1 2 3 4 5 Previous Next 73 Replies Latest reply on Nov 26, 2011 10:21 AM by kaczortrex Go to original post
      • 15. Re: Seam.Next discussion
        brvuga

        For me modular design of seam 3 is not a problem. While using seam 2 i tried to move out the stuff that i didn't use. I like to keep my dependencies to a minimum. So i like being able to use only what i think i need (seam servlet, seam faces, seam security for the most part). So modular design was a bingo if you ask me.
        JavaEE 6 is much better than javaEE 5. It is so much better that you don't need a full blown framework as seam 2 was. You can program most of the stuff your self without that much effort. Writing extensions is really easy. Writing interceptors also. So for me keeping all the modules separated (just as it currently is) would be the right approach. Only thing that is currently lacking is more rigorous testing of things working together on all the containers.


        What i fear is that by moving some of the modules outside to other jboss products, it will be harder to find a combination of versions that works. I also fear of being locked to jboss products.

        • 16. Re: Seam.Next discussion
          shane.bryzak

          Excellent feedback so far, thanks everyone.  If you are holding back from letting us know your point of view because someone else has already posted what you are thinking, please don't let that stop you.  In fact, if you agree with something that someone else has said you should be reinforcing that point as the message to us will come through clearer (remember, we're trying to build something that benefits the community, not a bunch of separate individuals).

          • 17. Re: Seam.Next discussion
            mdesignz

            Seeing how the Seam team handled the JSF2 support for Seam2, I'm a little gun shy about jumping on the Seam3 bandwagon.  Assuming Seam3 will eventually provide similar functionality to Seam2, having adequate examples, documentation, and a Seam2 to Seam3 migration doc will be absolutely essential.

            • 18. Re: Seam.Next discussion
              ozgura

              I agree with Robert. I have been following numerous requests for JSF2 support in Seam2 for a really long time now (in the forums). I can't say there is yet a certainty around that delivery. 


              Seam2 is useful and useable for all intents and purpose in real-life projects. It has pretty much all necessary components/libraries that I needed so far. I can find all those tools I need in a single package - and most importantly, they work out-of-box. To me, that was the attractiveness of Seam2 - I didn't have to go off to collect myriad of different packages from different sources (maybe some prefer this sort of approach, but that's undesirable to me).


              The way I see it, decentralizing Seam3 means not having Seam; probably that would make me consider another framwork or technology instead (again, this applies to my situation, others might think differently). I'd prefer to see a centralized Seam3 with Excel, PDF, Cron, etc. all working out-of-box just like Seam2; then I will definitely consider moving to Seam3.

              • 19. Re: Seam.Next discussion

                Here another heavy and happy Seam 2 user.


                I just can repeat what most people asked here.


                Examples,examples,examples, maybe starting off the examples of Pete Muir for JBoss 7 (kitchensink) but a bit more elaborating on other modules of Seam too.


                Keep it bundled so one can be assured that all modules work properly together on certain platforms.(Jboss 6, 7, glassfish,tomcat)


                Keep providing integration with more than just jboss stuff.


                Hopefully Seam 3 still has a bright future (the foundation is there)


                PS A book will help too, Seam in Action was a brilliant guide for Seam2, thanks Dan. Another good book came from Jacob and Michael

                • 20. Re: Seam.Next discussion
                  vata2999

                  you'd better never call this project Seam again



                  The strength of Seam was that it was an integration solution of others technologies like a kind of cement

                  and you r ruining it

                  • 21. Re: Seam.Next discussion
                    petercr4

                    More or less everything was written above but anyway:


                    Seam2 framework was great for us (and still is) because:



                    • we could focus on business stuff not on thinking how to get a value from one layer to another. Here I don't mean only injection stuff also classes like EntityHome were important they gave you a path of how to do stuff. From what I've seen in Seam 3 lots of stuff we got used to is missing.






                    • documentation and examples are great







                    • everything was tested together you just picked the latest package and knew the jars will have right version







                    • framework is open, if you like IceFaces you can use that (we still prefer Richfaces), if your customer is using Websphere or Weblogic you have a recipe in documentation how to deploy stuff





                    My worries now are:





                    • not much talking about Seam until now. I've got a feeling that JBoss folks just don't care anymore. Ok Weld is here, but what about us?







                    • development on Seam 2 and also Richfaces 3 stopped too quickly, step to move to CDI and JSF 2.0 is too big if you have a large codebase to move. Richfaces 3 for example does not have support for IE9.





                    About migration:





                    • i don't expect tools to do that but some documentation would be great







                    • ideally migration should be in smaller steps, making Seam 2 work on Jboss7, making it JSF2 compatible to switch to RF 4 and this should be enough







                    • please be honest if you are going to still work on Seam 2 or just say the project ends here and we'll all know where we are. There were some promises to make JSF2 compatibility but nothing happened there for a while now




                    • 22. Re: Seam.Next discussion
                      antoine.antoine.abside.com

                      I wrote a blog post to give my point of view on Seam future :
                      Please JBoss don’t let CDI become the “Betamax”of Java by destroying Seam 3

                      • 23. Re: Seam.Next discussion
                        titou09

                        We are using Seam 2 in most of our web projects since a few years now. The team love it and we are very productive with it
                        I personally participate to have Seam 2 work with WebSphere v7.0 and I almost entirely rewrote the WebSphere chapter in the doc and the WebSphere samples with Seam 2 in addition to many commits


                        I'm in charge of the development team here and the way to go for us is to use WebSphere v8.0 that includes OpenWebBeans and OpenJPA


                        As explained in the mailing list my current tentative is a big failure : link here


                        Now that WAS v8.0.0.1 is out I was waiting for Seam 3.1 to make another try but I still have difficulties to know from where to start to build a prototype application and :



                        • most of the JIRA opened in June are still opened

                        • there is still no nice example to start from. The booking sample in Seam 2 was great for a beginning

                        • If you are not a maven shop (and we will never be), it is not that easy to get all the pieces together with sample and code fast

                        • I was thinking about participating to Seam 3 as I did for Seam 2 but another obstacle is there : git . I have first to learn git to participate and then to understand the Seam 3 module structure, owner etc.



                        Also I post a mail with ideas some time ago on how to reorganize the WiKi too to be more clear from where to start but nothing changed.


                        I want to have Seam 3 become a success but this is very confusing even for old user. I don't know for new users, but for old users the step is very very high, to summarize we have to eat and digest: maven, git, CDI, JSF 2.0, Weld, Seam 3, JBoss-logging and others this with the lack of easy sample to play with and all the incompatibilities with WebSphere, IBM JDK and even standards. Of course many of the new standards are not the fault of Seam 3 (JSF 2.0 CDI etc) but some choices are and add to the difficulty (maven ,git, JBoss-logging, logging frameworks confusions etc). lack of good sample, confusion on the WiKi/doc etc)


                        IMHO Seam 3.0 was an alpha version as the number of incompatibilities with standards are too numerous. Seam 3 must be standard compatible and must stay as a bundle. If you remove that, most will use CDI without Seam and we'll have to write own own glue around JEE6 as our intention is to use WAS v8.0 out of the box, ie with JSF 3.0, CDI and JPA 2.0, ie with Myfaces, OpenWebBeans and OpenJPA as those are the implementation bundled with WAS v8.0


                        I expected also that Seam 3 would be the right way to learn JEE6 but it does not seam to be good for that now as Seam 2 was a great way to learn RichFaces and JEE5 for us


                        For me things changed so much that I'm very confused from where to start but we need to move on. I'm thinking into diving into the Seam 2/JSF 2.0 discussion in order to go forward and user RichFaces 4.0 for Browser compatibility reasons and continue for some time (years) with Seam 2


                        Also I would be curious to have the current situation on compatibility with popular AS and standard implementation around here. The compatibility worksheet could include the most popular AS, persistence, JSF implementations, CDI implementation, Web 2.0 frameworks, logging frameworks etc. This would be a good indicator where Seam 3 is as today


                        • 24. Re: Seam.Next discussion
                          titou09

                          We are using Seam 2 in most of our web projects since a few years now. The team love it and we are very productive with it
                          I personally participate to have Seam 2 work with WebSphere v7.0 and I almost entirely rewrote the WebSphere chapter in the doc and the WebSphere samples with Seam 2 in addition to many commits


                          I'm in charge of the development team here and the way to go for us is to use WebSphere v8.0 that includes OpenWebBeans and OpenJPA


                          As explained in the mailing list my current tentative is a big failure : link here


                          Now that WAS v8.0.0.1 is out I was waiting for Seam 3.1 to make another try but I still have difficulties to know from where to start to build a prototype application and :



                          • most of the JIRA opened in June are still opened

                          • there is still no nice example to start from. The booking sample in Seam 2 was great for a beginning

                          • If you are not a maven shop (and we will never be), it is not that easy to get all the pieces together with sample and code fast

                          • I was thinking about participating to Seam 3 as I did for Seam 2 but another obstacle is there : git . I have first to learn git to participate and then to understand the Seam 3 module structure, owner etc.



                          Also I post a mail with ideas some time ago on how to reorganize the WiKi too to be more clear from where to start but nothing changed.


                          I want to have Seam 3 become a success but this is very confusing even for old user. I don't know for new users, but for old users the step is very very high, to summarize we have to eat and digest: maven, git, CDI, JSF 2.0, Weld, Seam 3, JBoss-logging and others this with the lack of easy sample to play with and all the incompatibilities with WebSphere, IBM JDK and even standards. Of course many of the new standards are not the fault of Seam 3 (JSF 2.0 CDI etc) but some choices are and add to the difficulty (maven ,git, JBoss-logging, logging frameworks confusions etc). lack of good sample, confusion on the WiKi/doc etc)


                          IMHO Seam 3.0 was an alpha version as the number of incompatibilities with standards are too numerous. Seam 3 must be standard compatible and must stay as a bundle. If you remove that, most will use CDI without Seam and we'll have to write own own glue around JEE6 as our intention is to use WAS v8.0 out of the box, ie with JSF 3.0, CDI and JPA 2.0, ie with Myfaces, OpenWebBeans and OpenJPA as those are the implementation bundled with WAS v8.0


                          I expected also that Seam 3 would be the right way to learn JEE6 but it does not seam to be good for that now as Seam 2 was a great way to learn RichFaces and JEE5 for us


                          For me things changed so much that I'm very confused from where to start but we need to move on. I'm thinking into diving into the Seam 2/JSF 2.0 discussion in order to go forward and user RichFaces 4.0 for Browser compatibility reasons and continue for some time (years) with Seam 2


                          Also I would be curious to have the current situation on compatibility with popular AS and standard implementation around here. The compatibility worksheet could include the most popular AS, persistence, JSF implementations, CDI implementation, Web 2.0 frameworks, logging frameworks etc. This would be a good indicator where Seam 3 is as today


                          • 25. Re: Seam.Next discussion
                            asookazian

                            I have used the JBoss dev stack extensively for a few years (Seam 2.x, RF 3.x, Hibernate 3.x, ant/maven, etc.)


                            I would like to very heavily emphasize that performance and scalability are big concerns.  The injection/outjection/disinjection model in Seam 1/2 was a disaster.  This has apparently been addressed and fixed by CDI/Weld which Seam 3 builds/extends on.  The @BypassInterceptors being all or nothing was a limitation as well in Seam 2.


                            If performance and scalability cannot be achieved with EE6/Seam3 then forget about large high-volume projects (like eBay, amazon, bofa, etc.) using the JBoss stack.  They will opt for Spring or another stack (RoR/.NET?) that can handle these scenarios.


                            Also think about how to incorporate and use Seam3 in mobile and cloud scenarios.


                            How can EE6 and Seam 3 be more popular and widely deployed in prod envmts?  It's hard for Seam devs and employers to find jobs and fill positions.

                            • 26. Re: Seam.Next discussion
                              egore911

                              To me the migration path is most important. I work for a company that (more or less) is always in the process of updating the libraries it uses. We are currently looking into Seam 3 because it looks really promising. But we have many applications that rely on the Home and Query concept of Seam 2.


                              This is a really important and missing point right now. We can't use Seam 2 in the long term an need to update. So we are currently in the process of porting the Query and Home classes from Seam 2 to Seam 3. Not sure if this works out as a few concepts from Seam 2 no longer apply. If we succeed we'd be happy to contribute this as a seam-query-module (or similar).


                              Regards,
                              Christoph

                              • 27. Re: Seam.Next discussion
                                gaboo.gael.livre-rare-book.com

                                We're a very small company, and seam 2 brought us prodictivity, fast learning curve and powerful tools : RichFaces, Hibernate Search integration, PDF generation, easy email templating, fast / reliable CRUD operation (Home/Query), xxxx.page.xml metadata files are very useful to set security, description, page flow, parameters, RESTEasy integration, etc.


                                The fact that seam2 is a simple bundle to download and unpack and code is very important to us : we are two developers and I won't have the time to hunt down all the matching jars, setting them together etc.


                                As already stated, all the new tools are also time consuming. Which is unfortunate. Git, CDI, JSF2, Maven. I'm not against progress, but please provide examples so that we know how it should be done and where to look at for a working example when lost.


                                Seam 3 should be an integrated, fast and east to use framework.


                                My boss was impressed. I need xls export : 2 hours later, done. A PDF export here would be cool. 1 hour later, I'd have a first version. A REST API to use by mobile devices ? All needed stuff is here, examples are available and getting started is easy. I've never lost too much time figuring things out with seam2. It's like an almost endless box of tools and building box to play with.


                                I was considering migrating to seam 3 this month. I guess with all previous comments and being a small company we're better waiting for you to figure things out and continue with seam 2 for a little while.


                                Thank you for being open to criticism and listenning to your users!

                                • 28. Re: Seam.Next discussion
                                  mana.hotmana76.gmail.com

                                  Robert Morse wrote on Sep 30, 2011 16:54:


                                  Seeing how the Seam team handled the JSF2 support for Seam2, I'm a little gun shy about jumping on the Seam3 bandwagon.  Assuming Seam3 will eventually provide similar functionality to Seam2, having adequate examples, documentation, and a Seam2 to Seam3 migration doc will be absolutely essential.


                                  Robert,
                                  JSF2 will come in Seam 2.3 and if you read my post about Seam 2 and current development, you should understand the slow moving.


                                  Anyway Seam 2 is opensource project and if you like to have it earlier than I can release, you always can help the Seam2 as community member. I welcome all help from community.


                                  The first step was and is done - Seam2 is in maven build system now. Also JPA adn Booking examples are migrated to JBoss AS 7. It is slow, but in progress!




                                  • 29. Re: Seam.Next discussion
                                    kem

                                    Hi,


                                    seam brings me to javaEE ;)


                                    before I jump into seam world, I did a doc. search and prototyping with other frameworks. Seam with it's fabulous seam-gen (thanks King) was what I was looking for: a full productive stack for ee apps. seam-gen was very helpful in this regards.


                                    Recently, I had to start a new project. After prototyping with seam 3 and some discussion in seam2 forum, I have decided to continue with seam 2. Simply because I am more productive with seam 2 than with seam 3.


                                    seam-gen is a GREAT tool. Seam Application Framework is also great. I can't find something similar in seam 3. Honestly I still don't see what is the benefit of seam 3 over vanilla java ee 6? I also don't see the benefit of seam-forge over a modern IDE.


                                    I am not a java ee expert. My core business is not java EE. But I do need java ee for my business. Anything to improve productivity and hide java ee complexity is welcome. seam 2 was more successful in this regards than seam 3.


                                    That being said, I believe that seam 3 is a great project, leaded and developed by expert in the field. But I also have the feeling that it's also targeting experts. I have the feeling (I am probably wrong) that end users, who are not experts in the field, who simply need simplicity and productivity are not well served in seam 3.


                                    Finally, I would like to join Gael, and thanks to all seam 2, and seam 3 people for what they have done and for being open minded ;-)))


                                    Regards,


                                    Khalil