Are people still taking this EJB certification thing?

[Francis Amanfo]

Hi all,
This is just curiosity.
After all these heavy and serious discussions around the net about how bad the current version of EJB (at least the entity beans part) is and the drastic overhaul of it in EJB3 are you guys still bothering yourselves with this BAD beast?
My advise would be stop wasting your time. That thing would be thrown out in at least 2 years time.
Check some of the excellent discussions at http://www.theserverside.com/news/thread.tss?thread_id=25779 and
http://www.theserverside.com/news/thread.tss?thread_id=25804
Regards
Francis

[Roger Chung-Wee]

As someone who is involved with a couple of applications which have thick, remote clients (literally thousands of miles from the servers) and which use a lot of transactions and JMS messaging, I'd like to know what is a better solution than using an EJB container.
We use WebLogic, various versions up to 8.1 running on on Solaris servers. Granted, this isn't the cheapest solution, but as a big corporate with lots of other applications which also need a web container, clustering, scalability, etc, etc, WebLogic works fine. I believe that we've cut a deal with BEA, so the cost of licensing probably isn't a big issue, come to think of it.

[Francis Amanfo]

Originally posted by Roger Chung-Wee:
As someone who is involved with a couple of applications which have thick, remote clients (literally thousands of miles from the servers) and which use a lot of transactions and JMS messaging, I'd like to know what is a better solution than using an EJB container.
We use WebLogic, various versions up to 8.1 running on on Solaris servers. Granted, this isn't the cheapest solution, but as a big corporate with lots of other applications which also need a web container, clustering, scalability, etc, etc, WebLogic works fine. I believe that we've cut a deal with BEA, so the cost of licensing probably isn't a big issue, come to think of it.

I don't doubt there are a class of enterprise applications that definitely need the EJB container. Having said this, I must also say that I have no shadow of doubt that the majority of the enterprise applications just don't need that high end solutions like you've mentioned. For most applications a combination of JSP/Struts/Tapestry for the front end, Spring for business logic flow manipulation, and Hibernate for persitence all running on cheap but great Linux OS is all you need. EJBs might be just an overkill
Regards,
Francis

[Roger Chung-Wee]

Something like transactions do not necessarily equate to high level solutions as quite small applications can need transactions.
If you need to put your transactions through the ACID test, I think it's overkill for the developer to write the code to do this when it's already available in an EJB container ...

[Francis Amanfo]

Originally posted by Roger Chung-Wee:
Something like transactions do not necessarily equate to high level solutions as quite small applications can need transactions.
If you need to put your transactions through the ACID test, I think it's overkill for the developer to write the code to do this when it's already available in an EJB container ...

In Spring, there is also a declarative transaction demarcation as in EJBs but not via deployment descriptors- one of EJBs Hell. In fact transaction is well taken care of in Spring.

[Tejas Bavishi]

Can someone point me to Spring related resources and what it is ?
Thanks
Tejas

[Gabriel Forro]

http://www.springframework.org/
enjoy it!
[ May 10, 2004: Message edited by: Gabriel Forro ]

[Francis Amanfo]

Originally posted by Tejas Bavishi:
Can someone point me to Spring related resources and what it is ?
Thanks
Tejas

A link to Spring and its docs:
http://www.springframework.org
Regards

[Axel Janssen]

Its far more easier to understand Hibernate and Spring, if you allready have had some exposure to EJB.
My question: What would you use to replace MDBs in an non-EJB app?
[ May 10, 2004: Message edited by: Axel Janssen ]

[garth schneider]

Francis,
I am building large Portal Servers on OS/390, RS6000 and ISeries systems. I have 2000 users hitting at one time in some cases. Please tell me how I can use .NET. in this environment. Please tell me of a middleware solution that will work in this area besides EJB's.. I need to go against other Enterprise Information Systems like ERP, SCM, CRM systems on different systems like Solaris, IBM etc (sorry no windows here). And I need to be able to do 2 phase commit on these systems as well. I will need to replace Session Beans, Message Beans and a few Entity Beans if I go away from EJB's.

Originally posted by Francis:
"I don't doubt there are a class of enterprise applications that definitely need the EJB container. Having said this, I must also say that I have no shadow of doubt that the majority of the enterprise applications just don't need that high end solutions like you've mentioned. For most applications a combination of JSP/Struts/Tapestry for the front end, Spring for business logic flow manipulation, and Hibernate for persitence all running on cheap but great Linux OS is all you need. EJBs might be just an overkill. "
I do not agree with this statement. There are many Global Corporations that need to integrate large distributed systems. Enterprise Javabeans do a great job in allowing this to occur.

Best Regards,
Garth
SCJP
IBM Solution Developer VisualAge for Java and WSAD
[ May 10, 2004: Message edited by: garth schneider ]

[Francis Amanfo]

Originally posted by garth schneider:
Francis,
I am building large Portal Servers on OS/390, RS6000 and ISeries systems. I have 2000 users hitting at one time in some cases. Please tell me how I can use .NET. in this environment. Please tell me of a middleware solution that will work in this area besides EJB's.. I need to go against other Enterprise Information Systems like ERP, SCM, CRM systems on different systems like Solaris, IBM etc (sorry no windows here). And I need to be able to do 2 phase commit on these systems as well. I will need to replace Session Beans, Message Beans and a few Entity Beans if I go away from EJB's.

Originally posted by Francis:
"I don't doubt there are a class of enterprise applications that definitely need the EJB container. Having said this, I must also say that I have no shadow of doubt that the majority of the enterprise applications just don't need that high end solutions like you've mentioned. For most applications a combination of JSP/Struts/Tapestry for the front end, Spring for business logic flow manipulation, and Hibernate for persitence all running on cheap but great Linux OS is all you need. EJBs might be just an overkill. "
I do not agree with this statement. There are many Global Corporations that need to integrate large distributed systems. Enterprise Javabeans do a great job in allowing this to occur.

Best Regards,
Garth
SCJP
IBM Solution Developer VisualAge for Java and WSAD
[ May 10, 2004: Message edited by: garth schneider ]

Garth,
I agree with you that there are many Global Corporations that need to integrate large distributed systems. My question to you is do you implement integration solutions with EJB? Is that what EJB is meant for? I don't think so.
Anyway, what I want to say is that for the majority of projects we just don't need an EJB solution. Spring/Hibernate would do excellent work without all the complexities that EJB comes with.
That was my 01 cent.
Francis

[Ko Ko Naing]

Guys, don't try to make me doubt... I'm almost taking the exam within this month...
Anyway, the background in EJB is needed for such things, I guess... So go ahead with EJB...

[Francis Amanfo]

Originally posted by Ko Ko Naing:

Anyway, the background in EJB is needed for such things, I guess... So go ahead with EJB...

I disagree! Sometimes immersing yourself too much in a certain technology can serve as a stumbling block in learning another. This is because it becomes difficult for you to leave the old way of thinking you've developed. Since EJB 3.0 would have a completly re-designed architecture, I suspect the way of thinking to develop in it would also be different.
Just my 02 cents.
Francis
[ May 11, 2004: Message edited by: Francis Amanfo ]

[Ko Ko Naing]

Originally posted by Francis Amanfo:
This is because it becomes difficult for you to leave the old way of thinking you've developed. Since EJB 3.0 would have a completly re-designed architecture, I suspect the way of thinking to develop in it would also be different.

Mmm... But I believe that new things come from old things. If we don't have EJB 1.1 or 2.0, how in the heck EJB 3.0 will be able to come up? Even though EJB 3.0 has a different architecture, I believe it cannot throw most of its properties under EJB...
But I don't know why you are discouraging people out here in SCBCD Certification forum about EJB 2.0...

[Nicholas Cheung]

Guys, don't try to make me doubt... I'm almost taking the exam within this month...

How's your preparation? Go smooth? How's SCDJWS as well?

Anyway, the background in EJB is needed for such things, I guess... So go ahead with EJB...

I totally agree with you. The idea of certain technologies are ususally copied from others or a mix of others.
The idea of name services in COBRA was also borrowed from the DNS concepts, looking up the services as well, and now J2EE and .NET try to adopt the apporaches. I cant find any problems that certain technologies going through the iterative processes on keeping the improvements. Instead, I am very appreicate that SUN put so much effort on revising the existing technologies, to make them better.
Although I dont come across with Hibernate, or other open source projects, I do believe that there is NO prefect software. They are good at some areas, but maybe bad at some other areas. If something is really that prefect, EJB is already disappeared or venished, and replaced by that technology.
Nick

[Ko Ko Naing]

Originally posted by Nicholas Cheung:
How's your preparation? Go smooth? How's SCDJWS as well?

My preparation is going well... only for SCBCD... I guess I cannot take Web Services Exam anymore...
Even SCBCD preparation is a bit struggling... I cannot lift two heavy loads like this... And my current project is a bit making me busy as well...
How about your preparation for Web Services? Hope you'll do well in the exam...

[Nicholas Cheung]

My preparation is going well... only for SCBCD... I guess I cannot take Web Services Exam anymore...

Oh... I am a bit disappointed with this. I guess we can crack SCDJWS together.

Even SCBCD preparation is a bit struggling... I cannot lift two heavy loads like this...
How about your preparation for Web Services?

I am currently studying 3 exams simultaneously. I have scheduled SCDJWS on 23 May, 1z1-036 on 30 May, and SCBCD on 2 Jun. Since my discounted vouchers for those exams are subjected to the beta, and university policy (for SCBCD), the schedule is very very tight.
I finished reading Richard book, and found it is a bit difficult for me to understand those XML stuffs. I am waiting for the revision notes from lovely MZ and Sathya, for the 2nd revision.
In the meantime, I am reading Linux kernal stuffs for 1z1-036. As I am not a System guy, I also found difficult to work with OS.
Thus, finally, I come back to SCBCD. I read HF EJB twice, and testing myself with EJBPlus. However, the scores are not high, only 70-75%.
I hope I can get more time to be more well-prepared. Anyway, best luck in your SCBCD. Do tell me your experiences after taking the exam.
Nick

[Ko Ko Naing]

Originally posted by Nicholas Cheung:

I am currently studying 3 exams simultaneously. I have scheduled SCDJWS on 23 May, 1z1-036 on 30 May, and SCBCD on 2 Jun. Since my discounted vouchers for those exams are subjected to the beta, and university policy (for SCBCD), the schedule is very very tight.

Oh, I should be jealous of your hyper-speed reading skill... Since I got no commercial experience in EJB, I first time read HFEJB and it left me with lots of doubts... It's the second time I'm on HFEJB...

[Nicholas Cheung]

I should be jealous of your hyper-speed reading skill

My speed is not fast.
I used 3 weeks to finish reading Richard's book. I think it is only becos HF EJB is easy to read. I can finish the 1st reading within 1 week, and use 1+ week to re-read it.
BTW, when is your exam?
Nick

[Ko Ko Naing]

Originally posted by Nicholas Cheung:
BTW, when is your exam?
[/QB]

I haven't registered it yet... But I'm sure I have to take it before or on May 31... I guess I'll do it on May 31....
Since I have both EJBPlus and WhizLabs, I have to digest them before the exam as well....

[Nicholas Cheung]

How about softSCBCD?
Many people said that it is very similar to the real exam.
Nick

[Ko Ko Naing]

Originally posted by Nicholas Cheung:
How about softSCBCD?
Many people said that it is very similar to the real exam.
Nick

I dunnoe... :roll: So r u preparing with that simulator?

[Kathy Sierra]

I just want to encourage everyone who is preparing for the exam today!
You're studying what people are using today, and what people will be using for the next two years. Francis is partly right--*some* (certainly not ALL!) of the EJB 2.0 spec will change, parts of it will be dramatic (but in a *good* way). But a good portion of what you learn now will be beneficial and relevant, and in 18 months (more likely at least two years) it won't be very difficult for you to learn the new EJB 3.0 things and take the small *upgrade* certification exam to take your certification from 2.0 to 3.0. Remember, 3.0 is NOT a replacement for 2.0. I believe the radical changes from 2.0 to 3.0 have been overstated. Learning 2.0 today *will* mean that you've learned a few things you won't have to do in the future, but you're still getting an important and useful foundation for EJB, and you may have a greater appreciation and understanding for the 3.0 changes BECAUSE you've experienced 2.0.
In the meantime, don't be discouraged if this is what you've chosen to prepare for. I think Francis makes a very good point if you are not *currently* planning to use EJBs in your company. In that case, you might choose to wait until 3.0 comes out. But that is not the point here. I also think he would have a very good point if this discussion were taking place in January 2006. In that case, I *might* (depending on what happens--specifications almost always slip their deadlines) recommend that you consider preparing for the new version instead.
But if you ask this question one year from now, I will STILL be saying, "It's going to be a long time before the new spec is in widespread use and the Container vendors have fully upgraded and companies have migrated their enterprise systems, etc..."
All technical knowledge becomes largely obsolete. All of us have had to learn and unlearn and relearn as technologies mature. Heck, I remember having to build my OWN scrolling text box because in Java 1.02 (the first official release of Java) in 1996, the AWT (remember AWT??) didn't actually work. And if you look here on javaranch, I'm so proud of my hand-crafted double-buffered cows on the rule roundup game but today, Swing takes care of double-buffering now. But that doesn't mean I wasted my time, because I FULLY understand what's happening behind the scenes and can step in whenever I need to having done it *the hard way*.
So, cheers and good luck to all of you in your studies! Before I worked for Sun, I, too, had to take some of the Sun exams, and I still remember what that was like! I made a lot of flash cards, and took them with me everywhere. I would take them into the movie theater, and review them while waiting for the movie to begin. I would take them in the car, and review them while waiting in traffic (I lived in Los Angeles at the time, so I have the terrible LA traffic to thank for my passing the exam). I would make my friends quiz me with the cards, even if they didn't know anything about Java.
But as you all know, it's a great feeling when you pass, and you always learn much more in preparing for an exam than you would if you just read the parts of the API that you liked or needed to use. That's the great thing about exams, in my opinion, that you are *forced* to learn more about the technology BEFORE you actually have to use it. Working at Sun with EJB customers, I could not believe how many people I met that were component developers who didn't know some of the most fundamental things about EJB. Most developers learn only just-in-time, just what you need to get through the current project, and when you have a better understanding of the WHOLE technology, you can make better choices.
Studying for the exam will *never* be as good as hands-on experience, but hands-on experience does NOT guarantee a good knowledge and understanding. A good developer needs both, but I'd much rather start out with knowledge of how it all works than to just dive in and then some day come to realize that I had a wrong understanding of the technology and was creating things in a very inefficient, non-portable, or incorrect way.
I still want to say that I appreciate the point Francis is trying to make here; I just think he's making it too soon for the people on this forum, and now that he's made the point, I don't think we have to discuss it any further. Everyone knows what's happening, and if you choose to take your EJB certification, then I want you to be supported and helped and encouraged here!!
Have fun
--Kathy

[Niki Sharma]

well i've bought it, bought the head first and EJB book. Haven't yet started on it. Lacking the motivation to get going.

[Joyce Lee]

Kathy, thank you so much for your advice, encouragement and support!!!
Joyce

[Valentin Crettaz]

This topic is a duplicate of https://coderanch.com/forums/ which has been transferred to the EJB forum as it deals more with the future of the EJB technology than with the SCBCD exam itself.

If you want to take part in this discussion and express your opinion, you may freely do so at https://coderanch.com/t/314041/EJB-JEE/java/Advice-those-planning-SCBCD

This topic is now closed. Thank you for your comprehension