Kathy and Bert: Future of EJB?

[Rick Portugal]

Kathy and Bert,
I have a co-worker who contends that getting your SCBCD cert is a waste of time because there is no future in EJB. I don't agree, but I must admit that EJB hasn't taken off as much as I had hoped that it would.
Years ago I read that EJB would be a boon for third party companies who would develop EJBs for companies to plug right into their apps. I'm sure that some companies do that, but I haven't seen a lot of that.
I know of a company that has a J2EE system, but will not use EJBs because they feel it is opening a can of worms. They feel that the complexity of EJBs outweighs the benefit.
Do you feel that EJB is a growth technology? Is getting a SCBCD cert worth the time?

[Bert Bates]

Rick -
I'll squeeze my two cents in before that pesky cowgirl
I was just discussing this with a Java guru from Scandinavia the other day. He is seeing a steady growth of EJBs in Europe. His feeling is that EJBs kind of got caught the the dot-com disaster a couple of years ago, and that a lot of good things (like EJBs), got mixed in with some of the nonsense that was happening at the time. His theory is that after the initial hype there was a bit of a crash, and now we're in the phase of realistic steady growth.
That seems right to me. There's no doubt that the technology will continue to evolve, and it will certainly be hard to predict what EJBs will look like a few years down the road, but as computer technologies go, this one seems to have a reasonable 'set of legs'.
At any point in time the technology is evolving, and you'll never be in a situation where you can say "oh, I'll just wait for the situation to stabilize, then I'll jump in"
So I guess the short answer is 'yes', I think EJBs is a strong technology. It's certainly not right for every situation, but I think it's really great in a lot of environments.
[ October 28, 2003: Message edited by: Bert Bates ]

[Karen Gomes]

I guess the number of posts on this forum and the enthusiasum for SCBCD itself is a good indication of the popularity of EJBs
Long Live EJB
Long Live the SCBCD community
Cheers
Karen

[Ko Ko Naing]

Originally posted by Rick Portugal:
Kathy and Bert,
I know of a company that has a J2EE system, but will not use EJBs because they feel it is opening a can of worms. They feel that the complexity of EJBs outweighs the benefit.
Do you feel that EJB is a growth technology? Is getting a SCBCD cert worth the time?

Yeah, in some developing country in Asia, EJBs are not widely used, since they are afraid of the complexity of EJBs. But they will see the benifits of EJBs, only after they have tried to use them...
Thoughout my experience, most of the projects are done using only web components like JSPs and servlets... I have seen only one or two EJB-project here in my working life... It also depends on the individual's workign life too...

[Kathy Sierra]

Originally posted by Bert Bates:
Rick -
I'll squeeze my two cents in before that pesky cowgirl

"Pesky cowgirl"? Are you calling me a "pesky cowgirl"? Well, you can rest assured you're gonna pay for that one
(of course, as I write this, Bert is sitting about two meters away from me...)
Anyway, this time I guess I'll have to agree with Bert, and of course add MY two cents
From working at Sun, I can tell you this much: the Sun Professional Services team as well as the J2EE team are virtually overwhelmed with the amount of work required to satisfy loads and loads of huge customers using EJB. They are constantly responding to input from vendors and customers who are already deeply commited to this technology, and have no intention of using anything else in the near future. Because... what would they choose? If there's a strategic decision to NOT go the .not route, then J2EE is it, and for a specific class of large-scale enterprise projects, Web technologies alone (servlets/JSP) are not capable of handling the scalability requirements.
The problem with EJB is that in the beginning, too many people were using it who SHOULD have used servlets and been quite happy. Or it was used incorrectly. But there are tons of war stories of those who decided NOT to use EJB, and went with a Web-only solution, but who later found it difficult to extend and scale and then had to re-engineer a lot of their work.
One way to judge the health of EJB is to look at the efforts the vendors are putting in to their products to support it. They would not be spending that much R&D effort on products for which they did not see a healthy market (or at least a market worth pursuing).
That's my story and I'm sticking to it
This doesn't mean that many aspects of it won't evolve, or that the ways in which it is used won't change (as in the efforts to find integration between EJB and JDO -- I'll try to say more on that near the end of the week).
cheers,
Kathy

[Ko Ko Naing]

So can we conclude on the EJB technolgy as the following?
-EJB should be used in very large enterprise site, as it is famous for its scalibility and extendibility.
-EJB should not be used without possessing expertise of EJB technology, as it might affect the site itself, when the developers cannot handle technology well.
I just would like to share the knowledge...

[Rishi Singh]

Hi All,
The EJB's is supposed to be the main driving technology for mid- large J2EE projects and with so many vendors spending such a lot of R&D efforts
in trying to comply with the specs , i guess it will be refined and we may see some new version of the specs which has integration of Entity and JDO both , but as long as we talk of sever side java , we are referring to nothing but EJB...my two cents
Rishi

[Ko Ko Naing]

Originally posted by Rishi Singh:

but as long as we talk of sever side java , we are referring to nothing but EJB...my two cents
Rishi

If we are talking about server-side Java, we are also referring to JSPs and Servlets...
Just want us to remember the indispensible web components...

[Kathy Sierra]

The main issues often come down to:
1) do you even NEED the services the Container gives you? Like security, transactions, etc.
2) does scalability matter to you?

If you answered "no" to those, then you might not have a good candidate for EJB. Of course, there are many, many, many other questions to ask, but those two are the ones to start with, when wondering whether EJB makes sense for an application. Other questions include extensibility... like, do you forsee a time when clients might be coming from something other than a web browser?
But if you answered "yes" to either of those (or especially number one), you will probably be at least *considering* EJB as a solution. The complexity and economics of using EJB begin to pay off at a certain level of scalability. Below that, you probably have more server/services/overhead than you need. But above that... if you are *not* using EJB, then you may be in for trouble in one way or another.
I don't know what that scalability level is... but it is something our (Sun's) architects talk about with customers A LOT.
And yes, EJB is a bad solution for anyone not willing to invest the time to really understand what it is good -- and NOT good -- at doing.
cheers,
Kathy

[Ko Ko Naing]

Originally posted by Kathy Sierra:
The main issues often come down to:
1) do you even NEED the services the Container gives you? Like security, transactions, etc.

Kathy, do u mean that using EJBs makes the application more secure than using just JSPs and servlets? Coz I am going to have an e-commerce site soon, in which the customers want me to raise the level the secutiry as much as possible. I'm wondering if I should suggest them to use EJBs...
Thank you for your explanation...

[Pradeep bhatt]

I would recommend that you read the Bitter EJB book from Manning. It will let you know when to use EJB.

[Ko Ko Naing]

Originally posted by Pradeep Bhat:
I would recommend that you read the Bitter EJB book from Manning. It will let you know when to use EJB.

Thanx Pradeep... I have downloaded some sample chapters from Manning Site and am having a look on it... Thank you again for your recommendation...

[Pradeep bhatt]

Ko Ko,
Did you sleep last night or spent the whole night posting ?

[Ko Ko Naing]

Originally posted by Pradeep Bhat:
Ko Ko,
Did you sleep last night or spent the whole night posting ?

Yes, I'm in Thailand... Not America... I guess you overlooked the time difference...
I just want to help people out here in this forum...

[Pradeep bhatt]

How many hours is it ahead from GMT?

I just want to help people out here in this forum...

I know that.

[Ko Ko Naing]

Originally posted by Pradeep Bhat:
How many hours is it ahead from GMT?

I know that.

Thailand is 7 hours ahead GMT... Currently it's 4:30 P.M. in Bangkok...

[Pradeep bhatt]

It is just 1.5 hrs ahead of IST(Indian Standard Time).

[Ko Ko Naing]

Originally posted by Pradeep Bhat:
It is just 1.5 hrs ahead of IST(Indian Standard Time).

So it's 3:00 P.M... I guess you are still in your office...Aren't you?

[Pradeep bhatt]

Yes Sir