8 Replies Latest reply on Nov 19, 2009 8:43 AM by mpickell

    Opinion on swing-like layer on Richfaces

    mpickell

      I'm looking for richfaces-users opinions on a pure java "layer" that we use at our company. We find it very useful, but we want outside opinions.

      We have developed a pure java "layer" that sits on top of JSF and allows JSF application development virtually purely through java. The API is similar to SWING, we only use a single JSP page and a simple bean for the page to bind to, and all components have complete and easy access to the application's full object model (bean isn't needed for this).

      For our use, this has simplified JSF development and allows for amazingly dynamic pages (especially now with the ajax support).

      We have an overview article on theServerSide on this (written prior to Ajax implementation) at : http://www.theserverside.com/tt/articles/article.tss?l=IntroducingJavaPureFaces.

      We also have the source now available at http://www.b6systems.com/javaPureFaces.html

      and a blog with more information at http://www.b6systems.com/blog/

        • 1. Re: Opinion on swing-like layer on Richfaces
          jbalunas

          Hi Matt,

          I'll take a closer look early next week, but this does look interesting.

          What are you plans or thoughts on the future of this?

          Is it anything like wicket?

          • 2. Re: Opinion on swing-like layer on Richfaces
            mpickell

            We would like to open it up as an open source project. We took the first step towards that by putting it up on our web page (link is in original post), and we are now trying to see if there is interest.

            I am not totally aware of all of the differences between this and Wicket, but from a quick look at an overview of Wicket, I see:

            In Wicket it appears you still need to do a lot of configuration in static files (html and/or xml). With Java PureFaces you do minimal upfront configuration and then pretty much everything is in Java (except CSS definitions or JS).
            In Wicket (and some others) you need to extend the framework in order to write classes for it. With Java PureFaces you only need to implement the PureComponent interface to allow any class to also handle its own UI

            Of course Wicket is a framework and Java Purefaces is a layer overtop of JSF that simplifies implementation. I hope I didn't misrepresent anything on Wicket -- like I said, I am not extremely familiar.

            • 3. Re: Opinion on swing-like layer on Richfaces
              vgarmash

              Actually guys I don't see any reason why you need the framework like this. JSF also single servlet application with pure Java API. You can build all your components, action listeners, validators etc using Java API. I cannot say that this is the way to go but it depends on your level of comfort with Java or XHTML. I am working with JSF for 3 years and I never had any reason to move coding focus from standard way towards Java API. I can't imagine the problem that requires such solution like Pure Faces. On your website I couldn't find any example of this problems and explanations why you think your way of resolving this problem is better than JSF 2.0?
              I understand that JSF is not the simplest framework in the world and you need to have good level of expertise in Servlets, JSP and other J2EE features before start working with JSF. And by the way for my taste JSF is very similar to Swing but it is adopted to client-server architecture. I saw a lot of real examples when developers with strong Struts skill couldn't work with JSF efficiently and they never liked it because it is not Struts :). But yes - it is different! You need to wrap your mind differently (and stop asking for c:foreach and c:if support :) ). And in most cases problem was in poor JSF documentation where people couldn't find good examples to follow.
              So may be you really resolved some particular issues you were facing. And if you are really interested in promoting your tool you better explain what real benefits we can get from switching to your framework.
              Currently I don't see any use in extra layer. JSF+Facelets+Richfaces is working in 100% cases for me and I am able to implement stable and efficient applications with fancy functions in very short period of time. And the besta part of it that if requirements changed I can apply them more quickly than with other frameworks.
              All descibed above is complete IMHO.

              • 4. Re: Opinion on swing-like layer on Richfaces
                mpickell

                vgarmash, I understand that not everyone is looking for a framework extension like Java PureFaces, and that some people are just fine with what they're using now.

                But when i look through the forums I see a lot of people looking to build dynamic code, like we are building very easily, and having issues. We have had a lot of success using it and would like others to know about it. I've also received direct emails requesting more information on it (based on the article and our blog). That's why we worked to make the source available on our site and why we think getting the word out makes sense.

                • 5. Re: Opinion on swing-like layer on Richfaces
                  mpickell

                  by the way.. we put up a small demo of some of the components with source code shown next to them.
                  http://www.b6systems.com/pureFacesComponents/demo.jsf

                  • 6. Re: Opinion on swing-like layer on Richfaces
                    ilya_shaikovsky

                    mpickell, great work! we'll check it out after finish some urgent works regarding nearest release ;)

                    • 7. Re: Opinion on swing-like layer on Richfaces
                      jbalunas

                      Hi mpickell,

                      Unfortunately I agree with Ilya, and won't be able to take a look at this right now. I am curious about it, but with the 3.3.3.BETA1 release and JSFSummit out plates are full.

                      • 8. Re: Opinion on swing-like layer on Richfaces
                        mpickell

                        No problem at all.

                        I look forward to your comments whenever you have time.