Home

News: BEA WebLogic Server 9.0 and AquaLogic Service Bus Released

  1. BEA has released to general availability both WebLogic Server 9 and the AquaLogic Service Bus, available for download from BEA's web site. Documentation for both is available on EDocs.

    Here is a brief summary of what's new / released, highlighting what I think is important, but much more is available in the release notes:

    WebLogic Server 9.0 is fully compliant with J2EE 1.4, and is focused on simplifying and streamlining the day-to-day management of production systems. Production redeployment of updated web application versions can be done on a "side-by-side" and "sanity check" basis, with HTTP sessions gradually migrating over to the new version. WLS 9 also includes a new diagnostics service, administrative console with domain-wide change control, and full support for the Jython-based WLST scripting tool.

    WLS 9 also features support for wide-area cross-cluster failover to address catastrophic disators, along with automated server migration for services such as JMS. It also provides a new policy-based workload manager, and self-tuning facilities for shared resources, such as executed threads.

    WLS 9 web services support includes JSR-181 annotation-based programming ("Web Services Metadata for the Java Platform"), the Web Services for J2EE 1.1 spec, reliable, conversational, buffered and/or asynchronous web services, support for WS-Policy, WS-Addressing, WS-Security and WS-ReliableMessaging.

    AquaLogic Service Bus (ALSB) is a configuration-based, policy-driven Enterprise Service Bus, based on WLS 9, with message brokering and web service management (WSM) features. It supports Xquery-based routing, many transports including filesystem, FTP, HTTP/S, JMS (including Websphere MQ and JMS/XA), and Email (POP/SMTP/IMAP), many messaging approaches including REST-style XML, WSDL-defined SOAP/XML, plain text or binary, dispatching based on header, XSD type, WS-addressing, transport header, or dynamically based on the message shape.

    ALSB supports synchronous, asynchronous and pub/sub message exchange patterns, full XQuery-based transformation of XML, text, and/or binary data, WS-security authentication, encryption, and signatures, logging, monitoring, and SLA enforcement.

    Comments, anyone? What's WLS 9 missing? How do you think ALSB would fit into your world (if at all)?

    Threaded Messages (80)

  2. The wait is finally over[ Go to top ]

    Congratulations to the BEA team.

    :-)
  3. Diablo[ Go to top ]

    Is this their "Diablo code base " ?
  4. Diablo[ Go to top ]

    Is this their "Diablo code base " ?

    Yes, this is Diablo.


    Mike Jasnowski / WLS Console Team
  5. Does the AquaLogic Service Bus implement JBI?
  6. Congratulations -- I was guessing WLS 9.0 wouldn't be released till eWorld in September. Maybe 9.1 will be announced at eWorld :)

    --Vinny
  7. New WLW ?[ Go to top ]

    Does it include new Eclipse-based Weblogic Workshop?
  8. New WLW ?[ Go to top ]

    Does it include new Eclipse-based Weblogic Workshop?

    No. The new WebLogic Workshop will come by the end of this year.

    Jesper Joergensen
  9. Is there any plan for releasing new WL workShop ?
  10. Think Liquid[ Go to top ]

    In fact BEA now has enterprise products to manage services around the "Think Liquid" concept. This means more transparency in the enterprise which disentangles enterprise infrastructure problems.

    In addition to BEA AquaLogic Service Bus BEA have BEA AquaLogic Data Services Platform™, BEA AquaLogic Enterprise Security, and BEA AquaLogic Service Registry™ in the product line.

    AquaLogic Data Services Platform (a.k.a Liquid Data) is a SOA based Enterprise Information Integration (EII) tool. It introduces the concept of "Data Services" which can enable seamless integration. BEA’s XQuery Engine forms the heart of this product.

    BEA AquaLogic Enterprise Security is a SOA based security infrastructure. BEA AquaLogic Service Registry is a SOA based service registry.
  11. RE: Think Liquid[ Go to top ]

    Yes, yes... I got all of that from the website / marketing material. But does it implement JBI, since it seems to be exactly what JBI is specifying?
  12. RE: Think Liquid[ Go to top ]

    I am not sure if they would implement JBI after abstaining from the ballot........

    http://www.jcp.org/en/jsr/results?id=3226
  13. AquaLogic Service Bus Released[ Go to top ]

    Congrats to BEA for shipping this.

    Does this mean that the web is out and water is in? Is water faster than the web?
  14. What's the difference?[ Go to top ]

    I found the following products from BEA which seems confusing. Can anyone explains how these are different?

    Diablo - code name for WebLogic 9.

    "Diamond" - Is it the code name of QuickSilver?

    "QuickSilver" - First official SOA product?

    "Freeflow" - another code name which is made up of 'QuickSilver'(?), Liquid data, portals, security etc.
  15. What's in a name?[ Go to top ]

    "AQUALogic" - looks like a new toothpaste in the market. Doesn't it freshen up our brains...or ...?
  16. What's the difference?[ Go to top ]

    Yeah, the code names get a little tangled. Here's the secret decoder ring:

    Diablo is now WebLogic Server 9.0 (as you said). The bits for WLS 9 are live on our download site, and we're doing an official launch--with lots more info--on August 8. So stay tuned!

    Diamond is WebLogic Platform 9.0. This will include WebLogic Server, WebLogic Portal, and WebLogic Integration as in the past; however, it will be a staggered release rather than a simultaneous release. So far WebLogic Server 9 is the only WebLogic Platform product released so far.

    Quicksilver is now AquaLogic Service Bus 2.0.

    Free Flow is the code name for the whole BEA AquaLogic product family. Currently available AquaLogic products are: AquaLogic Service Bus, AquaLogic Data Services Platform (fka Liquid Data), and AquaLogic Enterprise Security, with other future products to be named later.

    And, finally, AquaLogic DOES in fact freshen up your brain! So go download it!

    Thanks,

    John Doppke
    BEA Systems
  17. EJB 3.0 ?[ Go to top ]

    Does it support EJB 3.0? JBoss already does!
  18. EJB 3.0 ?[ Go to top ]

    Great question... WLS 9.0 does not support EJB 3.0. As you probably know, the EJB 3.0 JSR is not yet final. Naturally we're working on EJB 3.0, but we don't want to drag our customers (and our engineers) on the "spec roller coaster" that could still easily happen with EJB 3.0.

    As an illustration: remember what happened with local interfaces suddenly appearing in EJB 2.0 draft 2? That caused a lot of pain for our customers who were already crafting solutions based on our early release of draft 1. We don't want to get burned again with early adoption re-coding and support issues as we did back then.

    So, as a mainstream, responsible software company, we are waiting until the actual EJB 3 specification is final before releasing it to our customers.

    John Doppke
    BEA Systems
  19. EJB 3.0 ?[ Go to top ]

    That's good to hear, John... Sounds like BEA's grown up :-)

    Now, about newly released specs... JBI for instance... ?
  20. lol[ Go to top ]

    "So, as a mainstream, responsible software company, we are waiting until the actual EJB 3 specification is final before releasing it to our customers."

    Got that JBossians?
  21. ROFL[ Go to top ]

    So, as a mainstream, responsible software company, we are waiting until the actual EJB 3 specification is final before releasing it to our customers.

    ROFL

    So, let me get this straight. Both Oracle and JBoss are non-mainstream, non-responsible, software companies, just because we have successfully implemented the EJB 3.0 spec before BEA? Are you getting all this, Mike?

    Now, certainly there will be minor changes to the final draft. But as everyone here knows, the spec is already at the *public draft* stage, where preview implementations are expected, according to JCP rules. I don't think I'm speaking out of place to say that neither I nor the rest of the expert group expect to see any truly disruptive changes to the spec in the next months before we go final.

    By the way, I'm sure there is plenty of previous precendent of BEA releasing preview implementations of specs during the public draft stage. Were BEA non-mainstream, non-responsible then?

    So, anyway, given that John just took a little swipe at us, I hope you guys will indulge we in a little swipe back. :-)

    In particular I'm wondering what BEA recommends their customers do if they are starting a new application now?

    Option 1: keep using EJB 2.1 for new development. Hmmmmm. Greaaaat option. I'll wait and let the giggles die down.
    Option 2: use Hibernate + Spring. Fine by me ;)
    Option 3: Switch to Oracle or JBoss. Fine by me ;)
    Option 4: not build any new applications, until BEA catches up

    Which option is it then John?


    P.S. We will soon release an EJB3 container that can be embedded in WebLogic, so this will add another option for WebLogic users.
  22. ROFL[ Go to top ]

    Gavin,

    No problem, swipes are fair game :-)

    I'm glad to hear that you aren't expecting any major surprises during the EJB 3 public review. Note, though, that in my example of EJB 2, the spec had reached public review before local interfaces were added four months later. So dramatic changes can happen, even at this stage.

    More generally: Obviously all companies in this industry must strike a balance between innovation and stability. As the Java industry in particular has matured, we've increasingly heard from our customers that they want stability and investment protection. This doesn't mean that we never head out in front of the spec--we have done so in the past, and we'll continue to do so in many cases--but it does mean we have to be responsible in balancing the benefits of future technologies against the costs, to us and to our customers, of going there early.

    So we continue to evaluate where we want to be on a case-by-case basis. For EJB 3, we're opting not to release an implementation until things settle down. I haven't heard of a single customer who is thinking of using EJB 3 in real projects today--that would be extremely aggressive, far too aggressive for everyone I've talked to. By contrast, I've heard of a few people who are curious about EJB 3 and may want to try it out a bit. That's fine. When EJB 3 is ready, we'll be there to support it too. In the meantime, we'll still be providing top-notch, enterprise-grade support for the many technologies (including many new standards in WLS 9) that we have already implemented in our products.

    John Doppke
    BEA Systems
  23. ROFL[ Go to top ]

    Note, though, that in my example of EJB 2, the spec had reached public review before local interfaces were added four months later. So dramatic changes can happen, even at this stage.

    And I bet you guys noticed, because I bet you had an implementation out at the time :-)

    And note that *adding* things (like local interfaces) is not a breaking change. I fully expect we will *add* things to EJB 3.0. In fact, there are certain things that I am right now working to add. That is not a reason for users to not adopt. The question is whether there will be major breaking changes. I do not expect this, and that is not the goal of the EG.
    I haven't heard of a single customer who is thinking of using EJB 3 in real projects today

    Go to the Hibernate and JBoss forums, and you will see numerous posts from people using EJB 3.0 persistence and EJB 3.0 today. Certainly these are not *production* projects, but obviously it is necessary to have implementations of the spec out months before people start going into production. We are aiming to satisfy people who want to deploy EJB 3.0 applications into production on a timescale of approx 6 months from now (at which point the spec certainly *will* be finished).
    When EJB 3 is ready, we'll be there to support it too.

    I'm honestly looking forward to it. A lot of my blood sweat and tears is in that spec and I will love to see you guys release an implementation. I hope it's not too far away.
  24. ROFL[ Go to top ]

    Note, though, that in my example of EJB 2, the spec had reached public review before local interfaces were added four months later. So dramatic changes can happen, even at this stage.
    And I bet you guys noticed, because I bet you had an implementation out at the time :-)
    Indeed we did, and it was quite a learning experience, at least for me :-)
    And note that *adding* things (like local interfaces) is not a breaking change.
    True... The big painful thing I remember from EJB 2 was that not only were local interfaces added, but they were suddenly required in order to use EJB relationships. So it wasn't the addition that caused the problem; it was the change in how relationships were handled.
    Go to the Hibernate and JBoss forums, and you will see numerous posts from people using EJB 3.0 persistence and EJB 3.0 today. Certainly these are not *production* projects, but obviously it is necessary to have implementations of the spec out months before people start going into production.
    Yes, assuming the APIs don't shift beneath people's during those months. We've all suffered through that on other technologies (I can think of numerous examples from my own experience). It sounds like you guys on the EG for EJB 3.0 are trying to ensure no such API changes come up, which is very encouraging.

    John Doppke
    BEA Systems
  25. ROFL[ Go to top ]

    Yes, assuming the APIs don't shift beneath people's
    ...feet...
    during those months.
    (editing error)

    John Doppke
    BEA Systems
  26. ROFL[ Go to top ]

    I would like to add to John's points. WebLogic Server 9.0 should not considered an "API-focused" release. Having said that, a significant amount of work and innovation went into this release including:

    Building an extremely robust and performant messaging infrastructure which includes:
    - store-and-forward messaging
    - automigration of messaging destination for high availability
    - simplified administration of remote endpoints
    - greater visibility and control over messages
    - significant performance improvements

    Improving application management through:
    - advanced automation capabilities
    - disruption free redeployment (side-by-side redeployment) and server upgrades
    - Monitoring and diagnostics capabilities providing non-disruptive control of data acquisition and management

    Enhanced enterprise Web Services with:
    - reliable delivery over HTTP
    - Conversational web service framework
    - improved asynchrony support

    So to imply that BEA is no longer delivering innovative technology ahead of our competitors is just plain wrong and missing the point. Clearly, a real enterprise application server is a lot more than just the APIs it supports. Its about delivering a reliable and performant platform for deploying and managing enterprise applications. Supporting the latest and greatest APIs is only a small part of delivering on that promise. WebLogic Server 9.0 is in fact a very significant release that solves real problems. I welcome you to download it and try it out for yourself.

    Jim Rivera
    BEA Systems
  27. At the end of the day...[ Go to top ]

    Having worked with the BEA WebLogic line of products since the late 90s, I would say that BEA would be crazy not to further build on their excellent infrastructural track record.

    Most of us are developers, but it would be interesting to know how many of you also have worked with application and system administration. Well I have, and what I can firmly say is that what I appreciate the most when sitting there and watching exceptions fill the logs and the application turning on its back is not whether the application server has support for the latest and sexiest APIs.

    No, I rely on the built-in thread pools to be able to configure and isolate certain requests, thereby avoiding the whole system from going down when e.g. a back end system stops responding or any other resource becomes unavailable. I turn to the administration console to get a bird's eye view on the system - thread pools, connection pools, heap usage, etc. And I rely on the built-in clustering to ensure that customers get a smooth ride even though server nodes take a nose dive.

    As developers we tend to devote all our energy and interest to new features and APIs and ignore the fact that what our customers (those big mouth/litte ears folks that give us our pay check/pay our consultant fees every month) want is a stable and performant platform on which to run their business.

    Not to say that JBoss, WebSphere etc. won't give you that, but if were the customer, or say a consultant acting loyally to my customer (mind boggling, I know), I would almost always choose proven infrastructure before the absolute latest API support. Then again, there are issues such as price, previous business relationships and investments, strategy considerations (open source or not), so once again, I'm not saying that WebLogic is better than JBoss, just that I think BEA did the right thing.

    That said, I hope that they don't underestimate the momentum and strategic value of the Open Source/innovation is king/API focused movement.
  28. At the end of the day...[ Go to top ]

    ... so once again, I'm not saying that WebLogic is better than JBoss, just that I think BEA did the right thing. That said, I hope that they don't underestimate the momentum and strategic value of the Open Source/innovation is king/API focused movement.

    Your point is well taken, and we are certainly very excited about what is happening in open source. It has been the breeding ground for some of the most significant innovations over the last couple of years. In fact, BEA has helped to drive some of this innovation with projects like XMLBeans, Beehive, and Eclipse WTP. The added transparency and portability of these projects can offer a level of investment protection otherwise not available from non-standard technology. Its proving to be a very nice and faster moving compliment to the open standards effort and has already done much for increasing developer productivity.

    Having said that, we also clearly understand the value of a reliable, performant, and manageable deployment platform for putting these technologies into production (hence our investments in the new WebLogic Server 9.0 functionality).

    Fortuntately the two ideas are not mutually exclusive. With a little integration work, the value you get from deploying on WebLogic Server can be retained while using leading open source application frameworks without impacting the portability of your application code. Along these lines, we recently announced support for the Spring framework, demonstrated some early integration work during our JavaOne keynote, and are investigating additional opportunities for integration:

    http://www.theserverside.com/news/thread.tss?thread_id=34875

    Jim Rivera
    BEA Systems
  29. ROFL[ Go to top ]

    Yes, assuming the APIs don't shift beneath people's during those months. We've all suffered through that on other technologies (I can think of numerous examples from my own experience). It sounds like you guys on the EG for EJB 3.0 are trying to ensure no such API changes come up, which is very encouraging.

    Actually, we are more focused on ensuring that the spec is right for what people want. I certainly won't guarantee that the API's won't change before the spec becomes final It is far more important that it be right than it be unchanged during the final stages. We are going through the spec right now and adjusting things that need to be adjusted before it goes out the door.

    I believe that people that have decided to use the EJB 3.0 preview releases from Oracle, JBoss and now SolarMetric have benefitted quite a lot. They have been getting a lead on developing their applications to the EJB 3 spec and have also been able to provide input and influence on what should get included or excluded in the spec.

    If BEA wants to wait until the spec is final that is okay as well. On the one hand they won't have to keep up with the task of implementing to a slightly moving target (which really is a bit of a pain, BTW) but of course they also won't be seen as a leader in the technology. It's a business and resource decision that every company has to make and what might be right for one company may not be right for another.

    -Mike
  30. ROFL[ Go to top ]

    It's a business and resource decision that every company has to make and what might be right for one company may not be right for another.

    I couldn't agree more. Our decisions for WLS 9 have been guided by what our customers have been asking for most. As Jim pointed out, for WLS 9 we've focused primarily on advances in administrative capabilities, performance, and web services. These areas are, far and away, most of what we've been hearing customers ask for--it's the stuff that really lowers the TCO of an application server.

    John Doppke
    BEA Systems
  31. ROFL[ Go to top ]

    As Jim pointed out, for WLS 9 we've focused primarily on advances in administrative capabilities, performance, and web services.

    When we are talking administration, I have to rub my eyes. I looked at the 9.0 preview with the new administration console. There was plenty of reason to dislike the old one. Finally I have a good reason to like it after all: The new one.

    I also find it somewhat unwise that BEA is pushing the server out before having at least something to show in the realm of workshop, portal and integration. What am I to think as a customer who is contemplating making a major investment in WebLogic Integration or Portal? The new versions will show large changes and improvements.
  32. ROFL[ Go to top ]

    I think what you are seeing is an overall change in development cycles and methodologies. When I was going through college, software development involved a lot of FUD. You go through months of 'talking' and writing about a project before you hit the ground with programming.

    Now days, developers are taking a much more agile approach. Engulfing themselves in technology and tools at a much more early stage in development while providing a quicker time to market.

    This is why you see a lot of developers switching to open source 'standards'. While your project's team is in development, so are the tools and frameworks you choose. Frameworks, tools, and technology, in general, ramp up as fast as the software you write. By the time you finish a project-- do you want it founded on technology from 2 years ago, or today?

    The point that EJB 3 shouldn't be approached until 'things settle down' shows complete ignorance to the way software is developed today. I'm *stuck* in weblogic world at my company and *still* only able to use JSP 1.2-- did it take almost 2 years after the JSP 2.0 spec final draft for things to 'settle down'?

    - unaffiliated
  33. A change in methods[ Go to top ]

    I think what you are seeing is an overall change in development cycles and methodologies. When I was going through college, software development involved a lot of FUD. You go through months of 'talking' and writing about a project before you hit the ground with programming.

    That's not exactly FUD so much as dealing with large numbers of people requires large amounts of oral and formally written communication. Agile approaches focus on smaller teams and oral and ("written oral", i.e. IM & Email) communication.
    Frameworks, tools, and technology, in general, ramp up as fast as the software you write. By the time you finish a project-- do you want it founded on technology from 2 years ago, or today?

    OSS has its dysfunction: "people co-operate on a project, until they don't feel like it, and then fork or start their own pet project that 5 people will use".

    Standards bodies have the "take 2 years co-operating until no one cares to use the standard" dysfunction.

    Arguably, yes, there definitely a place for "defacto OSS standards" -- things that don't have a formal standards body, but have a large level of support AND (and I think this is key) a core team that has very strict guidelines for official release status. This is largely why I think Struts became so popular.

    Having said that, there's lots of cases where a standards body helps to organize and clarify how groups can co-operate on interface / compete on implementation. It just takes more time.

    There is be a good argument that J2EE was standardized too early, which has impeded progress, hence the explosion in OSS innovation.

    Let's also not forget the third model: single-vendor-proprietary with a large base of support (Microsoft .NET) + a clone market (like Mono, Rotor, etc.). Given the noise in the Java OSS world, MIcrosoft's world seems easier at times, though their OSS community is certainly ramping up in popularity.
    I'm *stuck* in weblogic world at my company and *still* only able to use JSP 1.2-- did it take almost 2 years after the JSP 2.0 spec final draft for things to 'settle down'?- unaffiliated

    Well, I don't speak for BEA engineering, but to me it looks like there was a priority shift. Application servers have a hard time competing solely on J2EE standards compliance; being J2EE 1.4 compliant isn't going to send the world stomping to your door. Look at IBM, their slower approach to standards compliance certainly hasn't hurt their WebSphere marketshare!

    The focus for WebLogic 9.0 was on availability, system administration, diagnostics, and upgrades, a major upgrade to messaging, and more advanced web services support. There are huge improvements in all of those areas. But picking "tougher problems" means that it took longer.

    Stu Charlton
    BEA Systems (Consulting)
  34. A change in methods[ Go to top ]

    Application servers have a hard time competing solely on J2EE standards compliance; being J2EE 1.4 compliant isn't going to send the world stomping to your door. Look at IBM, their slower approach to standards compliance certainly hasn't hurt their WebSphere marketshare! The focus for WebLogic 9.0 was on availability, system administration, diagnostics, and upgrades, a major upgrade to messaging, and more advanced web services support. There are huge improvements in all of those areas. But picking "tougher problems" means that it took longer.

    Stu Charlton
    BEA Systems (Consulting)

    Your points are very valid, but as others have commented, people are looking for particular features within a reasonable amount of time. These aren't niche features per say, but standards included in things like JE 1.4 or JE 5. While BEA has taken the time to really emphasis certain areas, they haven't given the impression that they committed to keeping up with other standards within reasonable timeframes. This is fine, it's your business plan.
  35. Hallucination[ Go to top ]

    "
    I haven't heard of a single customer who is thinking of using EJB 3 in real projects today.
    "

    Wow...I need medication. I imagined myself on several Hibernate consulting engagements for WebLogic and WebSphere (as well as JBoss) customers who weren't sure when they could go to EJB3 because their appserver wasn't supported on JDK 5.0, but were anxious to go to EJB3. What was really funny is that I woke up from these engagements a week or so later. I even imagined many frustrations with airlines. Worse, I continue to talk to some of these people long after!

    from http://e-docs.bea.com/wls/docs90/notes/new.html
    "
    Now a J2EE standard, Web Services increase developer flexibility and choice by providing a common run-time environment and industry-standard support for Java annotations and Web Services extensions.
    "

    So ummm... your proprietary annotations which will inevitably have to be migrated when J2EE web services embraces annotations are ready for prime time? That's not too ahead of the spec? Oh wait...that's spec umm 181: http://jcp.org/en/jsr/detail?id=181 -- I wonder who leads that spec:

    Specification Lead
    Brian Zotter BEA Systems

    Hummm.. There is probably a message in there somewhere.
  36. 181 is final[ Go to top ]

    Andrew, JSR181 is already final and will not undergo any dramatic changes. (or any change at all) I don't see any relevance of this spec to the discussion of the EJB3.0 spec which is only in public draft stage.
  37. 181 is final[ Go to top ]

    It wasn't final when the 9.0 release was burned to disc for sure (since it only became final Jun 27).
  38. 181 is final[ Go to top ]

    It wasn't final when the 9.0 release was burned to disc for sure (since it only became final Jun 27).

    What's your point? BEA's not the only ones using JSR-181. I've been using the JSR-181 annotations with XFire to auto-register web services for 6 months under Tomcat. The annotation classes are in the Beehive jar, it's not like you have to include weblogic.jar in your classpath.
  39. 181 is final[ Go to top ]

    It wasn't final when the 9.0 release was burned to disc for sure (since it only became final Jun 27).

    Yes, it only became final on Jun 27, but the 9.0 release was burned after that.
  40. lol[ Go to top ]

    "So, as a mainstream, responsible software company, we are waiting until the actual EJB 3 specification is final before releasing it to our customers."Got that JBossians?

    Yes, absolutely, what I got was that a once-proud, once-innovative *technology* company just shitted on the idea of being ahead of the curve, and ahead of competitors. :-)

    How sad.
  41. lol[ Go to top ]

    "So, as a mainstream, responsible software company, we are waiting until the actual EJB 3 specification is final before releasing it to our customers."Got that JBossians?
    Yes, absolutely, what I got was that a once-proud, once-innovative *technology* company just shitted on the idea of being ahead of the curve, and ahead of competitors. :-)How sad.

    Well, it's worked for IBM... sort of...
  42. lol[ Go to top ]

    "So, as a mainstream, responsible software company, we are waiting until the actual EJB 3 specification is final before releasing it to our customers."Got that JBossians?
    Yes, absolutely, what I got was that a once-proud, once-innovative *technology* company just shitted on the idea of being ahead of the curve, and ahead of competitors. :-)How sad.
    Well, it's worked for IBM... sort of...

    Well, yes, but BEA used to pride themselves in being ahead of IBM with spec implementations. It's pretty nice for us, as BEA competitors, to see them let themselves fall behind, but on the other hand I'm frankly sad, I used to think BEA were a great company, with great technology leadership. And I've certainly been looking forward to seeing them implement EJB 3.0. I certainly didn't expect to see BEA folks bashing us and Oracle in public over this issue.

    Anyway, I hope John is just speaking out of turn and the responsible guys at BEA are working to get their EJB 3.0 implementation to market.
  43. lol[ Go to top ]

    I certainly didn't expect to see BEA folks bashing us and Oracle in public over this issue.

    Sorry, let me clarify... I don't mean to bash EJB 3 or those working on it. We're very eager to see EJB 3 get to "prime time" and will be there when it is. But right now our customers need solutions on technology available today, and that's what we're focusing on.

    John Doppke
    BEA Systems
  44. lol[ Go to top ]

    I certainly didn't expect to see BEA folks bashing us and Oracle in public over this issue.
    Sorry, let me clarify... I don't mean to bash EJB 3 or those working on it. We're very eager to see EJB 3 get to "prime time" and will be there when it is. But right now our customers need solutions on technology available today, and that's what we're focusing on.John DoppkeBEA Systems

    OK, understood.

    But, just so you realize, this technology *is* available today:

    http://www.hibernate.org/Projects/HibernateEntityManagerForEJB3

    :-)
  45. OK, understood.But, just so you realize, this technology *is* available today:

    Well, as long as we're turning this into an EJB3-preview-release-advertisement thread, we might as well be complete about it:

    Kodo 4 Early Access: http://www.solarmetric.com/kodo/beta/4.0.0EA
    Oracle's EJB3 preview: http://www.oracle.com/technology/tech/java/ejb30.html

    -Patrick

    --
    Patrick Linskey
    http://solarmetric.com
  46. And the list goes on......

    Versant:
    http://www.versant.com/opensource/orm/en-us

    Of course, there is also the APPROVED Eclipse ORM project that compliments the runtime and will have first Milestone release in August.

    http://www.eclipse.org/proposals/eclipse-jsr220-orm/index.html

    Robert Greene
    Versant Corporation
  47. EJB3 will get nowhere without a name such as EJB+ or EJB#, since EJB3 is no longer EJB.
  48. lol[ Go to top ]

    Well, yes, but BEA used to pride themselves in being ahead of IBM with spec implementations. [..]

    Actually, I think that company was called WebLogic. BEA is the company that owned Tuxedo, and they both WebLogic to get into Java.

    But the company called "WebLogic" was indeed always ahead of the curve.
  49. lol[ Go to top ]

    In BEA's defense, there is no legal way for them to release an EJB 3.0 implementation in a GA version of WLS. It's against the JCP rules, and Sun is not shy about enforcing them. BEA could ship an impl in a preview release, or even a beta release, but not a GA.

    In the interest of disclosure, I used to work at BEA.
  50. lol[ Go to top ]

    This is *NOT* completely true as I remember BEA always included early implementation of J2EE features in the past.

    The document proves that WLS 6.0 included features of EJB 2.x before it was finalized.

    http://e-docs.bea.com/wls/docs60/ejb/EJB_whatsnew.html

    "The Enterprise JavaBeans 2.0 implementation in WebLogic Server Version 6.0 will be fully supported and can be used in production. However, be advised that the Sun Microsystems EJB 2.0 specification is not yet finalized, and the WebLogic Server implementation of the EJB 2.0 architecture is based on the most current public draft of this specification."


    regards
    Debu
  51. Keep Reading[ Go to top ]

    http://e-docs.bea.com/wls/docs60/ejb/EJB_whatsnew.html#1069910

    "Before you can use the EJB 2.0 features with WebLogic Server Version 6.0, you must download the EJB20.jar file ..."
  52. Re: lol[ Go to top ]

    Gavin,

      No Production application uses a container based on a prerelease spec. What matters for any serious business application is robustness of the container and acheive 5 9's.

      I am glad BEA realised what end users really need and sad to see a smart guy like u turning into Marc Fleury's Clone.
  53. Re: lol[ Go to top ]

    No Production application uses a container based on a prerelease spec. What matters for any serious business application is robustness of the container and acheive 5 9's.

    Now, this is quite funny, since what I actually wrote was:

    "We are aiming to satisfy people who want to deploy EJB 3.0 applications into production on a timescale of approx 6 months from now (at which point the spec certainly *will* be finished)."

    It helps to actually read what people say before trying to insult them. (Yes, even on TSS, where I realize that the usual rules of adult human interaction are suspended.)

    Furthermore, "robustness" of the container is not in question, since our EJB3 implementation is basically just a very trivial wrapper around existing, mature, robust functionality in Hibernate3 and JBoss4, functionality that is already being used in thousands of production applications today. This is, by the way, also true of the two other EJB3 previews mentioned in this thread. "Robustness" usually has nothing to do with which API is used to access the functionality.

    I am glad BEA realised what end users really need and sad to see a smart guy like u turning into Marc Fleury's Clone.

    Since Marc is not only my boss, but also a friend, and a guy I have an enormous amount of respect for, I'm trying really hard to be offended, but just not succeeding.... :-)
  54. I think both groups on this thread are right in their own way. In late 2003 I flew to Stockholm to meet the JRockit team and to discuss further product direction specifically around application management and runtime performance analysis tooling. At the time I felt a serious tooling scalability issue existed in deployment of many large and complex J2EE applications into a managed IT environment. This issue still exists today though the WLS 9.0 release does offer a step in the right direction. When one does manage to get a J2EE application passed testing it would be nice to have some degree of assurances that it would be deployed into a managed environment that offered flexible and dynamic availability monitoring as well as problem resolution including redeployment to new nodes.

    EJB 3 does offer a simplified programming model that would appear on paper to offer less overhead during development and runtime performance. If one was starting a EJB project today that would be deployed in 9-12 months (when I am sure BEA is offering EJB 3.0 support) then yes it would be important to have something today to work with, ensuring that the project could avail of the performance benefits at deployment. This is what I believe Gavin is pointing out. But John (BEA) is also correct when we consider that migrating from WLS x to WLS 9.0 would take a relatively small amount time and customers would overnight benefit from increased performance and scalability in the underlying structure with an application integrating stratgey that will work with past, present and future J2EE applications whether they are based on EJB 1.1, EJB 2.x and EJB 3.0.


    <plug>
    JInspired recently announced an early access of JXInsight/JDBInsight 3.3 that includes a preview of our EJB 3 trace and profile extension.
    http://www.jinspired.com/products/jdbinsight/downloads/index.html

    An article describing how to integrate with OC4J EJB Preview is also available.
    http://www.jinspired.com/products/jdbinsight/oc4jejb3config.html

    Gavin, we will be publishing an Hibernate EJB 3 integration soon but please note that we recently announced a Hibernate 2.x and 3.0 trace and profile extension.
    <plug>

    Regards,

    William Louth
    JXInsight Product Architect
    JInspired


    "J2EE tuning, testing and tracing with JXInsight"
    http://www.jinspired.com
  55. EJB 3.0 ?[ Go to top ]

    So, as a mainstream, responsible software company, we are waiting until the actual EJB 3 specification is final before releasing it to our customers.John DoppkeBEA Systems

    Correct me if I am wrong: in the history, Weblogic did release some components/features at that time their specifications were not final.
  56. EJB 3.0 ?[ Go to top ]

    Does it support EJB 3.0? JBoss already does!

    No, JBoss does not implement EJB 3 yet. The EJB 3 spec hasn't been released so no one can support EJB 3 yet.
  57. EJB 3.0 ?[ Go to top ]

    Does it support EJB 3.0? JBoss already does!
    No, JBoss does not implement EJB 3 yet. The EJB 3 spec hasn't been released so no one can support EJB 3 yet.

    Jason, this is not correct; you need to re-read JCP rules. The spec is in public draft, and preview implementations are encouraged at this stage. This is obviously very important in terms of validating the spec, and teasing out any last holes and missing features. The kind of thing I'm thinking of is stuff like uuid identifier generation, which no-one in the EG, including me, thought was very important, but a bunch of users have been asking for, so we might need to standardize this feature. The kinds of things I am *not* thinking of is major changes to APIs and operation semantics.
  58. EJB 3.0 ?[ Go to top ]

    Does it support EJB 3.0? JBoss already does!
    No, JBoss does not implement EJB 3 yet. The EJB 3 spec hasn't been released so no one can support EJB 3 yet.
    Jason, this is not correct; you need to re-read JCP rules. The spec is in public draft, and preview implementations are encouraged at this stage. This is obviously very important in terms of validating the spec, and teasing out any last holes and missing features. The kind of thing I'm thinking of is stuff like uuid identifier generation, which no-one in the EG, including me, thought was very important, but a bunch of users have been asking for, so we might need to standardize this feature. The kinds of things I am *not* thinking of is major changes to APIs and operation semantics.

    Right, I was making a distinction between EJB 3.0 and the pre-release which is out now. What's out now may be CLOSE to what the EJB 3.0 spec will be, but it's NOT the EJB 3.0 spec.
  59. EJB 3.0 ?[ Go to top ]

    Jboss 4.03 (in RC1 as by now) _supports_ EJB 3.0 with integration in Eclipse IDE up to the deployment process.
    Look here: http://www.jboss.com/products/ejb3
    and here at the demo: http://trailblazer.demo.jboss.com/EJB3Trail/
  60. BEA is back in business[ Go to top ]

    I very happy for BEA for going live ahead of schedule..
    I feel this as the first step in the right direction.
    Congratulations.
  61. I don't see a clear path in the eDocs for doing this....
  62. What's WLS 9 missing?[ Go to top ]

    Comments, anyone? What's WLS 9 missing?

    Does Weblogic 9 support user defined Standard MBeans?

    Wei Jiang
    Perfecting J2EE!
  63. User-defined Standard MBeans[ Go to top ]

    Does Weblogic 9 support user defined Standard MBeans?
    Wei Jiang

    Yes indeed: check out
    Developing Manageable Applications with JMX (on e-docs). There's a section that specifically recommends using Standard MBeans.

    John Doppke
    BEA Systems
  64. Kung Fu Grip[ Go to top ]

    So essentially, this entire thread *thus far* can be summed up as follows (via illustration of elementary school children on a playground);

    Don: "Hey Gary I just got my super-secret, ultra-cool, ninjafied, kung-foo grip transformer and I'm gonna show all my friends"

    Gary: "Too late, I already have one 'cause my Dad works in R&D for the toy company and he stole a prototype for me. Plus, I already showed all of your friends. Oh, and mine has the mega-moo-moo blaster cannon"

    Don: "What? Well your mega-moo-moo blaster cannon won't work in a few months because they might not make anymore mega-moo-moo ammo. I heard it's going to be changed to a super-hyper-chicken-feed blaster."

    Gary: "Yeah, well you used to be cool."

    Don: "I'm still cool ... but that kid(IBM) isn't cool."

    Don: "Yeah, you're right! Let's go beat him up and take his lunch money."


    Names have been replaced to protect the innocent.
    -Adam
  65. Kung Fu Grip[ Go to top ]

    Don: "Yeah, you're right! Let's go beat him up and take his lunch money."

    Sorry, this should have been said by Gary, not Don!

    -Adam
  66. mac os x support[ Go to top ]

    Does Bea plan to support mac os x ?
  67. mac os x support[ Go to top ]

    Does Bea plan to support mac os x ?

    No, we do not currently have any plans to support MacOS X.

    John Doppke
    BEA Systems
  68. OS X support[ Go to top ]

    At least with 8.1, OS X users could install WebLogic using the generic installer which was the AIX version. This worked great and I and a number of people use OS X for daily development on WebLogic even though we are targeting deployments on Windows, Linux, and Solaris. With version 9, this generic installer has been ommitted and the installers provided are tied to the OS since they call native libraries to do stuff like check disk space. I really think BEA needs to provide a 100% java(read portable) installer package so that there is at least a way to get around this or risk losing a small but influential group of developers who prefer to use OS X for development.
  69. OS X support[ Go to top ]

    While I don't expect BEA to officially support OS X I'd still like to at least be allowed to install it and support it myself. It seems like they'd already have a generic installer working since there are many folks at BEA who currently run Weblogic on their Macs. So how about it BEA?
  70. Congratulations on the release.

    I'm also looking forward to Hibernate + Spring support in WebLogic. Gavin, how much of jboss does this "embeddable EJB3" container bring with it? How many features of WebLogic does it leverage versus replace?

    Also, Oleg, what is your background with EJB3? Are you using it?

    Peace,

    Cameron Purdy
    Tangosol Coherence: Clustered Shared Memory for Java
  71. Gavin, how much of jboss does this "embeddable EJB3" container bring with it? How many features of WebLogic does it leverage versus replace?Also, Oleg, what is your background with EJB3? Are you using it?Peace,Cameron PurdyTangosol Coherence: Clustered Shared Memory for Java

    We are committed to providing a full embeddable EJB3 container that takes advantage of as many platform facilities as possible. However, the first release(next week sometime) is mainly targetted at standalone apps, JUnit testing, and standalone Tomcat and will use Jboss specific lightweight services. Subsequent releases will be able to take advantage of some services offered on the platform it runs on. We will of course be restricted by the openness of the platform, so not all services on that platform may be usable.

    Bill
  72. x864-64 jrockit support[ Go to top ]

    Any plans on supporting x86-64 jrockit? It'd be nice to break that heap size barrier on 32 bit machines.
  73. The link
    http://e-docs.bea.com/wls/docs90/programming/overview.html

    is showing that the Weblogic server implements J2EE 1.3. What about 1.4 ?

    Looks to me that the docs are still not updated.

    Thanks
    Pradeep
  74. Various replies[ Go to top ]

    Just to answer a few questions I didn't get to yesterday:
    How to upgrade WebLogic Workshop 8.1 Web Services?
    I don't see a clear path in the eDocs for doing this....
    An upgrade tool will ship with WebLogic Workshop 9 when it becomes available.
    Any plans on supporting x86-64 jrockit?
    Yes. We will support AMD64, EM64T, and Itanium on RHEL, SLES, and Windows 2003. The target for these platforms is end of the year (approximately; the usual disclaimers apply). The good news is that JRockit 5.0 itself is already available on AMD64, EM64T, and Itanium. So it's just a matter of running WLS 9 through the QA cycle on these platforms.
    The link [...]
    is showing that the Weblogic server implements J2EE 1.3. What about 1.4 ?
    Oh you're right, this is a doc bug. Thank you for pointing this out! WLS 9 does indeed implement J2EE 1.4.

    John Doppke
    BEA Systems
  75. Various replies[ Go to top ]

    John, BTW where is WebLogic Builder in WLS9? I couldn't find it anywhere...
  76. There are always the EJB3 mad who hijack every opportunity to promote EJB3, from "Message Driven POJO" to "BEA WebLogic 9.0".

    Why the threads at TSS titled with EJB3 are so quiet? Does that tell you something?
  77. Thank you!!![ Go to top ]

    Thank you Bea for your hard work and dedication to quality.
  78. It's funny to see how the fact, that Weblogic 9.0 not support "EJB 3.0 Public Review" make it competitors happy.

    Competitors just not get the following points:

    1. Weblogic users love to smoke thick Havana cigars, while their Weblogic cluster just works.

    2. Weblogic users want to move for even thicker Havana sigars, therefore Bea offer in 9.0 for them availability, manageability and other abilities.

    3. Definitely, Bea right NOW working on implementing EJB 3.0, testing it, thinking about additional services for it (weblogic-cmp-rdbms-jar.xml) and so on. Of course, they can't allow become blamed that they not support "EJB 3.0 Final". But till EJB 3.0 not final, why they should care?

    4. EJB 3.0 persistence is important, but not SUCH important (for Weblogic customers at least) as it competitor want to show.
  79. I agree with you. CMP entity beans were never a problem for WLS users - EJB3 are only a big move for those who use other containers.
  80. Who cares EJB3[ Go to top ]

    For me it seems that Bea competitors wants to highlight that they are the technology leaders and big companies like Bea has been sloved down. BUT I think the Java world have to change.
    We are operating a lot of Weblogic,SunONE,Websphere etc. at T-mobile; and the most important for us the stability,support and ease of operation of an application server. Implementing the new standards are important; but mainly for the developers. The big money comes from the operation side; they will buy clustered licenses, they are pay the big support fees to the companies; so they have much more influence in the platform war that most people thinks. Most of the big tenderds starts with "you have to develop to this internal platform standards:..". So I think Bea Weblogic9 is a right step to a right direction to help Bea maintain it's leadership in the most importamt companies. Maybe they loose some home grown or small business application because some developers can't live without the latest standards but I think it is cheapest than risk stability. Sorry for my poor english.

    ps. Sun have the same strategy, the new J2SE5 more operational feautures than new Java features. It should continue because the JVM is still a big black box compared to Unix, Oracle etc..
  81. no comments on Aqualogic?[ Go to top ]

    I am not sure why people makes such as big deal of WLS 9 not implementing EJB3. Does implement the latest non-final spec/API so important that all other great features comes from WebLogic 9 do not worth mentioning?

    Besides, from this thread, I don't see many comments on Aqualogic Service Bus, though it is an important part of the annoucement. Any one?