope it makes sense. Romain Manni-Bucau @rmannibucau <https://twitter.com/rmannibucau> | Blog <https://rmannibucau.metawerx.net/> | Old Blog <http://rmannibucau.wordpress.com> | Github <https://github.com/rmannibucau> | LinkedIn <https://www.linkedin.com/in/rmannibucau> | Book <https://www.packtpub.com/application-development/java-ee-8-high-performance> Le lun. 31 août 2020 à 08:29, Jean-Baptiste Onofre <j...@nanthrax.net> a écrit : > Hi Grzegorz > > Enjoy your vacation, and thanks for the update about Pax Web 8. > > Waiting for your return, let me try to move forward on JSP/JSF support. > > I will keep you posted. > > Regards > JB > > > Le 31 août 2020 à 08:24, Grzegorz Grzybek <gr.grzy...@gmail.com> a > écrit : > > > > Hello > > > > I'm on PTO till Tuesday, I'm back to work on Sep 2nd. > > > > About Pax Web 8 - indeed, I don't want to hold Karaf 4.3.0 any longer. > > Whiteboard pattern is actually ready (everything defined in Whiteboard > spec > > is in place except tests for Preprocessors and per-context session > > handling). The problem is that what Pax Web 7 already had is still > missing. > > > > Long story short - Pax Web 7 handled ServletContainerInitializers well > and > > it's the fundamental thing to handle JSPs and JSFs. But what I had to > > change for Whiteboard compliance is actually a bit incompatible with > those > > customizers (most importantly - dynamic "ServletContext.addServlet()" > > according to Whiteboard specification should throw > > UnsupportedOperationException - so bye bye JSFs). > > > > I already have an idea how to "fix" this, but this will again delay Pax > Web > > 8 a bit. > > > > regards > > Grzegorz Grzybek > > > > pon., 31 sie 2020 o 06:19 Jean-Baptiste Onofre <j...@nanthrax.net> > napisał(a): > > > >> Hi guys, > >> > >> A quick update about the releases schedule. > >> > >> For both Karaf runtime 4.2.10 and 4.3.0.RC2, I have some dependency > >> releases in progress (Felix FileInstall, Pax Exam, etc). > >> I will submit these releases to vote this week. > >> > >> About 4.3.0.RC2 specifically, I will ping Greg today to see where we are > >> on Pax Web 8.0 and help him on that. > >> As I don’t want to hold 4.3.0 GA too much, we will evaluate a new time > the > >> effort for Pax Web 8.0 and eventually define Felix Jetty as default > >> (temporary). > >> > >> About Winegrower 1.0, I have several updates ready (I just have to > create > >> the PRs), a couple of new cepages to add. So the 1.0 release will be in > >> vote during this week. > >> > >> Regards > >> JB > >> > >>> Le 20 août 2020 à 15:23, Jean-Baptiste Onofre <j...@nanthrax.net> a > écrit > >> : > >>> > >>> Hi guys, > >>> > >>> As we are getting close to Pax Web 8.0, I think it’s the good timing to > >> release Karaf Runtime 4.3.0.RC2 with Pax Web 8.0-SNAPSHOT. > >>> It would give us time to move forward on 4.3.0 later faster. > >>> > >>> Karaf 4.3.0 is blocked by Pax Web for too long now, so, let’s move > >> forward at least on RC2. > >>> > >>> On the other hand, Karaf 4.2.10 is in preparation, as well as > Winegrower > >> 1.0 (I’m working on cepages). > >>> > >>> I will send some updates about new coming features on which we are > >> working (spring boot support, devx, flat resolver, and so on). > >>> > >>> Regards > >>> JB > >> > >> > >
