Opinion: Can BEA Outrace IBM?


News: Opinion: Can BEA Outrace IBM?

  1. Opinion: Can BEA Outrace IBM? (49 messages)

    Will BEA vs IBM be a repeat of Netscape vs Microsoft? Fortune magazine has written a piece about BEA's Alfred Chuang, and the race for BEA. It discusses the past, how BEA got where they are now, and ideas about the future.

    You are walked through how the company grew, and how it battles against companies that are a LOT larger (e.g. IBM).

    Is BEA setup to be acquired? Or does having HP's weight behind them enough.

    Ray Lane, an ex-president of Oracle, says that BEA faces a tremendous threat in IBM and perhaps a greater problem internally tackling a new market. With only a handful of exceptions such as Microsoft, software companies historically have run into trouble when they've had to expand beyond their core product line. It happened to Lotus when it tried to move beyond 1-2-3, to Novell when it expanded beyond networking into productivity suites, and to Netscape when it attempted to become more than just a browser maker.

    What will happen to BEA?

    Read the Fortune article: Can BEA Outrace IBM?

    Threaded Messages (49)

  2. <quote>
    Chuang is confident that he'll come out on top: "If we get even one-tenth the market opportunity that I think is here, we'll be the biggest company in the history of mankind."

    Any idea what market opportunity Chuang is referring to here? Sounds like hyperbole at best.
  3. Dark horses[ Go to top ]

    The real question is

    "Can BEA keep up with JBOSS"
  4. Let's see Jboss 4.x perf and then we can debate
  5. Let's see Jboss 4.x perf and then we can debate

    That's only part of the picture - the metric that really matters is cost per user (or similar). With a cost per server license of $10k plus for the competition, it's not hard for JBoss to win this battle.
  6. Whatever that title means I think I should clarify my point.

    To be honest at this stage whether JBoss outperforms BEA or underperforms is irrelevant to a large extent. I like what I am seeing in JBoss development with respect to going beyond J2EE and innovating in container design.

    I think JBoss may have the upperhand in terms of container features for the first time. How Open Source development can actually compete with proprietary vendors never ceases to amaze me.

    In my eyes the first casualty of JBoss are the second tier players. JBoss has effectively taken them out of the market. Microsoft is the natural competitor of Open Source and I suspect BEA may be leapfrogged altogether.

    I could be completely wrong.

    As far as the support goes, I talked to a local JUG member who claimed to be a heavy user of JBoss support and who claimed being directly supported by "the guys" was a unique experience for him. The speed was what impressed him. Maybe haven't tried it myself.

    It however points to one of the fallacies of this thread, that JBOSS needs someone large backer to gain credibility. Maybe. It seems to me that while BEA was sleeping on the .com wave, JBoss quietly established strong beach-heads in corporate america and we are all waking up to it with denial and astonishment.

    I wonder
  7. I think you got part of it right. Products like JBoss, Jetty, Jonas, Tomcat and even commercial "open" source products like Resin are making a very compelling alternative to the traditional "low cost model" provided by the previous commodity-based platform ("wintel"). Throughout the '90s, Windows went from running 2% of business processes, to running an amazing percentage (90%?) -- basically saturating every space except "enterprise" and "mission critical", and even getting a few wins there. Now you're seeing those Windows applications being run on the "open source platform" (i.e. Linux, Apache, Perl, PHP, Java/J2EE, MySQL etc.), which is (or at least appears to be) even lower cost and more commodity-oriented than the "wintel" platform, because any piece will run on almost any hardware/OS platform, and open source can be used for any piece of the infrastructure software.

    I know that BEA considers JBoss to be a threat to their core server market, but in the space that they play, I think IBM is a much more relevant competitor. I think Microsoft has a lot more to fear from JBoss, and Marc Fleury's comments seem to show that he grasps this reality pretty well too.


    Cameron Purdy
    Tangosol, Inc.
    Coherence: Easily share live data across a cluster!
  8. More Real Qns are[ Go to top ]

    More Real Qns are
    1. "Can BEA keep up with their current customers?"
    2. "Can BEA stop re-inventing wheels of J2EE Programming?"
    3. "Can BEA learn more about quality and customer's requirements?"
    4. "Can BEA stop comparing with companies such as IBM and Oracle?"
    5. "Can BEA try to understand JAVA is made for enterprise computing instead of Home PC?"
    6. "Can BEA stop delivering preparatory garbage into J2EE Development frame works?"
    7. "Can BEA learn the importance of J2EE engineers in project instead of HTML designers to complete the project with quality and performance?"
    9. "Can BEA stop making fool out of their business?"
    10."Can BEA educate their industry leadership and sales department about the strategies for learning the customer and their business?"
    11."Can BEA stop acquiring others firms and try to establish an in house JAVA scientist pools before trying to open their dirty mouth to be loud about being the leaders of enterprise computing?"
    12."Can BEA stop selling framework based ideas to the j2ee community and tell them to follow BEA's way?"
    Finally, "Can serverside.com stop comparing BEA with other firms and the deginity of their business??????
  9. Opinion: Can BEA Outrace IBM?[ Go to top ]

    Just wanted to point out that what's interesting here isn't the players that we are all quite familiar with, but of the time and place that marks the position of BEA, which to me seems to mark a turning point. The article also touches on some of the ethical questions in BEA's strategy of overhauling much of the developers which had helped to build the company. As I understand it, my opinion is that the best companies are run and maintained by an in grown staff.
  10. Opinion: Can BEA Outrace IBM?[ Go to top ]

    IBM and Oracle have just as much to fear with JBoss, I think the bigger question is how many installations are replacing any of these app servers with Jboss?

    I’ve reviewed both BEA and IBM platforms recently and the bottom line came down to vendor execution. You can look into all the costs, specifications, and wonders of the world these packages offer but it doesn’t mean squat unless the idea can actually get implemented. IBM is absolutely terrible at presenting solutions leading to what make me think they can actually do it? Microsoft is great at presenting solutions but they paint a picture so rosy you have to truly believe it can’t be true. .Net will exist regardless, no need in beating on that platform.

    That leaves BEA, IMO their plain attitude is the factor that propels them forward to getting the job done. The 8.1 product is very nice overall showing some positive vision over their competitors. The more Workshop starts looking like a Microsoft/Macromedia tool the better, development costs are just too high. This whole thing of BEA vs. IBM, JBoss is a waste of time, it’s just like Netscape vs. IE. If we don’t have a strong neutral vendor to support J2EE then what do we have at the end of the day, gee IBM and MSFT again which mean poor and proprietary worlds. I associate BEA to the old days of Apple and Netscape, innovation at its best while serving a better cause rather than total greed. I agree that new blood is good, people tend to get attached their work and refuse to change for the better.

    It’s a huge market out their, IBM is raising the price of their software and I’m sure others will follow. Regardless of the app server costs, vendor execution and support will reduce overall risk which will equate to a lower TCO long term. I hope BEA succeeds as a independent software company down the road, an acquisition in my mind would be the fatal blow.
  11. Opinion: Can BEA Outrace IBM?[ Go to top ]

    It's possible but not likely that BEA could be a buyout by HP, Sun or Oracle. Right now HP has the closest ties but neither Sun or Oracle have been able to establish themselves on the same level in software infrastructure as BEA or IBM.

    With respect to JBoss. I think those proposing it will have a significant dent in the market are completely off base. I happen to like JBoss and even proposed it's use at my last employer. But the IT political situation is such that it's extremely difficult if not impossible to get upper management to sign off on it without some big corporate support. Linux and Apache have had great success because IBM and others have decided to integrate it into their product line. This is not the case so far with JBoss and I don't think it will happen here in the US, though possibly in Europe.

    Anyway I think BEA can survive and even prosper since they have a proven record of success while providing the latest technology.

  12. Read this.....[ Go to top ]


    Give this article a read, you will find it very interesting.

    The landscape for Open Source is changing by the day. Look at the two areas they suggest will be owned by open source, Operating Systems and Middleware.

    This is just an interesting perspective and they have interviewed many CIO from big companies who have started to leverage open source in a big way.

    For all people facing a manager who is against open source, have them read this article. Open source is your competitive weapon against your competition if you are the IT organization of a large corporation.
  13. Read this.....[ Go to top ]


    The issue really isn't open source or not, it's how much of a risk the company is willing to take. There was a time not too long ago when Linux was considered risky because there were 15 or more vendors with virtually no support. In the US this changed once Red Hat was able to get licensing agreements with Fortune 500 companies (notably IBM). IBM and Red Hat can fine-tune Linux for IBM hardware and provide world wide support. JBoss needs to get a partner, the bigger the better to establish it beyond the sourceforge downloads. Also realize that BEA is offering free developer licenses for the complete platform. That's the entire suite of products.
  14. Read this.....[ Go to top ]

    Hi Ben,

    For all people facing a manager who is against open source, have them read this article. Open source is your competitive weapon against your competition if you are the IT organization of a large corporation.

    You forgot your signature:

    Ben Sabrin
    Director of Sales and Business Development
    JBoss Group, LLC
    404-467-8555 - office
    404-664-9466 - cell
    404-948-1496 - fax
    ben at jboss dot org

    It appears to be relevant to your post.


    Cameron Purdy
    Tangosol, Inc.
    Coherence: Easily share live data across a cluster!
  15. Cameron is da' man[ Go to top ]

    Thanks for pointing out Ben's true colors as a director of
    jboss :-)
  16. ok get them, keep on fighting[ Go to top ]

    To Ben,

    Only 14925 left to go!

    (BEA:15000 paying customers with support contracts vs. JBOSS:75)

    That's shows good spirits!

    Rolf Tollerud
  17. Makes sense[ Go to top ]


       Your comments make a lot of sense, but where I think it fails is not in what you say, but what you don't say. The same statement applies to Alfred's interview.
       All the other major vendors of enterprise Java solutions like IBM, ORACLE, SUN, and HP have one major business initiative in common: They view the EJB Container as a commodity. This is obvious given the state of the market for EJB Containers (ie. JBoss is free, Gartner Group all but sealed the coffin on the possibility of making a lot of money just by having a container alone).
       It is also obvious that the 4 companies view most aspects of the engineering/programming talent required to build this software as a commodity as well. Evidence that this is so can be found in these 4 companies willingness to outsource major portions of product development to other countries (outsourcing is not an easy decision to make, the economic benefit of reduced cost must overshadow many other factors). Take a look at HP, they're so convinced that making and promoting an EJB Container is such a waste of time that they gave up on it altogether. Many people consider that decision to be a very good one.
       The point is that these 4 companies are focusing on the "pilot fish" around the shark that is the EJB Container. Services is a HUGE factor in the selling of this product now. Buying an EJB Container means that you plan on doing some heavy integration between disparate software/hardware components throughout an enterprise.
       I didn't hear one word from Alfred about how BEA's "pilot fish" are doing. Don't get me wrong --- I'm a big fan of WebLogic, and I've worked with people from the firm and for the most part I've learned a lot from them. I don't doubt that there is some great talent over there at BEA.
       But how is BEA adapting to the change in the market that they staked their life on? From what I've seen, they continue to focus on "dot.com-day" issues such as "First-To-Market" and so forth. These issues are becoming less and less relevant because buyers of this kind of product are now looking for a long term solution when they struggle to make a decision on which of these vendors to work with.
       Although I admire the technology very much, I would love to hear some insight into how BEA can match the offers from IBM, HP, SUN, and ORACLE when it comes to the "pilot fish" benefits.

  18. Makes sense[ Go to top ]

    Scott, totally agree with your viewpoint except I would say that HP and Sun were just plain idiots at marketing their app servers rather than building/buying one.

    BEA does have the advantage in most cases with “First to Market” due to the limited number of products their selling. The EJB container may be a commodity but everyone of these products could use some help beyond the spec with add-on services (eg. Caching, deployment, management). BEA can still make a difference just like any other company. Microsoft does it with every product release.

    IBM will always be the last one to the party due to the dozens and dozen of products in the “Websphere” family that must be aligned to market.

    JBoss has the least amount baggage to deal with promoting a nimble approach but support is a glooming issue. If some large company took on that support like IBM did for Linux it would make them much more viable to promote.

    That outsourcing issue is huge and getting more difficult each day.

    I can’t wait until this app server discussion is over, then we can move on to replacing Oracle with an open source product and really save some $$$. That should keep us consumed for a few years.)
  19. Makes sense[ Go to top ]

    I understand that at their recent Florida convention BEA were exhibiting a set of new tools and integration features with the message that they are aiming to be the 'VB' of the J2EE world. How realistic is that?

    In a previous message somebody compared BEA to Netscape in the earlier browser wars. Personally I think they'll end up going the same way as PowerBuilder.

    As for JBoss I can only assume that BEA is grateful that JBoss is run by a collection of techno-heads. The lack of basic JBoss introductory documentation for tutorials, installation and customization must be killing its real take up rate.
  20. Aiming for VB?[ Go to top ]

    Davout: I understand that at their recent Florida convention BEA were exhibiting a set of new tools and integration features with the message that they are aiming to be the 'VB' of the J2EE world. How realistic is that?

    I don't know, but apparently Sun decided to announce the same thing in the past week. It was on the cover of eWeek or SDTimes or something.


    Cameron Purdy
    Tangosol, Inc.
    Coherence: Easily share live data across a cluster!
  21. Aiming for VB?[ Go to top ]

    I would recommend the folks here to download WLPlatform 8.1 (incl the workshop). Tools will make the difference in the J2EE platform landscape.

    Sat Nashikkar

    "import std.disclaimer"
  22. Aiming for VB? Why[ Go to top ]

    VB market became saturated and eventually fell down. J2EE is about the Enterprise, there are many parts of the Enterprise you can simplify with tools but the problem spots such as integration, tools fall short.

    IBM is still getting a VAJ wrap when Eclipse is over 2 years old and WebSphere Studio is in its second generation. WebSphere Studio Integration Edition is also in it's second generation.

    I can import a COBOL copybook (true integration problem CICS, IMS), generate a WSDL doc and JCA connector and Business Delegate all from WSADIE. I can also drag the JCA connector and drop it on to a Micro flow and have an EJB or Java Object call it. I can also wrap in into a Web Service (HTTP or JMS) and have it part of the diagram. All Drag and drop like VB. But you still need to know what you are doing with the Integration. This is second generation also

    With WSAD I can also drag and drop Struts Actions, Action Forms and JSP on to a web diagram and create a Struts Interaction.

    you can add Portal plugins to pull in Portlets

    We can do Workflow and Microflows both Async and Sync. With people and automatic interaction from an Eclipse Open Tooling platform. Also with Eclipse I can add my own type of node to connect to.

    I think people say they've tried the IBM tools (because they used VAJ 3 years ago) and bash it. BEA didn't have tools 2 years ago. The same people that say they hate IBM tools say the love Eclipse (huh???). IBM created Eclipse and all their tools are now plugins to Eclipse.

    I am not saying BEA doesn't have comparable features in their workshop 8.1 because I haven't used it, I tried the 7 version and it seemed limited to me. But when you need to solve complex problems, your tools need to have hooks for CICS systems, IMS, etc... IBM is way ahead, I can enable a CICS transaction on the Mainframe as a web service with drag and drop. BEA only really has connections into Tuxedo and don't know if they can drag and drop JCA connectors on to their workshop.
    How can I extend BEA's workshop to drag and drop my own JCA connectors. I have to manually wrap code myself with a Web Service to make it part of the flow.

    You can down load
    WebSphere Studio Application Developer Integration Edition Here with a 90 day trial:

  23. Makes sense?[ Go to top ]

    As for JBoss I can only assume that BEA is grateful that JBoss is run by a collection of techno-heads. The lack of basic JBoss introductory documentation for tutorials, installation and customization must be killing its real take up rate.

    I've found that to be the case with MOST application servers - the introductory information and demos are difficult to wade through without the help of someone who's been through those same problems before. The biggest difference is that in most shops, there's noone who is "the guy" when it comes to JBoss.

    I've been through BEA's documentation up, down, left, and right on several occasions looking for critical configuration and customization info, to no avail. This is a common problem that has much to do with the complexity of app servers and the difficulty in predicting exactly what people will want to do with them.
  24. Makes sense?[ Go to top ]

    I've been through BEA's documentation up, down, left, and right on several occasions looking for critical configuration and customization info, to no avail. This is a common problem that has much to do with the complexity of app servers and the difficulty in predicting exactly what people will want to do with them. <

    Which is why I'd suggest their dream of making BEA the "VB" of the J2EE market is such a joke
  25. Makes sense?[ Go to top ]

    I've been through BEA's documentation up, down, left, and right on several occasions looking for critical configuration and customization info, to no avail. This is a common problem that has much to do with the complexity of app servers and the difficulty in predicting exactly what people will want to do with them. <

    > Which is why I'd suggest their dream of making BEA the "VB" of the J2EE market is such a joke <

    Sounds a bit more like the "ASP" of the J2EE market...
  26. Opinion: Can BEA Outrace IBM?[ Go to top ]

    Hmm ... a surprisingly readable article. Great link!


    Cameron Purdy
    Tangosol, Inc.
    Coherence: Easily share live data across a cluster!
  27. CEO - Code Executive Officer[ Go to top ]

    I like the part where the article claims that Alfred contributes a piece of code that he himself created to test the WebLogic server. Regardless of how simple that code may be, that really says a lot about this man.

  28. Alfred tests the product[ Go to top ]

    last year it was Chuang's code that was used as one of the final tests for the company's software...

    The only way this makes sense...
    Alfred actually played with WebLogic Workshop to test it. The tool is supposed to make programming so much simpler, they used him as a non-expert to verify the claim.
  29. BEA Employee Comment[ Go to top ]


    I've been following this thread closely with a lot of interest. I'd like to provide to the community some additional information and clarity on our positioning.

    Yes, we are actively marketing Workshop 8.1 as a productivity framework that can make any developer immediately productive on any IT project. As long as developers have experience in a structured programming language, they should be able to do the bulk (not all) of major enterprise development tasks. This "bulk" includes sophisticated web applications, workflows, portals, legacy system integration. We claim this because Workshop does this work requiring non-OO skills (ie, variables, constructors and methods). As long as a developer has a "control" (please try product for more clarity) to a particular resource, a developer will be able to apply their existing skill set to successfully develop the web apps, workflows and portals.

    Workshop 8.1 is a productivity framework within the platform, but developers are not required to use it. Platform 8.1 is still an open system, and when you peel back the covers, you'll still see the engine layers driving as many open standards as we can. There is a lot more market variability in the integration / portal standards area, but we are implementing the standards that we can. Developers and companies that want to develop directly to this layer can still do that, and we document our support carefully. In the case of using Workshop, it's a productivity layer that creates applications that leverage the open standards. For example, if you build a web application in Workshop, we generate a pure Struts app that is in perfect synch with the workshop app. If you build a Web application that accesses a DB control, then we bild a struts app using JDBC that is in perfect sync with Workshop. This means that developers can develop directly to these applications at the layer they choose without stepping on each other's toes (ie, the wizard / generation problem).

    My summarizing point: BEA is working hard to make our technology consumable by a wider audience of developers that are doing web application, integration and portal work. By creating technology that can be used equally by J2EE experts and non-OO application developers, corporations as a whole become more productive and have a better ROI.

    Director, BEA.
  30. Reply BEA Employee Comment[ Go to top ]

    And then I will get hired to come in and rewrite all those apps like I do now for silverstream pages.
  31. BEA Employee Comment[ Go to top ]


    I'm a Borland customer since the Turbo Pascal days. Borland has a long history of providing developers with extremely productive IDEs, thus they earned my vote ($) for each of my ten years of software development.

    OK, times change, BEA provides a great J2EE application server and I'm impressed. So, now BEA wants me to vote for their J2EE IDE called Workshop. Last time I looked into Workshop, I found it to be a fancy Java GUI wrapping a proprietary black box. I was in disbelief why a company like BEA could think enterprise-distributed developers would ever consider such a tool. Last time I saw something like this was with Powerbuilder.

    But, I still acknowledge good hard working technologies like Weblogic, so, Tyler give me some reasons why I and others like me should give your Workshop another look.
  32. BEA Employee Comment[ Go to top ]

    George, let me attempt to provide some answers to your questions.

    BEA believes that Borland has some great products and we maintain a very tight alignment with them. This is because neither company views Workshop / JBuilder / Togethersoft as competitors. Let me try to clarify the distinguishing points:

    Workshop: This is a runtime productivity layer with an IDE that BEA built focused on developing enterprise apps. This means that the development impetus is on workflows, XML, portals, portlets, content management, personalization and web apps. Think of Workshop as a productivity layer (runtime + IDE) designed to make development with all WebLogic products productive for a wide class of developers. Workshop structures a programming model that eliminates the OO intensive activities so that enterprise development can be done without it. You can leverage advanced OO concepts through Workshop, but it's not necessary to get the job done. Everything that Workshop produces is a file that can be simultaneously edited in JBuilder, if that is your preferred development environment (in fact, the runtime will automatically pick up any external changes to any files).

    JBuilder / TogetherSoft: Enterprise development environments focused on sophisticated Java / J2EE development. Their integration with WebLogic Server is good, but they do not have tight alignment with all enterprise development activities. These full-featured environments incorporate sophisticated J2EE programming tools, full life-cycle J2EE development including OO design, UML support, etc. They also contain libraries and technologies for doing Swing / AWT / J2ME development. Essentially, these environments are all-purpose heavy-duty development environments for Java that have extensions for working with WebLogic Server ONLY. For customers that are adopting struts on WebLogic, WebLogic clustering, WebLogic Portal, WebLogic Integration, etc. this environment provides development capabilities to the extend that standard Java allows. This isn't a drawback of JBuilder / TogetherSoft, just not their intended focus area. Workshop is an environment that provides uniform development across all of these key areas that are relevant to today's enterprise.

  33. Tyler,

    "Think of Workshop as a productivity layer (runtime + IDE) designed to make development with all WebLogic products productive for a wide class of developers. Workshop structures a programming model that eliminates the OO intensive activities so that enterprise development can be done without it. "

    The question is, why should one use those stuffs without knowing the "real" background? IMO, in the long term, OO concepts should be familiar to all developers, no excuse for this. If I had a developer, who did not know about OO concepts, I won't let him work with BEA Workshop directly. I would send him to OO courses first, so he could understand the concepts. After, I would send him to J2EE courses, so he would know the "real background" of that technology.

    I agree that we need tools for more productivity, but it is important, that the developers understand the real technological background first, so we don't end up like "point, click and forget". I saw a lot of programmers just doing point&click, drag&drop and if something "different" happens, they just don't know how to solve, because lacking of the technological background.
    -> Knowledge is the power, and not the tools ;-)

    IMO, BEA Workshop is just another IDE specialized in enterprise development and BEA technologies. Actually, it would be nicer to have BEA Workshop just as plug-ins for other IDE available such as NetBeans and Eclipse (or is this already possible?). No need to write your own BEA IDE. That's why we call this "Integrated development environment", right? ;-)

    Lofi Dewanto.

    OpenUSS - Open University Support System
  34. Lofi-

    Everybody at BEA sees your points and understands the perspective that you take. However, those views are not uniformly shared in the industry. Let me clarify.

    We did exhaustive research with our customers, the market and medium and large businesses that have network operations centers or IT development organizations. Many of these organizations have developers with skill sets represented by TSS community. TSS members are savvy developers, want to have more control and ownership over the technology, take more pride in the overall architecture of an application, etc. We estimate somewhere between 500,000 - 1,000,000 developers around the world that fit this profile. These individuals like programming directly to J2EE, want complete control over the systems they use and have a tendency to require choice in their development environment / tools.

    However, what we also found was that many companies have a class of developers that want to get into enterprise development or Java, but have not either because of the perceived difficulty of J2EE, lack of skills, lack of time to invest in research, etc. This market is in the millions of developers -- and this type of developer traditionally not a TSS member. Companies told us (repeatedly) that they want the reliability and stability of a strong J2EE platform, but they needed a simpler programming model for their applicatioins developers that do not have a lot of time to invest in understanding the technology as well as everyone on this site does. These developers do not have the same development environment choice requirements that TSS members have. These developers also appreciate what OO does, but prefer an environment that allows them to focus on business logic rather than the plumbing.

    Platform 8.1 was carefully designed to appeal to both audiences. We have made improvements at the engine / open systems layer to appeal to the savvier J2EE / systems developers. But, we've also added Workshop to provide a simpler programming model that gives the broader audience of developers a simplified model for working within a J2EE environment without having to do OO development, unless they want to incorporate it. You do not have to leverage the Workshop runtime to work with Platform 8.1, so the fact that we've built a simplification layer shouldn't shun the TSS community; it's a completely open system that you can leverage at any level that you desire.

    Also, to your point about "why not plug into Eclipse". When designing Workshop, we strongly considered this, but there were two distinct reasons that we couldn't leverage these open source platforms for the base IDE. First, eclipse is so open, there is no consistency in the exposed programming model for plug-ins. What do I mean? Well, in order to make different pieces within Workshop to work together and to work for applications developers, we had to create a programmable abstraction layer that "bridged" the different technologies. This turned out to be controls and XMLBeans, which are both models that access very sophisticated resources using non-OO java. Basically, a control simplifies any resource down to a set of asynchronous method calls that can easily be invoked. This provides a clean model for working between subsystems in the runtime environment. Eclipse and netbeans do not standardize on a programming model such as this allowing each plug-in developer to create a different model that would create inconsistencies for developers, ultimately not achieving our goal of creating a simpler environment for all developers.

    Second, eclipse wasn't developed in pure Swing. And for reasons that are important for portability, we needed the GUI portion of the Workshop framework to be in Swing.

  35. Because IBM is doing it . . in a big way, and getting much bigger.

    Tyler's 3rd paragraph in his reply (msg 76956) is right on - that's reality.

    IBM is "moving beyond" J2EE in seeking productivity gains - check these EXTRACTS from recent articles and see where they're heading. Article URLs included (hope these aren't too long..) :

    The second initiative -- the rapid application development (RAD) features for WebSphere Studio -- will shorten time-to-value for WebSphere applications by providing a simple to use, visual environment for Web development with minimal coding.

    &#8221;It is our intention to make WebSphere as easy to develop for as Domino is."

     &#8220;So we are designing the right tools to allow Web developers to build apps on J2EE without having to learn Java."

    TITLE: IBM expands J2EE application development tools and increases Domino and WebSphere integration


    By the first half of this year IBM will aggressively extend the Java Enterprise Edition 2.0-based architecture of WebSphere to a services-based architecture that allows developers to build applications with integrated workflow, business rules, and network-based transaction capabilities

    &#8220;In the on-demand applications world, J2EE alone is like having a heart without the lungs. Putting a Web services veneer on top of J2EE is just not sufficient, according to Scott Hebner, IBM&#8217;s director of marketing for WebSphere, in Somers, N.Y.

    Along with the integrated workflow and business rules capabilities, IBM will also deliver a more complete Web services stack in the upcoming version of WebSphere 5.X, Hebner said. The improved Web services stack will include the Web Services Invocation Framework that allows for network-independent interaction among Web services, a precursor technology to a business process execution language, and the ability to have compensation patterns that allow users to back out of any given Web services workflow.

    Explaining the advantages the next generation of application servers can bring, Hebner said developers will be able to look at a palette of business services whether they are written in Cobol, EJB, or Windows and take them to compose and choreograph them into entirely new applications.

    TITLE: IBM eyes services-based role for J2EE


    IBM Corp. says the Java 2 Enterprise Edition (J2EE) is not enough to support its on-demand computing strategy, so the company's looking to implement a services-oriented architecture through its WebSphere application server.

    This "requires a next-generation application server that takes the J2EE kernel and makes more of a services-oriented architecture out of it," Hebner said. "The J2EE architecture is insufficient for this type of application and integration."
    A new class of application server usage is required, Hebner said. "It's much more than J2EE," he said.

    IBM will deliver the advanced features&#8212;integrated workflow, business rules and provision and audit capabilities&#8212;in WebSphere 5 this year, Hebner said.

    In addition, Hebner said, the move toward a service-oriented architecture would help developers.
    Today developers "are writing a lot of customized J2EE code, and with the service-oriented architecture it'll allow developers to take a much more visual approach to implementing business services instead of coding everything line by line," Hebner said. "The development productivity increase is huge."

    TITLE: IBM Looks Beyond J2EE

    "What currently excites me most is the change in computing as we know it."

    "However, we clearly understand that today's standards cannot enable all the application design patterns that are required by today's flexible enterprise architecture. IBM continues to add extensions to the programming model that provide additional application enablers needed to achieve the goals of service-oriented architecture."

    TITLE: Interview with Danny Sabbah, VP Development for WebSphere Software Platform


    Already in place to a degree, the Service Integration Bus paradigm anticipated by IBM will understand multiple protocols, in addition to SOAP and Web services.

    "We want to go into services-oriented architectures across the enterprise, and that goes way beyond [protocols such as SOAP]," said Andre Tost, solutions architect in the IBM WebSphere Business Development organization, during a presentation at the XML Web Services One conference here on Wednesday.

    The Service Integration Bus would integrate applications such as portals, b-to-b interactions, Web services, and existing applications and feature a common run-time environment. JCA (Java Connector Architecture) connectors would talk to mainframes and systems such as SAP implementations, Tost said.

    TITLE: IBM eyes services provisioning role for WebSphere

    "The rapid application development (RAD) features for Web applications that Lotus and the WebSphere Studio team are building is going to open up the ability to build J2EE applications to a whole new set of developers. The features will offer a much more visual environment, which hasn't been available in the J2EE world to this extent -- the kind of integrated development environment that has been there for the Domino developer.&#8221;

    TITLE: The win-win-win world of next generation software


    And there's other new "stuff" that fits into the picture too, such as their Information Integrator series.

    Best to all - GH
  36. a way forward?[ Go to top ]

    Considering that they have not learned to use J2EE correctly, still talking about of how fantastically xDoclet will generate the EJB interface classes, service locators and deployment descriptors etc..

    As the J2EE developers persist with their EJBeans maybe such tools could coerce or at least entice them in the right direction?

    just a though..

    Rolf Tollerud
  37. |
    |the rapid application development (RAD) features for WebSphere Studio

    Rapid AD? You could write a small application in the time it takes for WSAD5 to start up or do stuff (on a dual-proc 0.5Gig RAM workstation, no less).


  38. Opinion: Can BEA Outrace IBM?[ Go to top ]

    Makes a point..

    "IBM was supposed to put Microsoft out of business with its OS2 operating software. It was supposed to put Oracle out of business with its database software. And it was supposed to put Sun out of business with its mainframes and servers"
  39. Opinion: Can BEA Outrace IBM?[ Go to top ]

    Professional Sales team of IBM already put un-professional, loud speaking BEA sales team out of corporate managers office by providing customized business packages deal including( server + database + app server + professional service) for a reasonable price. Result is, a lot of people lost money in BEA's share and BEA desperately looking for a way to come out it.
  40. BEA Sales Team[ Go to top ]

    T Q,

    If you worked at BEA you would be aware that it's a goal to NOT have a direct sales force. The combination of a great core platform and increasingly strong and usable tools lead to software that sells itself. BEA is not only going after MSFT, they are doing it with former MSFT'ers using well-crafted MSFT-strategies, like dev2dev.

    You should strive to get over your being terminated from BEA as it was probably for the best. If they are as bad as you claim, you should be grateful to not have to work there anymore, right? Move on.
  41. Envy BEA[ Go to top ]

    Although BEA is the underdog in this fast-growing race to achieve market share, their tools and solutions make sense.

    Since IBM and others also offer tools and solutions that make sense, this may turn out to be a war of attrition.

    TQ, your generalizations on BEA are meaningless. You have not completed a survey of all customers serviced by BEAs sales team, so you are not qualified to say anything about their tactics. Your comments are the same as if a salesperson were to make judgments on the development team's code --- meaningless.

    I know you have hard feelings about being fired from BEA, I certainly hope that you can overcome your negative feelings and as the previous poster rightly stated, move on with your life.

  42. Envy BEA[ Go to top ]

    Who got fired from BEA?. I know BEA better than you. I know BEA Sales people better than you. I used to go with them to explain enterprise-computing ideas to BEA customers. I have participated for bench marking of weblogic app server for clients with other vendors such as IBM at client place. I know how IBM sales person presents the business and BEA sales team present the idea. Dude I am happy about what I am doing, proferssional vendors such as Oracle and IBM is also happy with their business. They don't show up in serverside.com in 24/7 to market their products.
       By the way, why don't you call yourself as Bill gates? It is a better name than the fake name u have right now.

       Dude, do some JAVA coding instead of making nonsense with my views or try to figure out how to create a web application using BEA Workshop 8.1 (It is funny, my application server start program is still hanging in a 1.7GHZ p-4 processor with 512 Mb ram machine). I know guys like you not going to change, cuz you don't even have a proper name.
  43. Opinion: Can BEA Outrace IBM?[ Go to top ]

    In his interview with TheServerSide, Doug Purdy from MSFT makes a good comment stating that the company with the best tools will win the race. Since both WebLogic and WAS are built off the J2EE standard (*relatively* similar set of core services), I think this statement is especially applicable.

    I think IBM WSAD5 and WAS5 are starting to come up to the ease-of-use standard of MS Visual Studio /.NET which really sets it apart as a great tool to use to build powerful J2EE Apps.
  44. Opinion: Can BEA Outrace IBM?[ Go to top ]

    BEA has been able to out maneuver IBM in the past by getting products to market much earlier. WebLogic 6.1 had J2EE 1.3 compliance well before WebSphere 5.0 was close to GA. Also BEA has the ability to focus it's entire corporate effort into software infrastructure products while IBM has thousands of products that require attention. The war will be won by the company that can move fast as new specs come out, get the best people to develop their products and build deep relationships with consulting firms and vendors. BEA has shown that they can do this so far. But any slip could prove to be fatal.
  45. Opinion: Can BEA Outrace IBM?[ Go to top ]

    "The war will be won by the company that can move fast as new specs come out,"

    No it won't. It will be won by the company that provides the best tools for quickly building robust enterprise software. The spec issue is a red herring.
  46. Think about this[ Go to top ]

    Just think about this one for a second:
    Everyone uses M$crosoft, does that mean that it is the best product?

    I have worked for a while with Weblogic, but my latest is an ear that has to work with both WebSphere as WebLogic. That's even harder. The only problem I have with migrating from WLS to WAS is the documentation. Bea support is great and i guess ibm thinks that everyone is a thinktank :) But I you are in the possibility of only working with WAS 5 and the WAS Developer Studio 5 and have a lot of RAM, then you are in luck in my opinion. The developer studio is great product. Now offcourse there are allways people saying, free stuff is the best, just give us opensource. Well i wonder if the software made by these people, is free for their client, i guess the end consumer would also like to benefit of the "free-sw-minds" ??? The developer studio and WAS cost a lot of money, but the end product is something you can rely one. Free is cool, but we're not living in utopia are we?

  47. about vendor lock in[ Go to top ]

    but my latest is an ear that has to work with both WebSphere as WebLogic. That's even harder. The only problem I have with migrating from WLS to WAS is the documentation

    Just that it is possible do not mean that it is easy. How easy/difficult it is to migrate from one J2EE app Server to another depends of course on how well you know both products.

    IMO when you talk about "portability" between app servers you have to take into account that you first have to learn the "target" environment. Most people who argue that "it is easy" to move from one to another already know both products intimately..

    I should say that the matter of porting a complicated enterprise app developed in one app server to another platform that you do not know well is a non-trivial task - to say the least.

    And before you kill me..yes, correct - I haven't tried..

    Rolf Tollerud
  48. Focus on Niche[ Go to top ]

    I believe, BEA can establish a niche for itself and should focus on the niche market and not try to be everything to everybody in the industry, since that requires a lot of resources at disposal, which is something BEA would always struggle against IBM
  49. Tools, and portability[ Go to top ]


    I'm seeing two issues being talked about here, which I can roughly summarise as follows:

    - "How do I support more than one platform (e.g. WLS and WAS)?"
    - "We need good (i.e. graphical) tools to improve productivity and make it easy to develop and deploy J2EE apps. The vendor that supplies the best developer tools will win."

    Curiously, the response to both issues is the same -- xDoclet (Open Source, of course).

    What's the issue with supporting multiple servers, anyway? The core business logic of your application itself doesn't change. It's just the deployment descriptors that are different for different app servers. xDoclet will generate them for you. If you include xDoclet tags for all the app servers you plan to support, xDoclet will generate them all and put them into your EAR file for you, and you can then happily deploy the same EAR to any server you want, -- JBoss, WLS, WAS or anything else.

    xDoclet also integrates nicely with Ant, which is *the* build/deploy tool you should be using in your development, whether or not you use an IDE. If you're using Ant with xDoclet, you will find little need to tinker with the vendor-specific aspects of your application. Just concentrate on the business logic, and xDoclet running as an Ant task will generate the EJB interface classes, service locators and deployment descriptors as part of the build/package/deploy process. It's extremely painless and addictive :-).

    You absolutely should not be locking yourself into any vendor's proprietary development methodology, not in this day and age. Just say No to a Studio ;-).

    Use an IDE, by all means. But save your money and go for an Open Source one like Eclipse or NetBeans, and run Ant with xDoclet from within it. It's a free-of-charge, free-from-vendor-lock-in, high-productivity way to build J2EE applications.

    Take a look at xPetstore to see how xDoclet is used.

    Ganesh Prasad
    Web developer/architect
    Sydney, Australia
  50. I think BEA will lost in the "war".[ Go to top ]

    How about Rational?