1 2 3 Previous Next 72 Replies Latest reply on Sep 4, 2009 12:47 AM by jkronegg

    Seam - a Product in Infancy or Just Another Framework

    brettshelley

      I worked with Web technologies like Struts a few years back and when I read about Seam, I put together a Seam project on my own.
      It was fantastic. I really liked the JBoss Tools, EJB3, JPA, Hibernate, AJAX, JSF integration - the one-stop solution (full stack) that Seam offers is really impressive.


      However, the technology does not seem to be taking off across the larger market.  I performed a keyword search on Dice for
      'Seam' and received 38 records.  I then did the same with 'Spring' and had 1595 results. That is a huge difference.  So my question is why? 


      I don't have the Spring or background with other web frameworks arising after 2004. Do other Java frameworks offer the same advantages?  Is SEAM more of a pure RedHAT JBoss consulting framework?  Or is it just so far ahead of its time that it will take another year or two before it takes off?


      I would love to hear some opinions...







        • 1. Re: Seam - a Product in Infancy or Just Another Framework
          amitev

          This is probably because Seam is not market-oriented like Sprint, right?

          • 2. Re: Seam - a Product in Infancy or Just Another Framework
            tony.herstell1

            JSR299 has been adopted which says a lot!


            JSF is mildly broken and Seam fixes a huge amount (slowing seeping back into core JSF).


            I am still partly in the legacy industry (and fighting tooth and claw to get out) working on customer site on some non-seam projects and i.m.h.o. people are loath to change unless they have a reason.


            .I have recently worked on 4000 line JSPs (client delivered code).


            .I have recently watched (deliberately avoided) a new JSF project struggle with all the things Seam can do without breaking a sweat.


            Personally I think bothering to LEARN Seam (and JSF) is something that a lot of 9-5 programmers won't bother with... if it ain't broke don't fix it. So they jump in trying to write seam like older frameworks and have problems.


            Which leaves the people whom really want to move on...


            Stick with it as the jobs our NZ Seam consultancy are looking at are all are very interesting and generally for forward thinking companies that want to move from legacy problems with old frameworks... very interesting work actually. I wake up and want to code using the Seam stack these days for less pay!


            With a new guy on our team, he learn't Seam and got out a client project 6 weeks early! ( must have been my training ;) including having to drop outside seam as seam pdf did not do enough ).


            If Seam didn't just work I suppose there would be more jobs!


            i.m.h.o. I suppose if the height of the bar to getting into Seam was lower more people would do it (thinks about a on-line video series done by technical authors or else getting technical authors to re-write the seam manual that is shipped with Seam .. or BOTH!)


            These are only my thoughts; and probably worth exactly what they cost you!

            • 3. Re: Seam - a Product in Infancy or Just Another Framework
              brettshelley

              Thanks for the responses.  I am certain they are worth a lot more than they cost me. 


              I agree with a number of points.  I think the height of the bar might explain a lot. A lot of junior developer types don't know Annotations and they certainly do not know old EJB.  So something I might view as an extraordinary simplification might be viewed as confusing to someone never exposed to EJB1.0-1.1.


              I can imagine maybe Red Hat /JBoss is explicitly not ramping up the marketing effort to get Seam widely accepted.  Using seam quietly, they can send their own consultants in, knock out great projects as light speed, and then move on to the next.  Silverstream had a similar model back in the day (only they were not open source).  But who knows, this explanation might be completely off, it might just be more along the lines of people are reluctant to change and learn something new. 


              I would love to hear from anyone with some Spring or other popular framework background to give perspective on what frameworks are the best,fastest,easiest to learn, etc.


              Thanks for the responses, Tony and Adrian.


              Brett

              • 4. Re: Seam - a Product in Infancy or Just Another Framework
                jeanluc

                I can imagine maybe Red Hat /JBoss is explicitly not ramping up the marketing effort to get Seam widely accepted. Using seam quietly, they can send their own consultants in, knock out great projects as light speed, and then move on to the next.

                I am very skeptical this approach would fly from a business standpoint. If Seam is not marketed enough (and is not), people starting new projects will simply choose the frameworks they heard most about. In contrast, Spring supports an active consultancy and training business: SpringSource (recently acquired by VMware). Quiet development doesn't make a framework successful, sustaining an eco-system does. Ours is rather small and often isolated from the Seam developers themselves (there are many threads asking for comments from the dev team). The search volume with Google has already been going down since 2007, see the graphs at Google Search Insight for Seam and Google Trends for Seam.


                Having worked in many corporations before, I suspect the reasons Seam is not promoted enough by JBoss is that at a high executive level it's not seen as a serious revenue source. Either after some market analysis or simply because the people in charge do not even bother (there is plenty of politics after any major merger and RedHat-JBoss is such. Within the landscape with all the products from both companies, Seam pretty much lost. In fact, they might ultimately be right: can Seam be indeed a revenue source, especially during recessions when fewer companies would be willing to pay $$$$ for basic training (most trainings are such, covering what one could learn from the reference docs or published books. Heck, Seam in Action goes deeper than any set of Powerpoints could cover in a 2-day workshop.


                I speculate Seam will stay for a while (at least version 3), but it will remain a minor player. By then, its developers will have moved to something else. Perhaps, by then, Arbi (who I think is the most active forum member) will become the new head of the project :-)

                • 5. Re: Seam - a Product in Infancy or Just Another Framework
                  israel.bgf

                  Perhaps, by then, Arbi (who I think is the most active forum member) will become the new head of the project :-)

                  Lol, i can say that i would vote in him too. :)


                  About seam in the market, well I don't know either but here in my City i started to spread the Seam knowledge and every one without exception really like it. Maybe to Seam be more famous it would need some evangelist like Rails have.

                  • 6. Re: Seam - a Product in Infancy or Just Another Framework
                    brettshelley

                    Magnificent commentary!  Thanks.  This analysis is far better than any of my speculation.


                    I don't know Arbi - but the technology/framework name Arbi has a nice ring too it. Instead of Indian warrior-like names, snakes that kill with Power, fancy jewels, aggressive animals, it would be a great change to incorporate the roast beef sandwich into the open source world.  Arbi is short, decisive, not spelled like the sandwich maker, and can easily be made into an acronym such as   Application for Realtime Business Intelligence.  It sounds like this Arbi fella has a great future. 

                    • 7. Re: Seam - a Product in Infancy or Just Another Framework
                      israel.bgf

                      I think that Arbi would be proud to hear that about his name.

                      • 8. Re: Seam - a Product in Infancy or Just Another Framework
                        asookazian

                        Well, when I started Seam (over 2 years ago with 1.2.1.GA), it was PMuir who was the most helpful and popular forumite.  Now it may be me, who knows.


                        But I definitely am not qualified to be an architect and lead a team/project as complex as Seam and JEE 5/6.  And I do not have time as I have an 8 month old son. 


                        Currently, there are very few Seam jobs in USA.  I'm talking a handful on indeed.com, etc.  Most of them in the Java space require some experience with Spring, JSP, JDBC, Hibernate, Flex.  Why would a Spring/Hibernate project add integration components with Seam or completely upgrade to Seam?  It's difficult to sell that idea to the business that is usually paying IT salaries.  If it's additional functionality, maybe.  Fixing a performance problem, maybe.  Otherwise, near to impossible.


                        At my company, they transitioned from .NET platform to Redhat/JBoss/Seam stack.  Maybe in those cases it makes sense to use Seam, but then again, performance and scalability for external-facing web sites is critical, and then using JSF/Seam is typically a problem.  How many big web sites are running Seam (e.g. twitter.com runs on RoR with Scala).


                        Even though I have ran the debugger thru the various core Seam components/classes, I still have a vague understanding of the internals (proxies, transactions, persistence, interceptors, bijection, remoting, jBPM integration, etc.) after more than 2 years.  Yes, I do know a lot, but I have spent hundreds of hours literally getting up to speed with JSF, Seam, RichFaces, JPA, Hibernate.  All this was new to me except EJB 2.x two years ago.


                        It is the huge learning curve, the lack of or poor marketing from Redhat, and lack of participation in the forum (nowadays, like I said, PMuir was very helpful, very often 1-2 yrs ago).  They are focusing their efforts on Web Beans and JSR 299 as is evidenced by recent activity in the Web Beans forum and the agenda for the upcoming Jboss World. 


                        After all the help I gave on this forum, you'd think I'd get a free ticket to Jboss World or something.  Nothing.  I do not work for Redhat and most likely will never be considered.  Management does not care about forum activity.


                        Sometimes I feel that there are literally no more than 100 qualified corporate Seam developers (who are not employed by Redhat) with more than 2 years experience in the entire world.


                        Knowing a technology stack that is somewhat exotic and advanced and yet totally not in demand is a very strange situation IMHO.  It's like having 10 exotic parrots (and only 100 left in the world), but nobody even cares...

                        • 9. Re: Seam - a Product in Infancy or Just Another Framework
                          asookazian

                          Seam has been in GA for over 3 years now (June 2006 IIRC).  That's a long time in the Java space.  The root cause problem here is that EJB 3 final spec was released too late (final release: May 2006).  And Spring/Hibernate was released in 2003/2004 IIRC.


                          So basically, if it was reversed and Seam and EJB 3 were released in 2003 and Spring was released in 2006, the EE space would look different in terms of traction/adoption of the different JEE stacks.


                          And yes, perhaps Spring is not JEE stictly speaking (as an EJB container is not required), but you get my drift...

                          • 10. Re: Seam - a Product in Infancy or Just Another Framework
                            brettshelley

                            Hi Arbi,


                            Thanks for posting. 


                            I do not know anything about JBoss/Red Hat management, but one would think that their forums would be the ideal place to find the best people. 
                            If they are too busy working on other stuff, then don't take it personally.  You have obviously gained a great deal of respect from external folks. 


                            Some company in Atlanta is looking for people with Seam experience...


                            You've got the right idea in focusing on your 8 month old son. 


                            Thanks for the tip on Web Beans and JSR299.


                            Brett



                            • 11. Re: Seam - a Product in Infancy or Just Another Framework
                              asookazian

                              Brett Shelley wrote on Aug 24, 2009 20:29:


                              I do not know anything about JBoss/Red Hat management, but one would think that their forums would be the ideal place to find the best people. 
                              If they are too busy working on other stuff, then don't take it personally.  You have obviously gained a great deal of respect from external folks. 



                              The way it seems to work (at least from what I noted from PMuir's and DAllen's experience) is you must contribute to the Seam codebase in order to be hired by Redhat/JBoss as a member of the core dev team.

                              • 12. Re: Seam - a Product in Infancy or Just Another Framework
                                asookazian

                                btw, thanks for all the nice comments.  I try to help as much as I can (good karma you know).


                                Break down the learning curve!


                                :)

                                • 13. Re: Seam - a Product in Infancy or Just Another Framework
                                  tony.herstell1

                                  More free thought (so worth every penny)




                                  The Present



                                  Q. Why has the forum support died?


                                  A. The seamsters are all off working on JSR299 (Gavin was probably doing 04:00 regularly to hit that deadline - bless...)


                                  Q. When will the support improve for Seam?


                                  A. When companies pay for it (support contracts equals $$$) and the free forums get the roll-on benefit. Its a very good company that supports free forums indefinitely.


                                  Q. When will companies buy support?


                                  A. DOH! When more companies use Seam.


                                  Q. When will more companies use Seam


                                  A. When companies discover that Seam saves them time/money.


                                  Q. When will companies discover that Seam saves them time money.


                                  A. When someone stumps up a few mil (ringfenced) to create an evangalising company (does not have to include all developers that contribute as that's like getting developers to write manauals) to go out there and show people, through training/hand on seminars, how to write a Seam app quickly (see Spring!)... show people what conversations are, show people how jBPM works and show people jPBM, and show people that yes, j2ee was sub-optimal but EJB3 sorts that out, show people that JSF is flawed but yes seam fixes that too, show people that they don't have to set lazy equals false in their whole project as they cant figure out how to stop LIE. And SEND THEM AWAY with a project skeleton that they created on the training course.


                                  Personally, for these types of demos, I would recommend not using seam gen as its too mysterious and I would use MyEclipse with hot-synch then you don't have to use Ant OR Maven and you don't have to wait for a build-deploy cycle (how 80's) - Yes Ant has its place when you start having to go outside seam and cant convince people to just connect through a Web Service but you DON'T need to do that at the start - it's a last resort as it kills productivity.



                                  Trying to tell people WHY you should use Seam is generally pointless (we have all tried). You can convince the hell out of people till you are blue in the face but they will still go off and use another framework (as someone in the left told them blah blah blah).



                                  The Future




                                  JSR266 has been selected by Sun for J636E (Jave 6 with EJB3 rolled out as J6EE) and so JSR266 has a future. Give the seamsters TIME to get JSR299 rolled back into Seam's core and the likes of Gavin, Pete and Dave will strangely re-appear on these forums when they Merge WebBeans forum with seam forum (have you seen the response time on the WebBeams forums?). Of course nothing can compare to how quick and helpful the Richfaces guys are.



                                  Now for Red Hat... come on guys... recruit some forum trolls from the community (freelance or paid). Specifically to support paying customers but 30% of the time to support the open source forums... as this is where people trying the technology live... NOT having questions answered is a REAL turn off for people looking at technology. Non-Active forums equals Dead Project.


                                  Myself and another guy were recruited (freelance in my case) for supporting ICEFaces a couple of years ago.. training done in Calgery. So it does happen.



                                  It's not about what can my country do for me; it's about what I can do for my country

                                  Right... you lot... Get out there and build smart projects and even set up companies (that get bought and thus crowbar seam into other companies ;) ). Get articles written etc.


                                  I am writing what could be a massive project at the moment. I am using a 10 year old server and my home PC... at some point I will go for VC (when the project is more established) and this will have to scale massively so at some point I will have to get a support contract with seamsters as I expect them to sort out that bit. If there was ever a project that had to scale then this would be it. I have used Spring, Tapestry, Struts etc. etc. and I chose Seam for this project... as I know that through JSR299 that Seam JSF is the future as it's the only technology backed by all the major players.




                                  plus 1





                                  I recommend Arbi to be recruited as first Forum Troll.


                                  I also know of a very very smart programmer that first told me about seam (but has never contributed) that may be very suitable for being a seamster (ex Sun Architect actually), he has contributed fixes to open source projects in the past... pm me for details.



                                  And can someone please change this editor to the Richfaces wysiwyg Editor (we should eat our own food!) it's buggy and having it on this forum WILL HELP it get fixed.


                                  As for saying it will get better when we get out of the recession... well... this is the world out of recession! So get used to it... this IS the norm! the age of excess is over.

                                  • 14. Re: Seam - a Product in Infancy or Just Another Framework
                                    jeanluc

                                    So, does Seam's license allow the code base to be forked? :-) :-) There's nothing like competition to make corporations move faster...

                                    1 2 3 Previous Next