Discussions

News: BEA unveils open source vision; support for Spring Framework

  1. In BEA's Java One keynote yesterday, CTO Mark Carges acknowledged that the majority of projects are combining open source and commercial tools to get work done. In addition, "there are a tremendous amount of innovations coming out of the open source community today." BEA is being practical about the growth of open source and particularly frameworks that work at different tiers in an enterprise app.

    However, Mark made the important distinction that frameworks are just frameworks – they assume that there is a robust runtime underneath such as transactions and security, that they can leverage, and that's where appservers like BEA, IBM come in. BEA's new vision for modern software development is a combination of open source and commercial, where open source frameworks provide ease of development and useful abstractions for developers, but which rely on professinal, enterprise grade run times below the surface.

    Thus, to better support their customers, Mark announced that BEA will be certifying a number of open source frameworks on top of weblogic server, and offering support for those frameworks. Specifically, BEA will be supporting the Spring framework, and partnering with Interface21 to take on 2nd and 3rd layer support.

    They will also be supporting Beehive, Struts, and JSF, but interestingly they will also support using Tomcat and Geronimo in a development environment, making it easy to migrate to WebLogic for a deployment.

    "The way we see the world is a blended mode of combining open source and commercial tools to solve problems".

    To illustrate, Rod Johnson was invited to the stage and illustrated how Spring on WebLogic can work well together. An interesting distinction Rod made is that Spring doesn't compete with WebLogic, it simply hides the need to work with proprietary API's. For example, Spring provides declarative transaction management, and when working on WebLogic, it can work behind the scenes with native BEA API's to provide more robust transaction management. In reverse, BEA’s sophisticated JMX-based management console can also integrate with Spring, allowing Spring beans to be managed from the WebLogic console.

    More information will be added to the TSS day 1 coverage, which will be posted shortly.

    Threaded Messages (37)

  2. Ouch[ Go to top ]

    In other words, "JBoss's model works better than ours."

    I really don't know how this strategy is going to help BEA. It's like making a public claim that they are a sinking ship.

    Let the flames begin.

    -James
  3. Ouch[ Go to top ]

    I don't agree with your analysis. They are simply acknowledging the fact that people want to work with simple to use open source frameworks, but at the end of the day you, based on your requirements, may still need an application server to provide robust runtime features such as XA transactions, management/monitoring, high availability, and all those other critical features that tend to be over looked in open source, or not the core purpose of the framework.

    Floyd
  4. Ouch[ Go to top ]

    ...XA transactions, management/monitoring, high availability, and all those other critical features that tend to be over looked in open source, or not the core purpose of the framework.Floyd

    Quite correct, but will you ever get those with WebLogic? Does depend on the framework of course, but if say, the open source servlet container does not support certain diagnostic features, it will never provide them if managed from WebLogic without a very significant investment from BEA.

    Also, this integration may very well "pollute" your original code, as shown by various flavors of annotations in the JavaOne keynote that would need concurrent management. And finally, a certain framework usually needs a certain (or various) programming style(s) to be used effectively and they might well be incompatible.

    Sort of reminds me of developing applications that are to be deployed on Oracle with MySql. While this is not generally bad, these "open" approaches to using certain resource classes (in this case databases) usually have very severe performance implications that you need to work around by bringing in another framework (say OR Mapping and caching) that in turn adds constraints that in turn require.....
  5. Do you question if Struts pollutes your original code. We would all be coding business logic and database acccess code in JSP and servlets.

    Sounds like BEA is embracing what has critcal mass now. Should be a good way to differentiate from IBM and bridge the not-EJB swarm.
  6. Ouch[ Go to top ]

    I don't agree with your position. This is, IMO, a huge endorsement of the Spring approach. First, this is as clear a BEA endorsement of the IOC approach, in general, as is EJB approach. You don't see BEA supporting every software design approach under the sun. Second, there are about 4 different IOC containers and BEA selected Spring. This is an equally clear endorsement of Spring.

    Despite protestations to the contrary, hundreds if not thousands very sharp people are poring over Spring, its code, and its philosophy and have found it valuble.

    As for the "certain framework, and programming style" that you imply Spring has, are you suggesting that Spring's encouragement of coding to interfaces is also suspect? Or perhaps their use of config files? Or their programmable configuration? If so, you'd better had hundreds of more products to your list of problematic software.
  7. Ouch[ Go to top ]

    You don't see BEA supporting every software design approach under the sun. Second, there are about 4 different IOC containers and BEA selected Spring. This is an equally clear endorsement of Spring.


    Well from the talk you could get the impression that they are indeed supporting every framework under the sun (or intend to). Just consider that BEA is already (sort of) providing Beehive, essentially an annotation driven IoC approach. They ccommit to _supporting_ Genronimo and Tomcat. Not just enable reuse, but _support_.
    Despite protestations to the contrary, hundreds if not thousands very sharp people are poring over Spring, its code, and its philosophy and have found it valuble. As for the "certain framework, and programming style" that you imply Spring has, are you suggesting that Spring's encouragement of coding to interfaces is also suspect?


    It is not suspect it is specific. And its core toolset (push configuration, basic AOP and so on) I find quite nice.

    The more programming models you will support at the same time the more stretched will your support become. As with Spring, my firm believe is that it is not so much the "programming model" or the "philosophy" that attracts people but the toolset. So essentially people use Spring because it is sort of the 200 Pound Gorilla of Push Configuration.

    In fact they care more about the tools that it gives you for quickly putting together a web app, manage transactions, add code tracers etc. And you know what: That is fine. As for programming by interface: Fine as well, but reallistically EJB forced you to program by interface as well. But people found it somewhat clumsy (heavyweight). And not so much for the programming model but for codeing/testing turnaround times.

    If the "programming model" would be the decisive factor in the toolset the very smart people use, they would certainly not currently pour over technology that breaks basic OO semantics ("generics") or dubios constructs like multiple argument method calls, covariant returns (typesafety has gone too far!) and so on.
  8. Spring seller[ Go to top ]

    Spring is now used as marketing tool to arg that a plataform is worthwhile. Weblogic looks like a nice frontend to manage spring back end beans.

    I don't believe that open source projects over look critical features if those are in scope of the project. These assumptions may be valid, but I need to see the facts.
  9. Ouch[ Go to top ]

    In other words, "JBoss's model works better than ours."

    I really don't know how this strategy is going to help BEA.

    It's like making a public claim that they are a sinking ship.

    James, I'm really surprised that you're not applauding the strategy. I think it's a huge endorsement of Spring from one of the top J2EE server vendors; what's wrong with that?

    Peace,

    Cameron Purdy
    Tangosol Coherence: Clustered Shared Memory for Java
  10. Ouch[ Go to top ]

    I see a non-EJB3 alliance here.
  11. Applauding the strategy[ Go to top ]

    Hey Cameron. Perhaps my comment came off as JBoss bigotry, but on the contrary I do applaud their strategy. They are one step closer to the Professional Open Source model. But this seems to be a catch 22. By moving towards a Professional Open Source model they are clearly stating that their market position has weakened considerably due to companies like JBoss. The positive side is that if they were planning on being bought out they probably would not have taken this route. So maybe BEA plans to create some competition in the Professional Open Source world. From my viewpoint this is a good thing since competition spurs innovation. However, from an investor's viewpoint it would be better if BEA had not done this and instead done all they could to prepare to be bought out.

    -James
  12. Applauding the strategy[ Go to top ]

    By moving towards a Professional Open Source model they are clearly stating that their market position has weakened considerably due to companies like JBoss.

    I see your point, but this is not what BEA announced. In the keynote address they made it clear that they are doing this to meet the needs of their customers. In other words, they will offer support to people who want to use Spring on WebLogic. I don't believe that they are saying they will support Spring just for Spring's sake.

    I know that I get annoyed when I call my ISP for support and they don't offer it to me because I am using a Linksys router along with my cable modem. The same case here - BEA needs to help it's customers that are also using Spring. this isn't a position of weakness, I think they taking the bold step to better support their customers at the risk of having people construe their announcement as weakness, as appears to be happening on this thread. :)

    Floyd
  13. Applauding the strategy[ Go to top ]

    Hey Cameron. Perhaps my comment came off as JBoss bigotry, ..

    I didn't know whom you worked for. I was just referring to the seeming contradiction that it is a bad thing when BEA announces that it is providing customers with a support option for Spring. "Spring good, support good, Spring support bad"? ;-)
    But this seems to be a catch 22. By moving towards a Professional Open Source model they are clearly stating that their market position has weakened considerably due to companies like JBoss.

    Simple fact is that people use Spring, and some of those people are important customers of BEA. I think it's just an obvious common sense move.

    I doubt that BEA is thinking about "moving towards a Professional Open Source model" .. I'd guess that they're trying to build their revenues by offering the things that their customers are showing willingness to pay for.

    BEA is a business. Trust me, if enough of their customers were asking for JBoss or Geronimo support at the right price level, BEA would be offering it.
    The positive side is that if they were planning on being bought out they probably would not have taken this route. So maybe BEA plans to create some competition in the Professional Open Source world. From my viewpoint this is a good thing since competition spurs innovation. However, from an investor's viewpoint it would be better if BEA had not done this and instead done all they could to prepare to be bought out.

    I could be totally wrong, but I don't agree at all. By supporting a well-respected project like Spring (and doing so before anyone else claims the same), BEA has likely increased their perceived value. Further, most successful companies don't sit around hoping to get bought out; BEA is a pretty successful company, even if it is struggling to maintain its billion dollar revenues in a market that has seen most of the unit growth shift to open source.

    Peace,

    Cameron Purdy
    Tangosol Coherence: Clustered Shared Memory for Java
  14. BEA Struggling?[ Go to top ]

    I can't help but respond to Cameron's comment that BEA "is struggling to maintain its billion dollar revenues".

    This is a wide spread perception that simply is not true. Here are the facts. BEA just turned in a record Q1. It was a record Q1 for total revenues ($281M) and a record Q1 operating profit. Last year revenues were up 7%, operating income was $220M, up 10%, 32 consecutive quarters of positive cash flow, up 26% last year and $1.6 billion of cash in the bank, up 10% last year.

    Said anther way, BEA is growing and is very profitable. Yes, our competition likes to point out a decrease in license revenues but they neglect to mention that it was only down 3% year over year in Q1.

    Between new releases of WebLogic Platform (Server, Portal, Integration 9.0), the new AquaLogic product family (Service Bus, Service Registry, Data Services Platform and Security), the WebLogic Communications Platform (SIP Server and Network Gatekeeper), new releases of JRockit (Profiling, Memory Leak Detection…) and a new release of Tuxedo, combined with a far better team and operation, I am personally excited by the prospects.

    Eric
    BEA Systems
  15. Applauding the strategy[ Go to top ]

    I'd guess that they're trying to build their revenues by offering the things that their customers are showing willingness to pay for.BEA is a business. Trust me, if enough of their customers were asking for JBoss or Geronimo support at the right price level, BEA would be offering it.

    Man, I wish I could see faces of JBossers when BEA announces commercial (AKA Professional Open Source) support of JBoss. BEA JBoss sounds good :)

    Regards,

    Slava Imeshev
  16. Applauding the strategy[ Go to top ]

    Hey Cameron. Perhaps my comment came off as JBoss bigotry, but on the contrary I do applaud their strategy. They are one step closer to the Professional Open Source model.

    OH f**k. there isn't a day anymore w/o this proffesional open source shit. Wherever there is a decent discussion about something serious, here they come. BEAS is going proffesional open source and they recognized that JBOSS has stolen all their customers and money. Also IBM has acknowledged that JBOSS stoled all IBM Services customers and now they have sworn eternal war to JBOSS and are buying Geronimo and turning to proffesional open source, so that they can fight JBOSS and not close the doors and go home. You All come to us, all your money belongs to us.

    Come ON. Will ya ? This was a serious thread. BEA acknowledging the value of an OS project (Spring), as SUN has acknowledged the value of another one (Hibernate). No JBOSS. No professional open source. Gosh
  17. This is great news.

    Lot's of projects barely get off the ground. It makes sense to start a lot of them off with Tomcat/Spring/Hibernate etc.
    If they do get popular, great! Easily migrate to a high performance container like Weblogic.

    This makes Java a very smart platform to develop in.
  18. Potentially dangerous for BEA.[ Go to top ]

    This is a very interesting move to me. It shows that BEA knows where the market is going. But how many customers that actually use WebLogic actually need it? You might gain customers in the short term...

    I'd say that BEA will either have to crank up the services and consulting business and support Interface 21, or raise their prices for WebLogic.
  19. I'd say that BEA will either have to crank up the services and consulting business and support Interface 21, or raise their prices for WebLogic.

    I agree with Bruce on this. But in response to the others, how is this NOT a move towards the Professional Open Source model, ie. Services and Support around OSS?
  20. I'd say that BEA will either have to crank up the services and consulting business and support Interface 21, or raise their prices for WebLogic.
    I agree with Bruce on this. But in response to the others, how is this NOT a move towards the Professional Open Source model, ie. Services and Support around OSS?
    I think Dorel may be mixing 2 things here: moving towards POSM is one, and fighting JBoss is another (even though JBoss started waving this POSM banner first). Both things can be totally unrelated, although I am not sure this is the case here...

    Regards,
    Henrique Steckelberg
  21. Spreading thin[ Go to top ]

    I think bea may spread itself really thin here. Also, I don't think the message to customers is very reassuring: Hey we have a gui/ioc/some other framework but we will support a couple of different ones just in case? Does not make to much sense to me,

    I would assume supporting something like pageflow (let alone portal) is quite a job already, let alone adding three other gui frameworks, all getting more complicated by the hour. BEA still has a fine reputation for support and documentation that they put at risk without good reason.

    Also, this happens as a lot of customers would like to see them throw more weight behind integrating with commercial tools in the marketplace, for example Messaging, BI etc.
  22. However, Mark made the important distinction that frameworks are just frameworks – they assume that there is a robust runtime underneath such as transactions and security, that they can leverage, and that's where appservers like BEA, IBM come in. BEA's new vision for modern software development is a combination of open source and commercial, where open source frameworks provide ease of development and useful abstractions for developers, but which rely on professinal, enterprise grade run times below the surface.

    There is an area where opensource does not bring only ease of development and useful abstraction : Spring designed applications comes also with global POJO approach, and on persistence area, Hibernate is very often chosen. Would they help their customer using Hibernate, that is directly concurrent to their J2EE Entity engine ? :-)
  23. OSS support[ Go to top ]

    I can see a two fold benefit;

    1. Developers can use the tools they want and get commercial support from BEA. This lessens the risk for companies and might lead to a greater uptake of OSS projects.

    2. By offering support for the frameworks, BEA would probably get more sales where applications and companies grow.

    Personally, having used JBoss and Tomcat a lot in other companies, I would like to be able have the weight of BEA behind me when justifying architectural decisions and have a clear upgrade path from using an open and free application server to Weblogic should it be required in the future.
  24. Developer licenses for WLS are already free (for a year)... so I am not sure why this approach is attractive (except to avoid introducing any BEA specific code).
  25. BEA listens to the users![ Go to top ]

    This is great news for BEA and a clear indication of where the market is going, i.e. away from classic J2EE with EJBs and into a more Spring-like framework. Rather than being a acceptance of the JBoss model I think this is another nail in their coffin. Users will actually move to BEA because they are actually closer to what the community want.

    This is a very smart move by BEA and of course for the Sring guys. Congratulations on what is probably the biggest movement in the J2EE world for several years.

    -John-
  26. BEA... :-( not good...[ Go to top ]

    After seeing both webcasts: BEA and Oracle from JavaOne (http://wcdata.sun.com/webcast/archives/VIP-1981/) I conclude following:

    - Oracle has overall a better strategy. Oracle supports standards JSF, EJB3, BPEL, etc. and gives a lot of those development tools and frameworks back to Eclipse (EJB3, BPEL, JSF development environment) and Apache (MyFaces). Their strategy is very clear: Use our server for the lifecyle of your enterprise apps, we support standards and we contribute to Eclipse, so that all developers can develop enterprise apps easier. If you need a real world enteprise application with SOA use our server and management tools. Additionally they add Meta Data Management on the top of their solution which is very good, so that they can transform everything to follow Model-Driven development. IMO, Oracle with their platform (maybe because they have bought PeopleSoft) has the same direction as SAP with their NetWeaver...

    - BEA has no long-term strategy at all. They support that (JSF) and that (Spring) or that (Beehive) but at the whole it is very unclear why they do or did that stuffs. I also don't see any use of Beehive. It's just an implementation of BEA but not more than that, no standard, nothing. So, I feel that if I use those products from BEA I'll be dependent on one implementation (Spring, Beehive) - JSF is an exception - and this is not the case with Oracle (JSF, EJB3, BPEL, JBI). One thing I like from BEA is the JRockit implementation. And here BEA beats Oracle (SAP also implements their own JVM, maybe it will be similar to JRockit)...

    So to conclude, if I were a mighty man from Oracle, I would buy BEA:
    - drop all their development tools
    - maybe still use their Weblogic server but try to integrate
      with Oracle App Server for one solution,
    - and for sure GET and INTEGRATE JRockit in my solution!

    ;-)

    Cheers,
    Lofi.
    EJOSA - OpenUSS
  27. BEA... :-( not good...[ Go to top ]

    I also don't see any use of Beehive. It's just an >implementation of BEA but not more than that, no standard, >nothing

    And Yes, I work for BEA as an engineer.

    In regards to this statement, I think this is an unfair characterization. Beehive may have started within the walls of BEA , but is now part of Apache. Struts (while more mature, which Beehive works alongside), is also not a standard (perhaps de-facto). BEA is also contributing code to Eclipse WebTools project. They also support standards, so most of your argument about what Oracle provides is also tru for BEA.

    >but at the whole it is very unclear why they do or did >that stuffs

    From my own perspective and opinion. They are doing it because these are the frameworks with which applications are being written by BEA and NON BEA customers period. To help ensure they are successful, they would like to support the frameworks these people use to build these applications. What's unclear about that? This way you don't need to call different support teams, or scour newsgroups, etc. to find the information you need. It's not uncommon to certify frameworks (app dev or not).
  28. BEA... :-( not good...[ Go to top ]

    I don't want to attact BEA engineers for what they are *not* doing. That was my personal opinion after seeing those webcasts from BEA and Oracle performed by their respective bosses (I'm waiting for the IBM one to see what they are after for ;-))

    <quote>
    BEA is also contributing code to Eclipse WebTools project.
    </quote>

    Yes, this is true (integration of Workshop to Eclipse). It's very good that BEA supports Spring, Beehive, nothing against that. But that's not the point. The key point is that I don't see any *long-term strategy* from the webcast:
    - All app servers can support Spring just like what you said with Struts. No special things.
    - We (as experience devs) all know that BEA will support standards like EJB3, JSF (Java EE), etc. in their app server, but hey, not everyone are experience devs! I did not hear anything mentioned about EJB3 in that webcast. And therefore after looking the webcast I have the feeling that BEA just don't want to implement standards anymore ;-)
    - What is the target customer for Workshop? I also did not hear anything about it, no SOA (BPEL, JBI), no demo. Look at the cool demo from Oracle with JSF and AJAX... ;-)

    In contrast, look at the Oracle webcast: very clear strategy, standards and implementation of standards, no more no less.

    IMO, BEA should stick with infrastructure and deployment that is Weblogic and JRockit. Make those both products as the best ever and don't go into the development area. In this area you can't earn money anyway, because a lot of things are already open source ;-)

    Cheers,
    Lofi.
  29. BEA... :-( not good...[ Go to top ]

    The key point is that I don't see any *long-term strategy* from the webcast:

    First, I also work at BEA and thought this comment was worth directly addressing. Yes, it is true we did not get into the entire company product strategy for SOA and other areas, but Java One is a developer centric audience so we choose to focus our presentation on short set of topics that we thought were most relevant. If you want to learn about all the things we are doing at BEA please visit www.bea.com or dev2dev.bea.com

    Among several areas that our CTO communicated in the Java One keynote was that we are expanding our support for application frameworks because of the increasing trends of using multiple frameworks together and the fact that innovation is happening in these OSS frameworks. Examples include Struts, Beehive, and Spring. We will also continue to provide support for J2EE standards in this area, which means JSF, but the fact is that most customers are mixing and matching these frameworks to best meet their business needs. Our plan is to provide the best solution for these customers and the industry.

    These announcements and product plans are in line with how BEA has always provided enterprise capabilities and products that support the dominant programming models in the industry and then drive innovation and standards into these industry programming models. First we did this with Tuxedo, and next with WebLogic Server. In addition, we have added many product extensions like Workshop, Portal and Integration that provide higher level capabilities and tools to increase productivity and the ability of customers to successfully use IT as a strategic business asset.


    Cheers,
    Shane Pearson
    Platform Product Management
    BEA Systems, Inc.
  30. BEA... :-) not bad...[ Go to top ]

    Does this mean we should see a BEA supported ACEGI security adapter that works with WebLogic?
  31. BEA... :-) not bad...[ Go to top ]

    Does this mean we should see a BEA supported ACEGI security adapter that works with WebLogic?

    I am not going to commit to specific feature support for Spring as that if coming in the near future. Since ACEGI has support for JAAS authentication one would expect this can be used with WL Server and AL Enteprise Security products since they both leverage our extensible JAAS based security framework.

    In addition, security is one of the value added areas that we would expect many developers would want to leverage the BEA JAAS framework so that they do not have to hand-code security policies into their applications. Security along with reliability, performance, production, operations and management are many of the benefits we believe are enhanced for people that will choose to develop and deploy Spring applications or mixed applications that include Spring plus other frameworks on BEA.

    Shane
    Platform Product Management
    BEA Systems, Inc.
  32. hibernate is not in the loop[ Go to top ]

    The hibernate people still have this on their website.

    http://www.hibernate.org/250.html#A23

    Maybe BEA could let them know that this is or has been resolved.

    Still, I think it's great that bea will be supporting open source projects like Spring and Hibernate.
  33. BEA... :-( not good...[ Go to top ]

    We will also continue to provide support for J2EE standards in this area, which means JSF, but the fact is that most customers are mixing and matching these frameworks to best meet their business needs. Our plan is to provide the best solution for these customers and the industry....

    Which I think make sense,valuing customer feedback and keeping an open mind in terms of going forward.
  34. BEA... :-( not good...[ Go to top ]

    Lofi, I share your general feeling about the respective strategies of BEA and Oracle. But I think the reason for this is simply that BEA is in a much more difficult position. BEA sells high priced software in a field where free (as in beer) software is not far behind in terms of features. Sure there are things like massive clustering, high availability, manageability of big installations and so on, where vendors like BEA are still ahead. But how much of the market is this?

    All this stuff about Spring or EJB3 or Hibernate is not tremendously important. These are just variations in programming style. I think people here at TSS (myself included) tend to over estimate the importance of middle tier architectures and frameworks.

    What's much more important for corporate users is data and applications. If you don't do databases and you don't do applications you have to do a very broad range of IT services. You simply can't just sell middle tier software to developers. There's no strategy in the world to make that work in the long term. At least not if you want to be a platform provider. It may work in very small niches if you are very good at what you do and you integrate exceptionally well with a lot of other infrastructure components.

    Let's look at what platform providers typically do:

    Microsoft: Data, middle tier, desktop, applications, OS.
    Oracle: Data, middle tier, applications, services.
    IBM: Data, middle tier, services, systems management, hardware, OS.
    LAMP: Data, middle tier, (increasingly) services, OS.

    The only powerful IT company that doesn't do data is SAP, but they do applications. I can't see what BEA's strategy could possibly be. Maybe buy some stuff that brings in maintainance revenue and become a zombie like CA? I hope they have some smart people there who can figure this one out.
  35. BEA... :-( not good...[ Go to top ]

    <alexander>
    But I think the reason for this is simply that BEA is in a much more difficult position.
    </alexander>

    Alexander, I agree with all your points... All those companies like you said are offering same stuffs (OS, data, middle tier, packaged applications, application development, services).

    I also saw the same thing from IBM webcast today. So for me, IBM == Oracle... It's just the name which makes they different (and maybe quality).

    For BEA is surely hard just to stay in OS (JRockit), middle tier and services to compete with Oracle, IBM, SAP and Microsoft...

    Cheers,
    Lofi.
  36. Meeting our customer needs[ Go to top ]

    Just to throw my two cents in, from a BEA field perspective (though I don't speak for BEA, I speak for myself).

    Almost all of our customers are using Struts these days for new web development. Some are using our Java Page Flow extensions to Struts, especially if they use WebLogic Portal. Many of our large enteprise customers are starting to seriously look at Spring, and a few are already standardized on it for their Enterprise Java development.

    This announcment, in my opinion, has zero to do with Professional Open Source Services. BEA is a software vendor, its future direction is to remain as a software vendor. Software vendors are measured by Wall Street on license revenue growth, and that's an area that BEA definitely seeks to improve: growth through WebLogic, AquaLogic, and our other new products, such as the WebLogic Communications platform. Services -- whether education, consulting, or support, are there to ensure long-term success with the software.

    Now, such services might be everything from business planning and architectural consulting to tuning and troubleshooting. The point is that I don't think BEA is a company looking to drop off a busload of services personnel to make the software work, and then ensure you're dependent on consultants forever. I think BEA ultimately wants customers to be self-sufficient and to have options after they've purchased its software.

    Again, just my opinion.
  37. Tricky, Tricky, Tricky …[ Go to top ]

    I believe that BEA strategy is to diversify its revenue streams through professional open source. And why not, it makes sense. IBM does it through consulting or any sort. I do not see why I need WebLogic to run Spring. If I am part of 1% of the companies that truly need a distributed solution I will just use WebLogic or any other application server, no Spring needed. Otherwise I will use Spring on Tomcat. Anywhere you read about Spring is it hailed as a replacement for EJBs. I am sure Application Server “big boys” do not like that. Sure there are synergies between a lightweight container and a full blown J2EE container, but mostly for already developed applications, otherwise it does not make much sense. I think that BEA had realized that selling Application Server, just by itself is no longer possible or will be impossible in the near future. Also, beyond technical considerations I am suspicious of this type of alliances. “Whatever it is, I fear Greeks even when they bring gifts.”, Virgil (not to offend Greeks. I am alluding to Trojan Horse) It is like wolf teaming up with sheep. BEA is a software vendor as heart. I believe, and I could be wrong on this one, but in my opinion commercial software conflicts with Open Source.
  38. Tricky, Tricky, Tricky[ Go to top ]

    I do not see why I need WebLogic to run Spring
    You don't. Who said you did? It was obvious from Mark Carges's keynote at JavaOne that BEA are happy the fact that Spring is portable. This announcement in no way ties Spring to WebLogic.
    If I am part of 1% of the companies that truly need a distributed solution I will just use WebLogic or any other application server, no Spring needed.
    Again you've missed the point. It's about portability. BEA actually like the fact that you can take a smaller app running on Tomcat and scale it up to an enterprise-class app with robust distributed transaction management on WebLogic without rewriting the code. That way you get choice, so you benefit as consumer too. Spring is as valid in either of those scenarios. Its simplification is just as relevant in a high-end application server as in a web container.
    Anywhere you read about Spring is it hailed as a replacement for EJBs. I am sure Application Server “big boys” do not like that
    It's a gross oversimplification to say that Spring is purely a replacement for EJB. And surely we should let BEA's actions speak for themselves; they are rather more meaningful than your speculation about what "Application Server big boys" may or may not like. Clearly BEA do not feel threated by Spring, and see a value for their customers in using Spring.
    It is like wolf teaming up with sheep. BEA is a software vendor as heart. I believe, and I could be wrong on this one, but in my opinion commercial software conflicts with Open Source.
    This is an extreme view which I've never agreed with. I believe that there are two kinds of software: good and bad software. There are both types of both commercial and open source software. And open source has delivered excellent results in many areas, but not so well in others.

    While I think the spread of open source is a good thing (and have done my bit towards it!), I don't see anything morally wrong with selling software. (Please note that I've been consistent in my views on this for years, so the current deal has not changed them.) Many companies work with both open and closed source: look at IBM's positive contributions to open source, in addition to their commercial product portfolio.

    There is no such thing as free enterprise software (as in beer). Writing high quality enterprise software costs money; it cannot be done purely by enthusiasts after hours. Maintaining and enhancing the software costs even more money. TCO does not purely involve license fees, but cost of training, performance on given hardware, support--in short, license fees are just one part of TCO.

    Btw when you refer to "commercial software" I believe you mean "closed source software". Open source may well be--and, in the enterprise space, inevitable is--"commercial".

    Rod Johnson
    Interface21 - Spring from the Source