Discussions

News: Sun's CEO McNealy Suggests Open Source Could Be Hurting J2EE

  1. In a recent interview, Sun Microsystems CEO Scott McNealy questioned the need for a J2EE certified JBoss application server, claiming that we already have Sun One. McNealy went on to suggest that open source models may be screwing up the industry's ability to market J2EE against .NET due to open source undercutting revenues (and marketing budgets).

    Since when is Sun One open source?

    An excerpt of the related questions have been published here:
    http://www.oetrends.com/cgi-bin/page_display.cgi?77.

    Threaded Messages (179)

  2. Needs more background[ Go to top ]

    I feel after reading this that I just get fed a bunch of FUD with no real substance. I certainly hope there is a lot more to this interview, because it feels like good sound bytes were take out and placed into this summary to sell magazines.

    Just my .02.
    Jim
  3. Needs more background[ Go to top ]

    There is no FUD here. Sun needs to adapt to a changed market.
  4. I thins this is a first step towards SUN charging for future services, advance products etc from SUN.
    However good JAVA, J2EE is, SUN is in deep shit as its not making money thro it , but spending a lot to support the whole Java community.
      Besides selling servers, SUN has no means to make money and their stock price is below $4. I know that they have a lot of cash and they wont crumble down, but there is always a chance of SUN starting to find ways to make money thro this java community soon.
      
        
  5. The writing was on the wall for a while... J2EE infrastructure should become a ubiquitous commodity, allowing 3rd-party solutions to proliferate. That's what will make Java/J2EE successful. JBoss is just making it happen, faster; and it's a pretty darn good piece of software. This should also help to circumvent .Net in their Web services-led game. For example, Collaxa's Web Services Orchestration Server (J2EE-level Container) for JBoss.

    Cheers.

    Jill.

    "Neo, this is the sound of inevitability" (The Matrix)
  6. Some would contend that the problem isn't with the application servers; it's with the J2EE specifications. The complexity of J2EE has violated Wiseman's law. Wiseman's law unequivocally states:

    A successful technology will saturate an 80% sampling of programmer/analysts only if 80% of the technology can be understood by those same programmer/analysts without forcing them to work beyond their regular 45 hour work week.

    Look back 10 years to PowerBuilder and Visual Basic. At the peak of Client/Server revolution, PB reigned as king, and VB held nearly as much market share in terms of number of applications rolled out into production. Why? Because a regular individual could learn how to be _productive_ with the tool in an extremely short period of time.

    I love Java programming. However, EJB's are far too comlicated for the mere mortal to learn AND be productive with without a serious investment in time outside of the regular 40 hour work week. How many of us have spent our Friday nights reading the latest spec just to keep up with the times? Even our peers who are publishing their works in O'Reilly bindings are echoing simmilar sentiments regarding the complexity of the specifications.

    While my love for Java has never been stronger, I believe the sun is setting, pun intended. Until our community can figure out how to de-mystify J2EE's complexity so that the mere mortals out there can learn it AND be productive, the step-by-step wizards in .NET that are make writing distributed applications a breeze will continue to gain ground on J2EE.

    At the end of the day, open source isn't hurting J2EE. Complexity is hurting J2EE.

    Jason Weiss
  7. Complexity Hurting J2EE? C'mon![ Go to top ]

    <quote>
    At the end of the day, open source isn't hurting J2EE. Complexity is hurting J2EE.
    </quote>

    Neither is hurting J2EE - J2EE is completely learnable and is truly not that difficult if you just apply yourself ever so slightly (and you don't even have to do it on Friday nights ;-). And of course it offers more solution-bang for the buck than .NET because of the wide range of solutions (including the so-called web services) that development shops can implement on a wide range of hardware/OS platforms.

    The comments have been right-on in this thread - BEA, IBM (and Oracle is making headway) provide products that people want to buy - Sun's product (at least the last time I used it) is horrible. I don't particularly care if its free. JBoss is good and free so I use it. When Sun produces good software I'll use it, too, if the situation warrants. To blame others on one's own bad software reflects poor knowledge of both your product and how the software industry has evolved over the past 5 years or so. The Open Source comments sound like the paranoid meanderings of people like Steve Balmer at Microsoft. The industry shows that if you have a good product, people will buy it - and for chrissake, if you are Sun Microsystems, your name alone will at least prompt others to try it - the fact that it hasn't made money reflects the software that Sun has produced not the evil Open Source products out there.

    Cheers
    Ray
  8. Complexity Hurting J2EE? C'mon![ Go to top ]

    You have to be kidding me --'And of course it offers more solution-bang for the buck than .NET'. Maybe you are talking about the open source solutions only.

    I use both WebLogic and .NET. Sun is yet to make up its mine on JDO, .NET already has a full-feature ADO.NET, which is what JDO is emulating. JAXP was released belated and full of inconsistency (Ref: JavaPro article or your own experience). Remember Sun's J2EE architect's comment about J2EE is using Microsoft Transaction Server as prototype. They frog leap each other. J2EE was ahead COM+ before .NET. In this round, .NET is ahead.

    Have you even looked the .NET framework library from end to end? I have worked with a bunch java developers. Some of them have tried .NET and most of those who have tried like it better. You have to build an app with .NET and build another app with JBuilder and WebLogic to say this, not by reading a bunch of white papers.

    How we dislike Microsoft's business practice is one thing. We have to acknowledge that the company has over the years attracted a lot of top talents from top schools and its competitors and have retained them. No matter how narrow minded some of them are, they are smart, competent and competitive engineers. And their products are embodiment of such talent pools and hugh R & D budget behind them -- for the same reason WebLogic is better than JBoss.

    By the end of the day, how good a product is really how good it actually is times how well we understand it. Let's don't kid ourselves. Microsoft has been playing catch up games, but they are doing a damn good job.

  9. Apache OJB project[ Go to top ]

    K k,

    <quote>
    I use both WebLogic and .NET. Sun is yet to make up its mine on JDO, .NET already has a full-feature ADO.NET, which is what JDO is emulating.
    </quote>

    Despite the names similarity ADO is much closer to JDBC than to JDO. JDO is a transparent persistence layer on top of JDBC that provides object/relational mappings, object query language, cache, etc. Check out Apache's OJB project for JDO implementation.

    Regardless of the above, many IT people do not see Widnows/.Net as a viable deployment platform due to lack of security, proven track record, scalability, choice,
    cost, etc. If you are interested, you'll find more details on this subject in numerous Java/J2EE vs .Net threads on this site.

    -- Igor
  10. Apache OJB project[ Go to top ]

    There is a fundamental difference between ADO and ADO.NET, although the names are confusing. Roughly speaking, ADO.NET, JDO and maybe even EntityBean are in-memory DB if you just look at the value object piece of it. ADO.NET does not support custom methods on data objects, but does maintain relationship amoung multiple recoresets within ADO.NET. The old ADO supports caching of single disconnected recordset long ago, and ADO.NET adds caching of multiple Recordsets and their relationship. And you can query/filter cached data objects and the filtering cascades to child data objects.
  11. Apache OJB project[ Go to top ]

    K k,

    The main idea behind JDO and EB (CMP) is to provide means for *transparent* persistence of java objects. On top of that you’ll get rich query language, cache, optimistic locking, declarative transactions (EB), security (EB), etc.

    Since neither ADO nor ADO.Net can provide you with this type of functionality, I'd suggest that you prepare yourself for looong hours of dealing with boring database access code and shuffling data between your objects and recordsets.

    To spend your time more productively (and quite in the spirit of this thread), check out free, open source Apache's OJB project.

    -- Igor
  12. Apache OJB project[ Go to top ]

    How is querying in JDO any simpler than querying database? As for security, cacheing, etc, MS has COM+ for that. It's the flaws in Entity EJB that motivates JDO. I think many people will agree dealing with EJB is too much a pain compared to dealing with DB. It's transparent alright, but only 20% for J2EE apps use EJB, let alone entity beans. We have more productive things to do, such as spending time with family.
  13. Apache OJB project[ Go to top ]


    It must be wishful thinking on your part.

    It is pathetic to see the Java community repeat this mantra over and over. It is thousands of large Windows/.Net installations already. And there are no more issues with security in WinNT/XP than in Solaris or Linux, WinNT/XP is attacked more that is the whole of it. But as Niezte says "All that does not kill you makes you stronger", so will all this attacks in the end only result in the most robust operating system on the planet.

    In the meantime you can enjoy 8 times faster performance, 4 times less code base, and 2 times more productivity not to speak of the sheer joy of working with the .NET who is happily enjoying the staying away of comical impractical theorists.

    Performance, maintainability of code and productivity. Why do you not try it? A much better mantra…

    Regards
    Rolf Tollerud
  14. More .Net misinformation[ Go to top ]

    In the meantime you can enjoy 8 times faster

    >>performance, 4 times less code base, and 2
    >>times more productivity not to speak of
    >>the sheer joy of working with the .NET
    >>who is happily enjoying the staying away
    >>of comical impractical theorists.


    All of the above propaganda claims have been refuted and discredited. So don't abandon the truth and keep developing with J2EE.

    See the following for details:

    http://java.sun.com/features/2002/07/rimapatel.html

    http://otn.oracle.com/tech/java/oc4j/
    pdf/9ias_net_bench.pdf

    http://www-3.ibm.com/software/info1/websphere/news/ibmnews/compreview4.jsp

    >>enjoy 8 times faster performance, 4 times less code base

    From the article:
    Secondly, .NET PetShop broke complex join operations to simple SQL queries and placed these SQL queries into stored procedures in the database.

    Now evaluate Microsoft's .NET PetShop from a design perspective. Clearly, their PetShop violates several recommended best practices. Imagine a real-world application designed just the way .NET PetShop is. It would run into a maintenance nightmare because of all the business logic in ASP .NET pages (written in C#), which are, in fact, meant for presentation.

    Microsoft took advantage of an application intended to demonstrate usage of J2EE APIs and best practices, by trying to use it as a benchmark, something for which the ECperf application is better suited. (See http://ecperf.theserverside.com.) Meanwhile Microsoft's .NET PetShop violates many basic design principles in a tradeoff for better performance.

  15. More .Net misinformation[ Go to top ]

    Actually in .Net the code is in a seperate file and compiled to a seperate ddl totally seperate form the asp page. Secondly, using stored procedures is considered a "best practice" because of of the performance benefits. There is no reason other than poor coding that the Java implentation of the project did not use stored procedures.

    Bottom line is both programs have the exact same funcitonality, but the .Net implementation is faster and development time was shorter. Without J2EE using Java Servlets a good java programmer could write an application that could compete with that .Net application without all the complexity of J2EE. In the real word 80% of the applicaitons do not require the complexity of J2EE. And by looking at the simple comparison, productivity and performance is hurt when you do follow the J2EE standard and all its "best practices".

    To me a program put out there should show me the "best practices" that give the quickest way to develop a program and still optimize the code so that it runs as fast as possible. There is no excuse for a program that touts "best practices" to be so poor in the performance.
  16. Apache OJB project[ Go to top ]

    Rolf,

    Regretfully for MS, it is not wishful thinking on my part. Based on what I read in the press and conversations with my colleagues, .Net adoption has been poor so far.

    Regarding MS's claims about performance, code base and productivity, they are simply ludicrous, MS marketing department is not doing a good job.

    Take a look at free, open source JPetStore by Clinton Begin. Make sure you download JPetstore white paper, it's an interesting read. Note that the JPetstore demo runs on Pentium 233Mhz with 64MB. You'll have hard time running Windows alone on such hardware, forget about Petstore ...

    Have fun
    Igor
  17. Apache OJB project[ Go to top ]

    Rolf,

    Performance aside, even Bill Gates, in interviews over the last couple of weeks, is saying that .NET is getting a slow start out of the gate. Poor marketing, poor adoption, poor strategy unification, some other things contribute to this according to Gates.

    -Newt
  18. Apache OJB project[ Go to top ]

    Newt: "Performance aside, even Bill Gates, in interviews over the last couple of weeks, is saying that .NET is getting a slow start out of the gate. Poor marketing, poor adoption, poor strategy unification, some other things contribute to this according to Gates."

    Having spent a few hundred hours looking at .NET, it really isn't that good to start with. To a Java developer, it offers absolutely nothing new, and there is no benefit to switching.

    OTOH To a Microsoft C/C++ Windows developer that for whatever reason is still a Microsoft C/C++ Windows developer in 2002 and has no reason to support anything but Windows, there are a lot of good reasons to go to .NET. Basically, it is the Microsoft-endorsed Java for Windows development. And frankly, the Java community has only itself to blame for not having taken better advantage of its 6 year headstart to provide better alternatives in this area.

    So I would say that outside of Windows-only development, .NET is DOA. For enterprise apps, .NET is DOA. For the common little IIS-based projects and for VB conversions and for C/C++ Windows-only type projects, it will do well.

    Peace,

    Cameron Purdy
    Tangosol, Inc.
  19. Apache OJB project[ Go to top ]



    "Poor marketing, poor adoption, poor strategy unification"


    Poor Microsoft that are never doing anything right. (Do I need to put Irony-tags?)


    "To a Java developer, it offers absolutely nothing new, and there is no benefit to switching"


    Except 8 time’s faster performance, 4 times less code base, and 2 times more productivity.


    "So I would say that outside of Windows-only development, .NET is DOA. For enterprise apps, .NET is DOA"


    FUD, FUD and more FUD.



    The question "How many SUN developers go on one Microsoft developer" (in quality of code) I can not say for sure but probably lies around 20 something.

    Regards
    Rolf Tollerud

  20. Apache OJB project[ Go to top ]

    Rolf,

    You've got to be kidding me.

    Cameron: "To a Java developer, it offers absolutely nothing new, and there is no benefit to switching"
     
    Rolf: "Except 8 time’s faster performance, 4 times less code base, and 2 times more productivity."

    I spent several weeks comparing performance numbers with some .NET freaks (in the nicest sense of the term) and the concensus was that most JVMs and the Windows CLR are pretty close in terms of performance, with Java having the edge in a number of tests and the CLR having the edge in a number of tests. Most were within a few percent either way.

    Cameron: "So I would say that outside of Windows-only development, .NET is DOA. For enterprise apps, .NET is DOA"

    Rolf: "FUD, FUD and more FUD."

    I call it as I see it. Since the release of .NET, I've been to visit internal enterprise development groups at about 15 of the Fortune 100, and I have seen no (zero, nada, none) usage of .NET for enterprise application development. I know that it's not a statistically valid sample, but until I actually see any company using .NET, how do I know it exists? ("Faith is the substance of things hoped for?")

    Unless, of course, you meant that saying that .NET is Windows-only is FUD, in which case I would like to ask how many .NET applications are in production deployment on:

    1) Solaris
    2) AIX
    3) OS/390
    4) HP/UX
    5) Tru64 Unix
    6) Palm handhelds
    7) Nokia phones
    8) Smart cards
    9) Mac (you know, about 5% of all client computers)

    I await expectently your ridiculous answer.

    Rolf: "The question "How many SUN developers go on one Microsoft developer" (in quality of code) I can not say for sure but probably lies around 20 something."

    Yes, now I remember, you hate Sun because your company was using J++, and Sun sued Microsoft when Microsoft breached their contract, and Microsoft decided to kill J++ instead of adding RMI and fixing the incompatibilities that the contract required (such fixes which would not have required the removal of the "non-standard" Microsoft technologies that you were relying on).

    Well, frankly, I don't see what your irrational hatred of Sun has to do with anything, but if it makes you feel better to hate Sun and as a result to use .NET, then I hope you haply bask in your silliness.

    From a technical standpoint, Java and .NET are very similar and both have a good number of points. From a business standpoint, .NET is a dead end. It's still DOA.

    Peace,

    Cameron Purdy
    Tangosol, Inc.
  21. Apache OJB project[ Go to top ]


    "From a business standpoint, .NET is a dead end"

    Did you go to Havard Business School?


    Let’s calculate, We have products that are going to be distributed to our 250 customer companies. None of us use Visual Studio, preferring text editors + DbgCLR.exe, the standalone debugger.


    .NET cost

    Development environment 0 (zero)
    Cost for app servers 0 (zero)

    Current Weblogic cost (please fill in)

    .WebLogic cost

    Development environment ? * 4
    Cost for 250 app servers ? * 250


    Vendor lock in both cases. Please don't tell me that the Weblogic app can be easily moved to WebSphere. I heard that before..

    Regards
    Rolf Tollerud
  22. Apache OJB project[ Go to top ]

    Rolf: "Let’s calculate, We have products that are going to be distributed to our 250 customer companies. None of us use Visual Studio, preferring text editors + DbgCLR.exe, the standalone debugger.
    .NET cost
    Development environment 0 (zero)
    Cost for app servers 0 (zero)"

    Those servers are running all free software? No Windows? No "Core CALs"? Last I checked, it was $1000 for Windows 2000 Server, and that was with only 5 CALs. You may also need Exchange, SQL Server, ISA, HIS, SMS, BizTalk, Commerce, CMS, ... and that all adds up to $0?

    If you really wanted to go the most cost-effective route, it would not be Windows.

    The J2SE and J2EE SDKs are also zero, although it's not really a fair comparison because there's a lot more functionality between those two than you can find in the .NET libs + IIS.

    If you want full-featured application servers for J2EE, they range in price from $0 to about $100k per CPU.

    Rolf: "Current Weblogic cost (please fill in)"

    $10k per CPU for the "advanced" edition, $17k per CPU for the clustered version thereof.

    Rolf: "Cost for 250 app servers ? * 250"

    Assuming you buy the advanced edition of WebLogic at its list price for 4-CPU boxes, then 250 servers will cost $10MM.

    Rolf: "Vendor lock in both cases. Please don't tell me that the Weblogic app can be easily moved to WebSphere. I heard that before.."

    You should have listened. It is fairly easy to move apps from WebLogic to WebSphere, or JBoss for that matter. Some tools now completely automate the multi-app server support (this is only for the server-specific EJB deployment descriptors). JSP-based apps, the corrolary of ASP.NET (basically the only thing that .NET/IIS offers on the server side), don't require any changes to move between different app servers. We have J2EE apps that run on WL, WS, Orion, Tomcat and Resin without any changes.

    It looks like you made a bad business decision and now you're trying to make yourself feel better about it at other people's expense. That is unfortunate.

    Peace,

    Cameron Purdy
    Tangosol, Inc.
  23. Apache OJB project[ Go to top ]


    Sorry, our customers already have all the software they need, apart from our application.

    If you are saying that open source software is best why are Weblogic still selling in spite of JBoss being available.

    I see only commercial proprietary software in all our customers’ mission critical applications.

    So according to your own information, the cost for Weblogic will be ca $10MM. The cost for .NET solution is 0 (zero), including developing licenses.

    Except our own software of course, we can charge a little more knowing that we do not have to share with anybody.

    Regards
    Rolf Tollerud
  24. Apache OJB project[ Go to top ]

    Rolf: "Sorry, our customers already have all the software they need, apart from our application."

    And if they already had Weblogic, the same would be true.

    Rolf: "If you are saying that open source software is best why are Weblogic still selling in spite of JBoss being available."

    Please point out where I said that open source software is best. Apparently you've never had a "choice" about anything.

    Rolf: "I see only commercial proprietary software in all our customers’ mission critical applications."

    For the most part, I concur. I would add that I've yet to see .NET, and IIS (the few times I encounter it) is usually a glorified router to a J2EE app server.

    Rolf: "So according to your own information, the cost for Weblogic will be ca $10MM. The cost for .NET solution is 0 (zero), including developing licenses."

    Well, at least one of us was honest. Windows server licenses are still not free. They did pay for them, or they will pay for them, or they are stealing them. Pick one.

    Peace,

    Cameron Purdy
    Tangosol, Inc.
  25. Apache OJB project[ Go to top ]


    Your logic escapes me. They have already paid for their NT license, in order to run Exchange, SQL Server, BizTalk or whatever. They don't need to pay it again, or steal..

    Of course $10MM is only pocket money for Java developers, but if you save $10MM here and $10MM there before you know it you have a sizeable sum. Even small amounts add up!


    Regards
    Rolf Tollerud
  26. Apache OJB project[ Go to top ]

    <Q>
    Those servers are running all free software? No Windows? No "Core CALs"? Last I checked, it was $1000 for Windows 2000 Server, and that was with only 5 CALs. You may also need Exchange, SQL Server, ISA, HIS, SMS, BizTalk, Commerce, CMS, ... and that all adds up to $0?
    </Q>

    Does WebLogic have all these functionality? If you run it on 'Enterprise' class server such as Solaris, what does it cost? I know the numbers. It's ridiculous to try to even argue that MS platform costs more.

    It's only true for Linux and JBOSS. But why are we all WebLogic/Websphere people? My take on that is total cost of open source solution as it is now if still higher.
  27. Apache OJB project[ Go to top ]

    k k: "Does WebLogic have all these functionality? If you run it on 'Enterprise' class server such as Solaris, what does it cost? I know the numbers. It's ridiculous to try to even argue that MS platform costs more."

    Solaris is free up to a certain number of CPUs per server. Above that point, it is free if you ask (ask at end of quarter ;-). Remember, Sun is a hardware company.

    k k: "It's only true for Linux and JBOSS. But why are we all WebLogic/Websphere people? My take on that is total cost of open source solution as it is now if still higher.
    0 replies in this thread"

    From what I've seen, total cost even for an extensive Weblogic system is low compared to the systems that preceded it, and .NET (and JBoss etc. for that matter) doesn't even get considered in that space. The cost of the hardware and the Weblogic licenses and the Oracle licenses often runs well into the millions for a single application, and those costs are only a small fraction of the overall cost of the system.

    As for the open source solutions, they are very good for ISVs with small-scale apps. I'm only implying that they are _most_ applicable for the 1x or 2x CPU server that has everything self-contained and needs zero admin. This is more like what Rolf is talking about, except he's talking about running on Windows, which is not free and is not anywhere close to zero admin.

    Peace,

    Cameron Purdy
    Tangosol, Inc.
  28. Apache OJB project[ Go to top ]


    "As for the open source solutions, they are very good for ISVs with small-scale apps"

    This discussion reminds me of the fight with Mainframes in the 1970 something. The arrogance of the mainframe people was unbelievable and when we needed some data they did all to put every obstacle in the way. A normal answer if you called was something like "why do you bother me, we do real work here, bust". Many years after I met one of these persons, he was driving a taxi.

    It’s all about the psychology of the individual. If you are used to having an exalted position (men in white rocks) and only see DOS you don't understand the evolution of Mac and Windows who eventually conquered the world. Anyway, there are impossible for any young programmer to imagine the crappyness of the command-window programs from the old mainframes. If they ever should see one they would probably laugh themselves to death.

    Having won this war its a little taxing to see the same thing happen again with Unix, this time. They see only Win95 and not NT/XP/.NET Server, the best choice for common browser based intranet applications with about 2000-to 4-5000 users. The Unix people have inherited the old arrogance, they call the Oracle Pet store application a "toy application" (700 files, 30 000 rows of code). Only Google is a real app, according to them.

    The real fight is going to be between Open Source and Microsoft but I am sorry, for a year or two we still have to endure the "big application Server guys". But have faith, the big application Servers from Bea, Sun, IBM and Oracle are soon going the way of the dodo.

    Meanwhile, if I want entertainment I find a site that discuss "J2EE Patterns and best practice".

    Regards
    Rolf Tollerud
  29. Apache OJB project[ Go to top ]

    Rolf,

    Most of the people pointing out your logic flaws come from a Windows background. Been there, done that, and it's yesterday's news. You liken Unix to mainframes. Quite unimaginative. When you take the next logical step, to realize that Windows is now becoming much less relevant, then you will have at least caught up to where Microsoft strategy is (tens of billions invested in Xbox, cable, TV set top, phones, mobile devices). The next logical step, to realize that Microsoft has struck out (or is bleeding profusely) on all those initiatives, is to realize that the attempt to force yesteryear's paradigm onto tomorrow's technology is a mistake that Microsoft learned from the likes of Burroughs and Digital.

    IBM was just as cocky back in 1980 and even 1990 as Microsoft is today. A few visionaries realized then that IBM had already lost the coming battle. Today the same is true of Microsoft: They've already lost, and you don't even see it.

    Peace,

    Cameron Purdy
    Tangosol, Inc.
  30. Apache OJB project[ Go to top ]


    It's the same old game, called "take down the leader". Been played thousands of times in history. Refresh your memory of "Pericles" for instance (a hint: search for the Golden Age of Athen).

    Regards
    Rolf Tollerud
  31. Apache OJB project[ Go to top ]

    Percules?
  32. Apache OJB project[ Go to top ]


    Ian Chamberlain:

    "You miss the point. If MS is forced to give up the Windows monopoly they just move one level higher, up to the virtual machine. All the existing MS products are busy being rewritten for .Net as fast as MS can churn them.
    There won't be a Linux version of SQL, IIS, Exchange, etc, there will be a .Net one. All MS then has to do for plan B to become operational is to port the .Net framework. How long do you think that would take if MS really put resource into it? What makes you think they aren't already doing it in the labs?

    Sun shouln't be trying to wrest the windows monopoly from MS through the courtroom and the State stooges. It might be good tactically but is real dangerous strategically. Imagine what MS products available on Linux would do for Solaris and Java. The biggest threat to java is not .Net on Windows. That is not a threat in the enterprise space where J2EE holds sway. The real threat is if .Net is ported to Linux. MS won't give up the Windows revenues
    to do that at the moment, but if the Windows revenues are taken away you have a different ball-game.

    Any-one who thinks Microsoft is stupid hasn't paid attention for the last 25 years. You don't get as big and powerful as MS by being stupid. Whenever MS has been put in a corner they've managed to wriggle out of it and surpirse
     verybody. Six years ago they had no internet plan, now everyone is scared of them hijacking it. Don't look for MS to make the obvious play, look for them to make the smart play and work out how to deal with that.

    Anyway who says a portable .Net is a bad thing for the market? I personally think competition is a good thing and look forward to Java(tm) vs .Net(tm) wars raging for some time to come."

    Anyway I love the fight. The first to go will be Sun!

    Regards
    Rolf Tollerud
  33. Apache OJB project[ Go to top ]


    Rolf said: "The first to go will be Sun!"

    I'll give you the benefit of doubt....let's say Sun did go first. Java would live on! Java and J2EE are open standards implemented by a number of vendors. IBM would jump at the chance to be the next steward for Java....and a hundred other companies are lined up behind them.

    .Net will lose because it is proprietary, and the world has moved beyond that. If Microsoft falls, splits up, screws up or is tied up .Net will fail with it.

    "Java is not a product, it's an industry." --James Gosling

    Cheers,

    Clinton
    www.ibatis.com

  34. Apache OJB project[ Go to top ]


    I have no problem or nothing against Java, only Sun.

    I was in fact one of the early adopters of Java. Perfect for Com objects. Javascript with Java Com objects was kind of "lite JSP".

    But Sun deliberately sabotage/destroy/disapprove of all attempts to make Java work well with Windows.

    If an open standards body, or better still Jakarta or even IBM (excellent job, eclipse) controlled Java there would not be any problem, and we could start to patch together all problems that Sun have generated.

    Regards
    Rolf Tollerud
  35. Apache OJB project[ Go to top ]

    Rolf said: "But Sun deliberately sabotage/destroy/disapprove of all attempts to make Java work well with Windows."

    How about trying to post some PROOF. An article, a hyperlink....anything. Nope, none there. You're just posting garbage and looking to stir emotions. Sad.

    Clinton
  36. Apache OJB project[ Go to top ]

    But Sun deliberately sabotage/destroy/disapprove of all

    > attempts to make Java work well with Windows.

    I'd be interested to know exactly what you're referring to be that.

    If you're referring to the old MS lawsuit, then I think that your statement would be more correctly stated as 'attempts to make Java work _only_ with Windows.'

    BTW: Did you know that MS did eventually release a separate download of Windows specific Java API extensions for developers to use - and that AFAIK was not blocked by Sun. If you _really_ had wanted to make your programs work better with Windows, then you could have used that.

    NB: If MS had used that approach at the beginning, rather than trying to subvert the entire Java platform to being Windows-only, then the lawsuit wouldn't have happened.

    Interestingly, tried to find it and it looks as if it has gone (like the MS VM!), but it is always difficult to tell - it could still be part of their Java SDK, but MS _never_ mentions any products apart from their own. Guess it saves on all those tedious interoperability issues!

    > If an open standards body, or better still Jakarta or even
    > IBM (excellent job, eclipse) controlled Java there would
    > not be any problem, and we could start to patch together
    > all problems that Sun have generated.

    If Java had been submitted to a standards body early on, the dominant standard would have been controlled by MS, by the nature of controlling the dominant development toolkit, and rapidly building in a set of incompatible (with any other implementation) extensions.

    Don't believe me? OK, real-world example: Browsers. If you go to you manager and say 'We're going base our site on w3c standards', he'll say 'Will it work with IE'. 'Not really' you'll say, 'IE isn't very standards compliant'.

    Well, your manager will then tell you that IE _is_ the standard, and that the w3c is just (to him) an inconvenient irrelevance. So it would have been with Java under your standards scheme.

    You might not value your choice of OS, but I'd rather that you don't argue to have my choice removed...

    Thanks.
      /david
  37. Apache OJB project[ Go to top ]

    You've got to be kidding, .Net is taking off very strong and I've never heard any statements from Redmond to state otherwise. I've even met more than 1 java developer that has brought VS.Net and the more they get into it the more they are impressed (which didn't have hppened with VB6).

    They are not allowing anyone to get thme involved in the MS/Sun holy war and are making sure that no matter what happens they will be able to write good applications, no matter who makes the tools. Plus, there is an open source projects under way to port the .Net framework to Linux, and the target release date is sometime late this year for the BETA. VB was behind Java, but while SUN consentrated on attacking MS, MS has come out with a product that is just as good and maybe even better considering the RAD abilities that Java does not have.

    If it wasn't for the holy war between Sun and Microsoft getting industry standards would be much further advances and all of us as developers would be benefiting from those standards.

    Also for a list of companies, inlcude Fortune 100 and 500 companies you can go to MS web site. There are 100+ early adapter of .Net, and a good number of those are Fortune 100 and 500 companies.
  38. Apache OJB project[ Go to top ]

    Feel better now?
  39. Apache OJB project[ Go to top ]

    I have Windows XP, VS.Net, MS Office all running on a Pentium, 66mhz compute with 396 RAM without any problems or speed issues. The only program that really runs dog slow is Forte for Java when it initially loads.
  40. Don't kid yourself. Designing and constructing enterprise systems is not something you can adequately cover with wizards. Assuming they don't generate poorly designed, non-maintainable code (which is usually a bad assumption), they are nice for getting you started. But you need help with the inevitable tweaking (an understatement, really) that is *always* necessary. How about automated refactoring, real-time validation against specs/good programming practices, etc...

    IMHO, writing these kinds of applications with EJB and J2EE has become much easier thanks to the specs. Sure, there is room for any number of improvements, but I think Java really shines on the server side. I am not interested in using tools aimed at neophytes, because they remove flexibility that I need. They may get you 80% there, but end up giving you applications that don't scale and that are difficult to maintain. They also encourage poor programming practices. And what if you need to do something outside of the model they provide? Well, you are out of luck. I never could create systems I was happy with using VB for these very reasons.

    There are some things that could, and probably will be changed for the better in the specs, but the greatest room for improvement is in the tools space. Java by far has the best coding IDEs, but there is still much to do in the visual tools space. I am more optimistic here than you are. A number of these tools already remove the drudge work involved in creating components like EJBs, packaging them, and deploying them. Things will continue to get better. Instead of having just Microsoft working on improving tools, we have a number of companies/individuals in the Java space that will continue to innovate.

    I welcome .Net, because it will make Java better in the end. Java still has a bright future.

    Respectfully,
    Bill

  41. /**
    Don't kid yourself. Designing and constructing enterprise systems is not something you can adequately cover with wizards **/

    I think Bill you nailed it on the head. Unfortunately lot of people pay too much attention on the Wizrads and tools becuase it is very easy for them to visualize. Throw some one the best Microsoft technology ever (VB) and the best of breed tools (Visual studio) and ask them to write multi threaded Trasanctional applications, on the other hand I can use my $16 Ultradedit editor using Java/J2EE API, guess who can acomplish the task first:-)

    Again it is the right framework that counts in the long run, not some fancy Wizrads/tools.

  42. On the wizards topic (off-topic a little):

    I just don't think that you can type descriptive function names into ejb-jar.xml (both the bean declarations and security and transactions), MyBeanHome.java, MyBean.java, MyBeanEJB.java for any project worth using J2EE as quick as a tool could generate it. I agree the tool should not write any real code, just declarations so the programmer still has to think about the code. There are finite options for attributes of functions (Local/Remote, return type, arguements, security role required, transaction attributes) that require a lot more time to type than is neccisary. I'll have to take a look at XDoclet, but I'm skeptical. It would just be great to reduce the chance for typos, forgotten or bad throw clauses, forgotten security/transaction attributes for functions, etc. It would be extremely cool if there was a way to specify all pertinant information about a bean in one place. As it is now, you have about 6 different files to look in. I don't particularly care for all the redundant typing even with auto-completion (which is a lot better than the old way). Refactoring should be handled by the same tool so that it is "EJB Aware" unlike most editors that are only aware of direct references from other files. I for one would pay for such a tool if it had a nice GUI :)
  43. Hi Sam,

    You may want to take a look at ObjectAssembler. It initially generates all of the necessary code for your EJBs. It allows you to view your enterprise beans as one single component, while still allowing you to easily get under the covers. You may choose to work with your beans with this visual representation or from the code editor, and both are kept in synch. It also validates (against the specs) your beans in real-time as you work.

    I would be interested in your opinion of how we do this. FYI, we are planning a maintenance release in about a week (just let me know if you run into any problems before then).

    Happy coding...

    Regards,
    Bill
  44. <quote>
    tool could generate it. I agree the tool should not write any real code
    </quote>

    WebLogic 7 ships with a GUI tool called Builder that will automatically create deployment descriptors for you if all you have is EJB classes. Builder will go as far as figuring out relationships, CMP fields, etc...

    --
    Cedric

  45. I'd disagree with you since VB has had transacional processing since VB5 and although limite support for threads in VB6, some things could be done. With .Net it has full mutlithread support, including thread pooling. I would bet that someone considered a .Net expert (such as Dan Fox, Dan Appleman, etc) could crank out the code to do multithreading and transacitional processing faster. Once .Net has been out for 1 there will be a lot more programmers out there that could also do the same quickly.
  46. Excellent story explains a lot about Mr Mcnealy,

    A fool's paradise for CEOs
    http://www.salon.com/tech/feature/2002/07/23/solutions/?x
  47. <Quote>
    Q JBoss has had a hard time getting J2EE certification. Do you think it's important to have a certified open source J2EE implementation?

    McNealy: No, we've already got one. It's called the Sun ONE app server. It's certified. Do I think it's important? I don't know what that means.
    <?Quote>
    Nobody was ready to pay for iPlanet/SunOne so we are giving it away for free and hoping to make money on the Service contracts.
    Scott McNealy<Quote>
    So, potentially you could make an argument that the open source thing is just screwing up all the revenue models and we aren't getting the advertising, because it isn't the best technology that always wins, it's who advertises more.
    </Quote>
    Translation.
    We were making big bucks selling software that people don't really need. We need to advertise to convince people that they need all the bloatware.

    Sadly, The only major advantage Java/J2EE has over .Net is the huge amount of Open Source stuff that's available. The other being that it's been around longer.
    Of course you can argue that .Net is unproven/unstable/etc.
    But most business app managers don't really care. They want something that is good enough and gets the work done quickly. If they feel .Net can get the job done they will go for it. That's the reason why VB used to be so popular.
    Cheers
    Ravi
  48. The only thing hurting the sales of application servers like iPlanet is that many of them suck.

    Everyone needs to understand the purchase process and apparantly even Scott McNealy doesnt.

    Customers make a purchase when the value of an item exceeds the price of that item.

    That seems simple right? People forget it all the time. For instance, take Scott's example. He believes becase a product is free, people will always choose it over the pay version. If that were true Scott, we'd all drive free cars, live in free housing, and wear free clothes.

    Plenty of shops buy app servers. Look at BEA and IBM's revenue numbers. Thos shops believe that the value they get from IBM WebSphere exceeds the price of IBM WebSphere. Its that simple Scott. The sad truth you cant get past is that you were unable to put enough value in your product to overcome a cost above zero.

    JBoss being free indeed means that it is easier for value to exceed price. Welcome to the real world Scott. Put value in your product worth paying for, and people will pay. The laws of economics havent changed.

    First attack MS as evil. Next attack Open Source? Scott, get back in the boardrood and not the soap box and write some code that doesnt suck will you.

    Dave Wolf
    The Scupper Group
  49. Correct.

    iPlanet app-server were not selling, but both WebSphere
    and WebLogic are selling.

    The obvious conclusion is that customers prefer
    the IBM and BEA products for the SUN product.

    And is is neither fair or make business sense
    for SUN to blame an open source product for customers
    not liking the SUN product.

    And if SUN wants .NET to kill Java and J2EE, they
    should just claim ownership and start charging for
    everything.

    In my opinion the success of Java is more due to
    the myriad of free tools (all the Jakarta stuff,
    JUnit, JBoss, jEdit etc. etc. - sorry the list
    is too long to complete) than to SUN.

  50. And is is neither fair or make business

    >sense for SUN to blame an open source
    >product for customers not liking the
    >SUN product.

    Perhaps Sun should open source their iPlannet application server.
  51. I agree with this 100%. I have worked on a number of enterprise products and not one of them has used JBOSS. I think JBoss is an interesting idea and something I might personally use, but customers want to see names like WebLogic and WebSphere when discussing Application Servers.

    The bottom line is that back in the early days of this J2EE world, Sun had a choice: they could have bought Weblogic or they could have bought NetDynamics. They bought NetDynamics and they chose poorly! Now they are trying to blame OpenSource for their problems.

    IBM and BEA are going head to head in the Application Server space. For Sun the game is over and they are out of the running. They should look elsewhere in the their attempts to generate revenue from their Java efforts.

    As for Sun's future, I worry about that too. That's even more reason for Java and its associated standards to be regulated by industry standards organizations rather than Sun.


  52. Sun's purchase of NetDynamics was not the big mistake; at the time ND was far ahead of BEA's product. The big problem came over the next 18 months, as Sun and AOL stumbled badly in getting an iPlanet product out the door. I will definitely give BEA their props, the team there took advantage of this by strengthening their product line and making some smart marketing decisions. Also remember that for sales (and not technical) reasons, iPlanet executives chose to use Netscape App Server as the basis for iPlanet App Server.

    If Sun had never gotten into the iPlanet deal, or if the executives had managed the transition better, the market might look very different now.
  53. <quote>
    I have worked on a number of enterprise products and not one of them has used JBOSS. I think JBoss is an interesting idea and something I might personally use, but customers want to see names like WebLogic and WebSphere when discussing Application Servers
    </quote>

    Isn't that the point of the question in the article?

    If JBoss had the 'J2EE Certified' stamp, customers would be more likely to consider it alongside WL and WS.
  54. Hi all,

    I think J2EE must have:
    - Very good IDE like Microsoft Visual Studio. It’s very hard to deploy a J2EE application to Application Server.
    - Many and many companies, such as: IBM, SUN, BEA, Borland… and many open source organizations support J2EE standard.
    - Added standard like SOAP to communicate from J2EE platform to MS .NET platform. (The HTTP standard is a technical revolution)

    Best wish,
  55. <Q>I think J2EE must have:
    - Very good IDE like Microsoft Visual Studio.
    </Q>

    It does.

    <Q>It’s very hard to deploy a J2EE application to Application Server.
    </Q>

    There are Java IDEs that make it simple.

    <Q>- Many and many companies, such as: IBM, SUN, BEA, Borland… and many open source organizations support J2EE standard.
    </Q>

    ???? They do.

    <Q>- Added standard like SOAP to communicate from J2EE platform to MS .NET platform. (The HTTP standard is a technical revolution) </Q>

    There is.
  56. I agree with most of you. I have always seen Scott making stupid comments about microsoft. I agree that to some extent .Net etc is not that scalable, windows crashes , not robust etc etc etc etc. But Scott always makes it a point to say atleast 5 bad comments at microsoft. I ahevtn heard bill gates doing it every now and then. Visit microsoft.com for the latest comments from bill gates to the gathering of Microsoft employees. Whatever is the condition of software of microsoft makes, Bill has the professionalism top speak in public, make logical comments. Its a SHAMe top hear such a shallow and unthoughtful comments from Scott about open source stuff. I think open source gave us more than what any company gave us ever.
    Also if you provide good service, there are companies ready to pay you. As someone here said, lok at BEA and Web sphere and even oracle back with a bang, they dont given everything for free, but still make money.
      Anyway .. I hope Scott gets his act together ( he seems to be frustated with CTO leaving last month, then stock price tanked and he has a BIG EGO to lay off)
  57. *Sigh*

    I strongly belive that Open Source is the only thing that can keep Java from obsolescence. For Sun to start posturing about OS Java is just another in a long line of reactionary positions the company has taken.

    I hate raising the issue of .Net, becauser this forum has a tendency to get all evangelical about it, but Microsoft always pushes for the volume market, and if you look and .Net this is no different from their strategies in other areas. MS is really pushing .Net to a much 'lower-end' market than traditional 'Enterprise' vendors have approached, because what is lacking in BIG$ is made up in volume.

    My take on it is this:

    Entry: ? (whatever is cheapest & easiest)
    Middle: .Net, Open Source J2EE, ColdFusion & JRun
    High: Big J2EE vendors .Net(?)

    Open Source J2EE provides an excellent migration path for companies to move to higher-end products and systems as their business needs require, and really provides the only real alternative to .Net in the middle space of Enterprise systems.
  58. The problem is Sun failed to capitalize in their Java technology. They let opportunities slip through their fingers only to have others make a ton of money on it.

    IBM, and WebLogic did pretty well with Java. I see that they have a marketing problem and not a technology problem.
  59. If people at Sun are reading this, there are still markets for Java technology that are wide open where they can make billions if they are smart enough and bold enough to capture those markets.
  60. I agree... the tools market is still wide open. Even though no one I know is going to trade in Eclipse or Idea, a good tool for EJB content generation that has an intuitive interface and can work well with existing database structures (automatically generating Entities) would be great. I've tried a few of them out there, and I can still code it quicker than those tools can produce it, if they even produce it properly. I'm sure there is a tool out there I haven't tried that might be good, but if SUN had it, I would know about it and probably purchase it if it could generate my entity beans and some skeleton Session beans. Some tools use custom java libraries to make things easier (Borland comes to mind), but most coders (at least me) hate the idea of adding another layer of fluff that isn't part of the official specs or we have the source and a license that allows us to play with it if we find a bug. I moved to Java to get away from Borlands BPL's that are buggy in places, and they take precious time to fix it leaving you to cludge your way around it. I moved to Java to avoid the tens of _documented_ bugs in windows that will never be fixed. (would be more, but there were about 10 undocumented bugs in office/windows as well and these are only the ones that costed my project thousands in developer time) Java is robust, but honestly, should you have to tell the computer 4 times in XML, the interface, the bean, and in the security descriptors what functions are available remotely and locally? Wouldn't a bean declaration itself and another simple XML file generated by a wizard be enough? In fact, couldn't the wizard with simple check boxes eliminate half your ailes in typos of function names and force developers to make more compatible code? SUN should stop relying on IBM, BEA, IntelliJ and the like to get the tools right. If they would get off their a$$es, they could make something better than Forte or their silly EJB wizard.

    Another market that hasn't been explored by SUN is the Bean Provider role. They would make EJB's for sell to handle complex financial calculations or statistics. Who says everything they make has to be for free. I want to see them make some money too. They have some nice hardware, but it's really hard to compete with AMD, Intel, Motorola, and IBM. Intel's on their way into the high-end server market, and it's only going to get more difficult. They could revive the whole "Java Accelorator" concept, but I think they just found that creative software on faster hardware was cheaper than more complicated hardware. It would be really nice if they combined efforts with someone like transmeta to make a "Java CPU" with "code morphing"(TM).

    Honestly, they can't hope to win without the technology being ubiquitous, and that's what open source does. They need to embrace open source to make it popular and sell support or other productivity software. If JBoss was certified, it would be everywhere, and those companies would have more money to spend elsewhere (like tools and components).

    Finally, if all else fails, they could sell J2EE solutions themselves. Who would be more compitent at producing good Java code than the company that invented it?
  61. Java is robust, but honestly, should you have to tell the

    >computer 4 times in XML, the interface, the bean, and in
    >the security descriptors what functions are available
    >remotely and locally? Wouldn't a bean declaration itself
    >and another simple XML file generated by a wizard be
    >enough?

    Use XDoclet, tag the remote methods in your source code's javadoc comments, and all the rest is generated automatically for you. There's even a GUI being worked on to help you add the tags...


    Andrew.
  62. I agree with you that they should seel J2EE themselves. SUN should onsider diversifying in:
    1. Being a J2EE consultant
    2. Being a J2EE Consultant's consultant.
  63. The venerable Scott McNealy needs to go, it's as simple as that. During his tenure, Sun has done way too many things wrong and failed to capitalize on chances to make a big impact (and money).

    Oh, what the hell - though I prefer writing code, I could clear my busy schedule for a few months and take over at Sun. Sun board members: you know where to reach me.

    :-P

  64.  I agree with everything said here (well almost).
    Sun was slow to pick up the ball in regards to app servers.
    Sun One is not in the "A" catagory yet.
    Their hardware is overpriced when you compare it to something like Oracle's unstoppable Linux.
    Mr. McNealy should not bad mouth some of Java's most devoted followers.
    Repeat after me : Open source is good !

     But, let us take the time to remember, there would be no Java if it was not for Sun. IBM would not have done it.
    IBM is deftly using open source to bludgeon Sun (and hopefully Microsoft) with. It doesn't take Sigmund Freud to see why IBM named there open source IDE - Eclipse - in response to Sun's Open Source Netbeans; and has helped shut Sun out of some key web services committees. The only reason I point out IBM is because: who else is there ?
    At least IBM is giving to Open source. I am not saying that Sun was being purely philanthropic; but, I think what Scott
    is really saying is : HELP!!
    Sun is being eaten at the bottom end by Linux and MSFT and by IBM and (soon) HP at the top end. Open source is Scott's scapegoat. If he could only see that it is his savour.
    Wake up and smell the coffee Scott. You gotta compete on price with Linux; make a Sun Linux on AMD's Opteron; sell the boxes pre-configured with open source software; then beef up the SPARC and invest in Solaris. Keep the full version Star Office open source and invest in your support
    team.
     All this advice is open source.
  65. Another way Sun can compete is to agressivley push their iPlanet server onto third party software that needs an application server to run on. Software application that use servlets often ship their demo software and, it would be better if it had a full blown app server as part of their installation instead of just a servlet/JSP engine.

    Sun need to work with the J2EE software application developers to make it easier for them to embed their product into iPlanet and to help them produce a demo CD of their app embedded with iPlanet.
  66. I don't understand his arguement.
    A) Open Source takes revenue from people that would advertise J2EE
    B) J2EE isn't being advertised because of A.

    But I don't understand how this could be possible considering JBoss's popularity. The first time I a lot of my friends heard about J2EE was the JBoss article on Slashdot.

    McNealy shouldn't be upset at all. If people are choosing .Net because it's advertised and not because it is a better product, then those companies are seriously lacking in research and intelligence. When Company.Net gets their product to market and have to compete with J2EE-Company, Company.Net is going to have to match J2EE-Companies prices. When they do, every licence fee for Windows X.X Serever and every license fee for MSSQL is going to come out of Company.Net's pocket. J2EE-Company will be able to sell their product with JBoss/Jonas/Tomcat/Jetty and MySQL/PostgreSQL/SAP/HSQL on a Motorola/AMD/Intel/Sparc/Arm/Toaster-Oven running OSX/WinNT/Linux/Unix/Aix/Solaris for much less overhead. When .Net companies start dropping like the .Coms, who advertised better isn't going to make one bit of difference. The server/enterprise market is virtually unpenatrable by MS anyhow because of previous and ongoing security, performance, licensing, or just plain stability issues. Advertising is for week products or for products targeted at week minded people.

    Honestly, do you want the fool who didn't research to blame his failure on J2EE because he saw a cool ad for it, bought it, but was too incompetant to make it work?

  67.  Is it a breach of etiquette to reply to one's
    own post.
     Well, I just had one more piece of open source advice
    to Mr. McNealy. Make the Intel?AMD(32bit) version of Solaris open source. Your using Gnome now on Solaris just port KDE and add gcc to the package.Solaris on Intel/AMD could start to have a mucher wider audience and converts. When they need more performance and stability let them have Sparc.
     Intel/AMD Solaris is still ahead of Linux in quite a few
    kernel internals.When the open source community gets a hold of it, you will have more branding than you'd get from a Superbowl commercial; and you'll get more development than
    MSFT; and you can still control the release like Linus(and his committe) does.
     Open source is here to stay; come onboard !
  68. McNealy: "I think it's important to have a community process, open specs, and choice for the customer, absolutely. How that gets implemented, I don't particularly care."

    Yes, indeed, community process and open specs are key parts of Java’s success. Open specs give companies confidence to select technology from an innovative but not necessarily 'big' vendor ... you know that a spec will be supported one way or another. Open specs give small vendors a chance to compete based on merits. IMO, Sun deserves a lot credit for establishing this environment and for making a good job of maintaining it during all these years.

    McNealy: "I actually think we need more revenue in the J2EE space, so that we can do more advertising to get the message out, because right now the world is getting blitzed with Microsoft advertising, and promotion and branding and propaganda, and big lies, and that's why they're going, not because it's a better product."

    Perhaps more advertising dollars would not hurt ;-) Hopefully, Sun also realizes that Open Source projects have the same effect as millions of dollars spent on ads. Apache, NetBeans, Jboss and many others are extremely successful and well-known Java projects, they provide excellent technology at very low cost, that’s a great message for the Java platform and a direct benefit to its users.

    Sun has done a lot of good for the industry and, surely, it’ll be able to adapt and to make money in this (highly competitive) environment (thorough certification (yes, I do believe Jboss should pay for its certification), services etc). However, I doubt that stepping on Open Source is the way to Sun’s financial success (hopefully, that's not in Sun's intentions at all).

    -- Igor
  69. Its about time SUN produced some intutive development tools for creating J2EE enabled application.

    We are developing J2EE application for two years and we are still looking for something like VC++ that doesnt cost a fortune and eat up all the memory of the PC.

  70. Open source isn't hurting J2EE, it's not even hurting Sun. McWhiner is just sour because the other vendors still have a higher market share than the open source alternatives (Sun does not).

    The fact is, if there were no open source alternatives, we still wouldn't buy iPlanet/SunOne garbage. Why? Go to their website and look to see when iWS going to be Servlet 2.3/JSP 1.2/JDK 1.4 compliant. They can't keep up with their own specs!

    So, is open source hurting Java/J2EE? No. It's the only thing keeping it alive in my opinion. If there were no open source/freeware alternatives in the Java space, then Microsoft might actually have a case for .Net. I know I wouldn't be very excited about Java any more...but maybe Scott wouldn't mind that so much. ;-)

    Cheers,

    Clinton Begin
    http://www.ibatis.com/jpetstore/jpetstore.html
  71. Interesting thoughts from mr McNealy, and I for one agree with most of them. Specifically:
    >No, we've already got one. It's called the Sun ONE app server.
    > It's certified. Do I think it's important? I don't know what
    > that means.

    Absolutely true, he has no idea what that means.

    >I actually think we need more revenue in the J2EE space, so
    >that we can do more advertising to get the message out,
    >because right now the world is getting blitzed with Microsoft
    >advertising, and promotion and branding and propaganda, and
    >big lies, and that's why they're going, not because it's a
    >better product.

    Yes, we need more revenue in the J2EE space, but not in terms of J2EE server sales, but rather in J2EE tools and applications sales. What's more important? 10 ads saying "Here's our great J2EE server!! Buy it!" or "Here's our great app, more functional, more secure, faster than anything else on the market. Powered by J2EE". I'd prefer the latter, thank you very much. More substance.

    In JBoss we have said from the beginning:
    J2EE is infrastructure. Infrastructure should be free and open. Applications written on top of J2EE is where the money is, in terms of product sales. It just makes sense (unless you're BEA or IBM).

    >So, potentially you could make an argument that the open
    >source thing is just screwing up all the revenue models and
    >we aren't getting the advertising, because it isn't the best
    > technology that always wins, it's who advertises more.

    Is OpenSource screwing up all the revenue models? No, just the old ones. Newer models built with OpenSource in mind, such as service models or apps-on-top-of-OSS is better for everyone involved, except those who cling onto the old ways. Well, wake up and smell the napalm. Evolve and adapt or die.

    >..because the open source community is cutting the legs out
    >from under all the R&D and promotion efforts of all the open
    > interface strategies -- not open implementation, but open
    >interface strategies.

    ROTFLMAO! Let's start with the R&D point. Generalizing wildly there are three types of R&D folks: those in academia, those at corporations, and those in OpenSource. In my experience the most motivated and out-there ones are in OpenSource. Why? Because (as the saying goes) "99.9% of the best developers in the world doesn't work for you". So, if you're a really talented developer at a company or academic instituion the best way to develop your 1337 skillzz is to find other talented developers at other companies, and work with them. How? Through OpenSource, or rather "Open Development" a la SourceForge. Kinda obvious, isn't it.

    When it comes to "open interface strategies" I'm not quite sure what to say. It's such a big lie. Pretty much all of the OpenSource projects either a) implement open interfaces or b) use open interfaces or c) both implement and use open interfaces. OpenSource is *the* biggest proponent of open interfaces, because it freakin' makes sense. Open interface strategies for closed source development is always a kind of "yes yes, but no no" because they want it because customers like it but they don't want it because they need an edge, and what better way to do it than to use closed interfaces in some strategic places.

    So while I agree with most of his view of reality, I *like* it this way whereas he don't.

    /Rickard, one of the founders of JBoss

    ps. .. who also expects now to be flamed by employees of companies who shares Scott's fear of what's happening. ;-)

  72. Rickard: "What's more important? 10 ads saying "Here's our great J2EE server!! Buy it!" or "Here's our great app, more functional, more secure, faster than anything else on the market. Powered by J2EE". I'd prefer the latter, thank you very much. More substance."

    Ultimately, the only substance ;-) ... app servers without apps are somewhat useless.

    Rickard: "In JBoss we have said from the beginning: J2EE is infrastructure. Infrastructure should be free and open. Applications written on top of J2EE is where the money is, in terms of product sales. It just makes sense (unless you're BEA or IBM)."

    I would agree that the standards and the ability to interoperate should be free, and nowadays companies rarely choose to base important investments around "closed standards".

    Assuming that the server must be free is a bit overboard. While developers and groups like JBoss have every right to give their work away for free (in any meaning of free they so choose), that should not mean that commercial ventures are not welcome. As I've pointed out before, the goal of JBoss should be to be a great server, not to be an answer or a foil to BEA. The sooner the entire JBoss group gets over its BEA envy (<g>) the sooner JBoss will be free to succeed on its own merits.

    And by the way, access to BEA documentation is free ;-).

    Peace,

    Cameron Purdy
    Tangosol, Inc.
  73. Assuming that the server must be free is a bit overboard.

    >While developers and groups like JBoss have every right to
    >give their work away for free (in any meaning of free they so
    >choose), that should not mean that commercial ventures are not
    > welcome. As I've pointed out before, the goal of JBoss should
    > be to be a great server, not to be an answer or a foil to
    >BEA. The sooner the entire JBoss group gets over its BEA envy
    >(<g>) the sooner JBoss will be free to succeed on its own
    >merits.

    FWIW, I never subscribed to the "BEA envy" part of the whole deal. I agree that it's completely ludicrous and simply a bad approach.

    FWIW, I also never assumed that all servers should be free by people working for free. If someone can make a good server and make a living out of it (such as my good friends over at OrionServer.com), then I'm all for it (in principle). *I'm* not the one to judge *how* to do it: that's what other developers making J2EE apps should do. For me OpenSource is the right way, for other's it isn't. Big deal.

    I welcome any approach to making J2EE work. That's the ultimate meaning of "free" I guess. Which is, FWIW, one of my main reasons why I don't like the GPL: it enforces the view of the author of the code on the one using it. To me, that's not "freedom", because that involves a choice where all variants are "right".

    Aaanyway. Peace, Love, and Good Code.

    /Rickard
  74. <quote>
    And by the way, access to BEA documentation is free ;-).
    </quote>

    Interesting, I was puzzled about this myself. Does anybody know where JBoss documentation is? www.jboss.org "documentation" link points to a collection of ads for JBoss publications. The only "free" documentation available there is "A free volunteer maintained manual was developed in the early stages of JBoss 2.0".

    Does this mean that the product is free but the documentation is not?

    --
    Dimitri
  75. Does this mean that the product is free

    > but the documentation is not?


    Keep in mind that pricewise, you are comapting apple and oranges when comparing Weblogic and JBoss. They have different pricing model.

    Weblogic costs $10K per CPU and JBoss is free, but the extensive documentation for JBoss costs around 10 dollars.

    10 dollars is small price to pay for documentation for a quality product when the alternative is $10K for Weblogic.

    I have seen whiners who complain about the 10 dollar price for the documentation. Nothing is free the time the sponsores of JBoss puts into developing the product is extensive and they are entitled to be conpensated for their efforts.

    As for the whiners, if they have not contributed to JBoss through development or through supporting others then all I have to say is be grateful that you don't have to pay the $10K.

    If people still want to whine about the price then I say go pay the $10K for Weblogic and get their free documentation. As for me, I prefer the free JBoss app server and paying 10 dollars for the documentation.



  76. >>10 dollars is small price to pay for documentation for a
    >>quality product when the alternative is $10K for Weblogic.

    I think we need to keep things in perspective - keep an eye on the real costs in a project. There is way too much fixation on license costs.

    Depending on the respective sizes of the development and deployment, even 10K/cpu is a relatively small cost. In general, the licenses costs make up a small proportion of the *total* costs of a project.

    1) Development costs FAR outweigh license costs.
    2) Support and maintenance represents 70-80% of the total cost of a project over its lifetime.

    Its been summed up well in an earlier post:
    Open Source software is good - so long as the value is greater than its cost. And the license cost is a small part of the total cost.

    -Nick
  77. Depending on the respective sizes of the development

    >and deployment, even 10K/cpu is a relatively small
    >cost. In general, the licenses costs make up a
    >small proportion of the *total* costs of a
    >project.

    I am in complete agreement here. Choosing the right app server also depends on the type of project.
  78. Peter: "Weblogic costs $10K per CPU and JBoss is free, but the extensive documentation for JBoss costs around 10 dollars."

    Just to be clear, the original post did have a grin attached. I was kidding with Rickard (who has my greatest respect) because there are a few people involved with JBoss that see the JBoss project as being ordained to destroy BEA (perhaps I overstate it, perhaps not). Personally, I feel that such a goal (or perception) taints an otherwise good project.

    Within reason we try to support open source efforts. We'll probably buy the JBoss manuals ;-). Regarding the original comment about the documentation, I do think that some of the open source mantras have seemingly blatant contradictions, particularly when it comes to what has "value" (what is OK to charge for) and what doesn't have "value".

    Peace,

    Cameron Purdy
    Tangosol, Inc.

  79. Well I cannot completely agree. While I have no problem
    paying a fair price for the JBoss doc's (The price is fine, and I have brought the basic 2.x book)

    I do take issue with the state of the JBoss docs. There is virtually NO doc on the various Jboss configuration and deployment files. I recently wasted days looking for info on configuring the server to run through firewalls with only partial success.

    Note to JBOSS folks: Thoughtly document your installation and configuration steps, and make money on the advanced docs (Clustering and CMP)

    You are driving folks like myself who WANT to use your server and buy your support as an alternative to outragous per CPU pricing models, but just carn't get enough confidence to make the leap.

    -- Success is about listening to customers....

    Rob
  80. i wonder why he thinks open source is eating up his revenues when he is hell-bent on giving up everything free.
    why not charge ten dollars for java - with 3 million java developers that is a sizeable annuity driven revenue. the open in "open source" if anything makes java developers productive
  81. check this out 1-800-Flowers.com - a higher share price than sun and a lower p/E ratio. just goes to show if u want to make money u can - maybe he needs a strategic advisor from flower shops unless of course its a worldcom story
  82. The higher share price and lower P/E say nothing about a firms relative economics. Last I checked BHS was running at $70,000/share. That doesn't mean anything. How many shares are there? What's the price to book ratio (that's the ratio that _really_ tells you if there is a value being created for shareholders). If their PE is lower then what's driving their returns, assets?

    PE doesn't really mean anything in your context either. Flowers probably does depend somewhat on real assets (ROA), and to some extent so does Sun. However, a lot of Sun's value driver to the share price is going to be driven by intangibles like intellectual property (software and hardware engineering IP). So _of course_ their share price will be driven up by these things.

    Flowers probably has little IP, and little intangible assets, therefore their price is going to be more driven by assets, receivables, sales.

    Plus there are about 1200 other things that speak to a firms finances, cost of debt, cost of equity, structure, debt to equity, etc, etc.

    I don't mean to sound harsh, but you need to learn a little more about finance before you spit this stuff out. Someone might listen to you.

    -Newt
  83. To Newt

    I may not be a finance xpert but finance xperts from cnbc have a story on returns from a flowers website as compared to returns from tech firms - u can check it out. Sun definitely has IP value but IP can only have value attached to it if it provides a return to investors
  84. wonder how those 'xperts' portfolios are currently doing.

    My dad says experts are 'drips under pressure'. (not real sure what that means but I have a guess)
  85. Is .NET actually going anywhere? I've seen the ads but know of no one using it. We bought a copy & I've messed with it some - 20 something CDs!. Reminds me too much of MFC and early DirectX - poorly thought-out & thus overly complicated. I can write Java much faster & end up with more robust code.

    To me it seems like MS tends to build hooks into non-public features of Windows. I think they do this to increase the performance of their code. However, it ends up also breaking encapsulation. My experiences with MFC & early DirectX were that things weren't very straight-forward and there are lots of little 'catches' to learn about. Java has been much cleaner.

    I don't doubt that Sun is finding business rough, particularly as compared to a couple of years ago. I think they've been a notably good corporate citizen - certainly compared to MS and Intel. Sadly that's not enough (e.g., Digital Equipment Corporation).
  86. Taken by itself, the price of a share of stock has nothing to do with the "market's valuation" of the company. You need to look at the market capitalization, which is the price per share times the number of shares issued. So (the imaginary) myknickknacks.com could be at $1000 per share, yet with only 5 shares issued, it's just a $5000 company. Sun at $4 a share is a multi-billion dollar company.

    Anyone who has ever eaten at White Castle or Krystal knows this lesson well. ;-)

    Peace,

    Cameron Purdy
    Tangosol, Inc.
  87. JBoss is what keeps j2ee going with small consultancies

    Otherwise MS would have eaten up that market long ago.

    I have been to small shops which show me the quotes at elance and tell me how they will survive.

    I mean it seems u can earn more by creating logos etc coz atleast u have a barrier to entry because u have to buy adobe photoshop etc.But with java being free....

    They are killing their own market and if jboss was not there these guys would have hopped onto .net bandwagon long ago
  88. "Taken by itself, the price of a share of stock has nothing to do with the "market's valuation" of the company."

    Yes Cameron, on the other hand in the last 12 months Sun's share price has gone from 16 to 4, and its current forward looking P/E is -21.6, I'd say Sun *is* in a weak economical position right now (especially as compared to Microsoft or IBM).
  89. Edgar,

    Microsoft's stock lost more than 30% of its value during last 12 month. With the current rate of Linux adoption, I'd suggest you be very cautious betting your money on MS.

    Read
    HP to work with US government on Linux clusters

    Big names bang Linux drum

    HP, IBM Notch New Linux Customer Wins

    BTW, congratulations on your recent appointment to Microsoft's Regional Director Community!

  90. Edgar: "Yes Cameron, on the other hand in the last 12 months Sun's share price has gone from 16 to 4, and its current forward looking P/E is -21.6, I'd say Sun *is* in a weak economical position right now (especially as compared to Microsoft or IBM)."

    First, I have no vested interest in any of the above mentioned companies. (At least not directly; I do have money in mutual funds, and I would guess that they hold stock in all of the above companies.)

    Second, Sun appeared to be over-valued as a result of the previous "tech boom" and their close association with the "dot com phenom". It's not surprising that they dropped 75% from 12 months ago. They are now in the same range that they were in 1998. Considering that 1998 was "boom" and 2002 is "bust", that isn't too bad (unless you bought Sun stock when it was $60/share).

    Third, Microsoft cannot lose money. According to the US federal court, Microsoft used illegal tactics to gain a monopoly and continues to use illegal tactics to maintain a monopoly. Very few companies and individuals have a choice but to buy Microsoft software and continuously upgrade it. Microsoft has more than tripled the cost of its OS and office software in the past decade, while most other software and all hardware has reduced significantly in cost. I think -- other than the US and EU legal actions -- Microsoft is a very safe investment, since they can charge whatever they want for their software (at least in the short-to-medium term) and basically everyone will still buy it. If you think you have a real choice in whether to buy Microsoft software or not, try communicating mark-ups with your legal department or another company's legal department without having Microsoft Word.

    Actually, in summary, it's pretty surprising that Microsoft is doing as poorly as they are. They have more cash than most countries. They have more leverage in the computer software industry than anyone. I believe that they now have more employees working on software than any other company, except maybe IBM. Personally, if I owned a major chunk of Microsoft stock, I'd be pretty disappointed in how terribly they are performing, particularly since they are constantly on the wrong side of the law and thus threatening to lose the greatest monopoly position in history that they could otherwise milk with impunity for perpetuity.

    Peace,

    Cameron Purdy
    Tangosol, Inc.
  91. That's not always the case. Last I looked Ms was hovering around $100 a share, with just about as many shares as SUN has out there. Maybe instead we should look at Worldcom whose stock is now .12 and is still considered a multimillion dollor company. A stocks price says alot about what people's perceptions are of a company and whether they are a ggod investment. I would consider buying WorldCom stock becuase I think they will recover in the future, but wouldn't buy SUN stock as long as Scott McNeally is at the helm. He is to busy leading a holy war against the evil MS empire, when he should be consentrating on adapting to the industry and building his own empire. If he did this, maybe sometime in the near future SUN would be a true competitor instead of just a basher of MS.
  92. Because all the other companies/organzations/individuals who contributed to Java will want their cut.
  93. <Q>
    Does this mean that the product is free but the documentation is not?
    </Q>

    No, it means that you have a CHOICE, you can either use the free documentation contributed by volunteers, or you can get documentation with a little more consistency and effort being put in for as little as $10.

    Which ever you wish to read is up to you. Or just read the source, that too works for some people.

    --
    Thomas
  94. <And by the way, access to BEA documentation is free ;-)>

    Now I ask you, what is the better deal? Paying a six figure yearly license fee for the software and get free docs. Or pay nothing for the software and getting a monthly subscription to docs for $10 per month.
  95. All the open source fanatics forget where Java comes from:
    A commercial company

    This company has been more than generous in even letting JBoss exist. Now, its like the child is trying to eat its mother -It will be interesting to see when (not if) SUN takes a harder line against Open Source -just like microsoft.

    The truth is that Open source is counter-productive to innovation, investment, and competition. You don't have to be a CEO to recognize this.
  96. The truth is that Open source is counter-productive

    >to innovation, investment, and competition. You
    >don't have to be a CEO to recognize this.

    Actually, there has been alot of innovation in open source and companies like IBM have figured out how to profit from open source and invest in open source projects like Linux and web services. I see alot of competiton because of the markets that are created due to open source development.

    Anyone who does not recognize this should not be a CEO.
  97. As I see it, then SUN has not "let JBoss exist" - instead
    JBoss has helped promote J2EE.

    There are plenty of opportunities to make money
    on J2EE - IBM and BEA are doing it. SUN has just not
    been capable of doing the same.

    That is not JBoss's fault. It is SUN's fault. I doubt
    that SUN will sell one single iPlanet Enterprise
    license if JBoss disappeared tomorrow.

    SUN will not make more Money by going after
    the Java open source community. They migth
    hurt Java real bad (in the extreme they may kill
    it totally).

    And I do not understand your disrespect for open
    source - I think Linux and Apache has proved that
    open source can be a success.
  98. <quote>
    The truth is that Open source is counter-productive to innovation, investment, and competition. You don't have to be a CEO to recognize this.
    </quote>

    I disagree 100% - Open Source *drives* innovation and competition and hence investment. Anyone who doesn't recognize that won't be a CEO for too much longer.

    Cheers
    Ray
  99. Listen Mommy, if it weren't for IBM and BEA, java would have died in your womb.
  100. Sun wouldn't be anywhere without Open Source. M$ or IBM would be handing them their lunch within the end of the decade without Open Source. Go study some UNIX history. While I've always liked Sun and thought they had really good tech, they suck at marketing software. The proprietary UNIX wars bled marketshare to M$, just as the proprietary Java wars will do for .NET. M$ is banking on this. The only way Sun has been able to hold the proprietary Java community together is by benevolent dictatorship. They haven't shit, so it's time to get off the pot. Where is our vibrant Java component market? The closest thing I've seen beyond a couple of widget sets and report generating suites is Open Source. All the benefits that Java has promised have come from Open Source, not the proprietary vendors. Take a look at XDoclet, Struts, JBoss and some of the other great Open Source projects out there and you'll see exactly what Java can be.

    Of course, why should I care what myopic, shortsighted individuals like you or McNealy think? Java is bigger than Sun and there's no way it will go down with them, we have the specs and now we have the source.

    Sun can't take a harder line against Open Source, they need Open Source, we don't need them. They'd be slitting their own throats. Sun hasn't kept M$ at bay or made IBM play nice. Open Source has.

    Back up your arguments, I doubt your anything more than a troll.
  101. Hi Rickard,

    <quote>
    ROTFLMAO! Let's start with the R&D point. Generalizing wildly there are three types of R&D folks: those in academia, those at corporations, and those in OpenSource. In my experience the most motivated and out-there ones are in OpenSource.
    </quote>

    The generalization you just made is really too wide. People doing CS research (in AT&T labs, IBM, Microsoft research, academia etc) generally have nothing to do with actual implementations - quite often they are not coders at all - they formulate and prove (or disprove) theorems and lemmas. This is a very expensive process, which requires corporate funding, research grants, etc.

    --
    Dimitri
  102. Dmitri,

    You said

    "The generalization you just made is really too wide. People doing CS research (in AT&T labs, IBM, Microsoft research, academia etc) generally have nothing to do with actual implementations - quite often they are not coders at all - they formulate and prove (or disprove) theorems and lemmas. This is a very expensive process, which requires corporate funding, research grants, etc. "

    I work primarily in research right now (Northwest Alliance for Computational Science and Engineering). I don't remotely get how you came to this conclusion. While it's definitely not the same as commercial development (did that for a while before research), we definitely have to deliver, or we stop getting money to do research.

    The difference really comes down to mentality and paradigm. In commercial the objective is to reach the goal through the shortest-distance/least-effort/most cost-effective. In research the objective is to come up with a new way to reach the goal. Both require consistent, strong delivery.

    Sure there are plenty of ivory tower CS people, but you'll find them as much (if not more) in commercial than in research. In commercial it's easy to be Ivory Tower, as so often the people around you don't understand. In research at universities and R&D companies like IBM you can't get away with that. Too many people with good understanding, and too much money on the line.

    I feel as much, if not more pressure to deliver than in commercial.

    -Newt
  103. Hi Newt,

    <quote>
    we definitely have to deliver, or we stop getting money to do research.
    </quote>

    Of course everybody has to deliver ;-) I meant to say that in case of research these deliverables are not nesessarily java code (or any code for that matter).

    --
    Dimitri

  104. The real issue is not .NET or Open Source. It's this:

    As the world evolves, the line of what individuals and organizations are actually willing to pay for rises. Look at the kazaa and all those other file sharing programs. I'm not saying they are right, but obviously the world's standards for what media and music is worth is lowering every day. Eventually, these industries will die or will have to give some compelling reason to buy.

    The same goes with Open source. With more projects and skilled developers making free software, our tolerance things on what we are willing to buy. JBoss vs. a $50k implementation of BEA WebLogic + addons? Give it 30-50 years, the concept of money just won't exist anymore. People will make things because the world needs then, not because we have these N-dimensional partner-consumer-supplier relationships that are practically meaningless if capitalism just didn't exist.

    What we see is the first of these movements. Organizations like Sun and Microsoft better start to see it too.
  105. The real issue is not .NET or Open Source.


    I agree with you here. There are companies out thare that are willing to pay Weblogic, IBM, and Oracle $50K for a fully supported application server. Sun should be making money in that market space. If they failed to compete there then blame their own marketing department. Sun need to point the finger at itself rather than make excuses.

    I have suspected for a long time that there is a deficit of brain power in Sun's marketing department and the fact that they have not capitalized on their own standards and technology is proof of that.
  106. ...their stock price is below $4.


    Sounds like a bargain to me. If you have money in an IRA account, perhaps you should snap some Sun stock up while it is still cheap -- Just my opinion that Sun is a strong company that will figure out what they have been doing wrong in terms of marketing and fix the problem.
  107. At the risk of going a little bit off-topic, I'd say there will always be a kind of market economy where limited resources exist, including expertise. Yes, we might get rid of limits and control on physical resources and production but expertise and knowledge are still limited by the number and quality of individuals who know how to do "stuff". Ultimately, I suspect (and hope) that my grandchildren's children will come to see the current state of the world as a kind of global poverty.

    I think that open source/free software provide the backend to organizations that are the first of their kind to have the capability for affecting large scale changes in their environment mostly based on their influence/clout. The normal reaction from individuals and organizations who see themselves at the top of food chain is first to protect the existing situation or, failing that, protect their places in a changing world. It seems to me that the flares from McNealy are an effort to reconcile Sun's inertia to these changes with an unforgiving reality that keeps them off their feet.

  108. Give it 30-50 years, the concept of money just won't exist anymore. People will make things because the world needs then...
    <
    ???
  109. Completely contradictory[ Go to top ]

    I find the comments extremely bizarre. Sun has promoted Open Source, in fact has made significant investments in Open Source projects:

    Here's just a few:

    OpenOffice
    NetBeans
    JXTA
    Jini
    Mozilla
    Tomcat
    Ant

    You can find it all at: http://www.sunsource.net/

    Now to say that they are losing business because of it is completely perplexing. I thought it was part of their strategy!

    I think the problem with Sun has been pure and simple lack of execution:

    What happened to the Chorus acquisition? It was supposed to be a state of the art Object-Oriented operating system. Don't see them in the embedded space.

    What happened to the Forte acquisition? It was supposed to be a state of the art distributed object systems for business applications. Anyone ever see the fruits of this acquisition?

    What happened to the Magc processor? It was supposed to be a high performance processor that could crunch java at blazing speeds.

    Bea, IBM, Oracle have just out performed java in the enterprise space, you can only blame oneself for this. There's just something about how things are done in Sun, great technology, unfortunately poor execution. What the hell is going on over there?







  110.   >>
     Give it 30-50 years, the concept of money just won't exist anymore. People will make things because the world needs then...
     <
    ???
    <
    And we will all live in perfect harmony ... .

    Too many futuristic movies?
  111. money just won't exist[ Go to top ]

    Give it 30-50 years, the concept of money just won't exist >anymore. People will make things because the world needs >then...


    Well in Star Trek the Federation doesn't have a concept of money, unfortunately, everybody seems to be employed by the Military!
  112. money just won't exist[ Go to top ]

    <>Well in Star Trek the Federation doesn't have a concept of money,<>

    The Ferengi did though. And did they. :) Hmmm.
  113. money just won't exist[ Go to top ]

    <quote>
    >Give it 30-50 years, the concept of money just won't exist >anymore. People will make things because the world needs >then...

    Well in Star Trek the Federation doesn't have a concept of money, unfortunately, everybody seems to be employed by the Military!
    </quote>

    Now now. In Star Trek the "Federation" has no concept of money, but yet they do business with many other species that do "Ferengi etc."

    An interesting world....

    You know. A group where everything is free manages to live in harmony with a group for whom money is everything...

    ....sound like a nice place for J2EE to be? Most definitely. Are we anywhere near? Nope. Too many people bickering with each other. Microsoft must love it! Just sit back and watch them beat the crap out of each other....

    Whether you agree with me or not doesn't really matter. What matters is this. J2EE needs to differentiate itself from .NET (which was really what McNealy was trying to say.)

    And I mean DIFFERENTIATE. I don't mean "our technology is better becasue of xyz." I mean differentiate in terms that the people at the top (the ones who don't really understand the technology in depth) will understand. How much cheaper is my project with J2EE than it would be with .NET? Those kind of questions. And vague waffly answers like "Well it's open source so you can switch vendors if you don't like the one you pick" etc. are a complete waste of peoples time. It is _very_ rare for a large company (which is where they need to target) to change vendor. And I mean _VERY_ rare.

    Bottom line. Get in on the ground floor with *real reasons* why J2EE is better than .NET and you are fairly well set. Mess around arguing over whether .NET is up to the job or not and MicroSoft will have done an end run around you and the games over. It's really very simple and I cannot for the life of me see why Sun can't spot that.

    Open source is good. So is commercial product. But for heavens sake when will Sun simply dig in and push this technology properly! And I mean to the customers and people not using it (the rest of us get it ALREADY!)

    Just my 2c.

    Chz

    Tony
  114. money just won't exist[ Go to top ]

    <tony>
    I mean differentiate in terms that the people at the top (the ones who don't really understand the technology in depth)
    </tony>

    Did it occur to you that it's precisely those people who really understand the technology in depth who should *not* be at the "top"?

    --
    Cedric
  115. money just won't exist[ Go to top ]

    <Cedric>
    Did it occur to you that it's precisely those people who really understand the technology in depth who should *not* be at the "top"?
    </Cedric>

    That is fine as long they defer technology decisions to those who do understand. But who is going to determine who knows? My wife pointed out that I've never worked\done work for anyone who does - at least that is how I come across (she thinks I have a problem with authority - I guess I should do what she says and not complain :) or is :( ).

    Ok, 'back on subject'. If those in the know are not deciding or 'controlling' or somewhere important in the decision making then they will end up trying to implement .Net on the Z/OS. Or a least supporting OSS in one area and bad mouthing it somewhere else. Unfortunately Sun is not the only one in this sinking boat (sinking boat == 'the insase are incharge of the asylum'). We can only bail so fast. Of course we and our buckets will get the blame - not the holes in the boat and those who picked the boat.


  116. "Did it occur to you that it's precisely those people who really understand the technology in depth who should *not* be at the 'top'?"

    Cedric, this sort of statement coming from one of the lead developers of WebLogic is rather sad. Just because some engineers get let into the big house to polish the silver (compared to toiling the cotton fields like most of us) doesn't mean they should sell out and buy into the whole "you are not worthy" line pure business people feed them. Sure, people who understand both tech and business are rare, but where they do exist they generally kick everybody's ass. Think Hewlett and Packard and more recently Bill Gates. Personally I think it's a hoot that Marc Fleury, a former Sun engineering peon is giving Scott McNealy a run for his money.
  117. Actually, from a management perspective Cedric is right in a way. A good manager doesn't do everything, but he does get everything done. That's an important distinction. The really good managers don't have to know all of the details, they do have to be good at taking recommendations from their people. That means listening, and surrounding yourself with good people.

    So many techies talk about how their manager doesn't understand them, or the technology, but that's not their job; that's your job. If a manager is sour, it's more likely because he/she doesn't listen, or ignores recommendations, doesn't protect his people enough from the higher-ups, is an ass-kisser, doesn't make expectations known in a swell though out manner, whatever. That's a separate discussion though.

    -Newt



  118. <jason>
    Actually, from a management perspective Cedric is right in a way. A good manager doesn't do everything, but he does get everything done. That's an important distinction.
    </jason>

    Right, but it goes further than that. Obviously, any competent individual gets their job done, whatever their job is.

    My previous statement was simply and humbly acknowledging that while I know a little bit about software engineering, I don't know much about running a company or selling a product.

    Accepting your limitations is the first step toward becoming better at what you don't know yet.

    --
    Cedric

  119. I think the point was that business is so dependant on technology that having someone with no technology knowledge making decisions on techology and even worse - in charge of a technology company - well is more than likely bound for failure (at some level).
  120. I think the point was that business is so

    >dependant on technology that having someone
    >with no technology knowledge making
    >decisions on techology and even worse - in
    >charge of a technology company - well is
    >more than likely bound for failure
    >(at some level).

    Never allow an idiot to be in a leadership position. The entire company will be destroyed as that person continues to mismanage and make poor decisions.
  121. Where [Enron] would [Tyco] you [Worldcom] get [Atari] such [Adelphia] an [Commodore] idea? (BTW way in case any of you haven't seen Kevin Nealon on SNL - the '[]' are subliminal messages).

    So how would you suggest not allowing that to happen?
  122. Actually, Atari has been re-invented, albeit as "only" a game publishing company :).
  123. So has Commadore - well the Amiga part they bought and then screwed up. Anyway, both are different companies using the name.
  124. This is true in an org with a pyramid management structure.

    I have worked on a team were the manager handled the money (mostly) and somewhat other things. We did everything else. It wasn't perfect but it definitely better than "manager A handing a task to manager B and manager B handing the task to team lead A and then ..." .

    You probably are thinking of a good leader. A "good manager" doesn't need to manage.

    Well, back to reality. :(
  125. This man is amusingly silly! He says:

    >I actually think we need more revenue in the J2EE space, so
    >that we can do more advertising to get the message out,
    >because right now the world is getting blitzed with Microsoft
    >advertising, and promotion and branding and propaganda, and
    >big lies, and that's why they're going, not because it's a
    >better product.

    Yet he seems to want to "Blitz" Open-Source in particular JBoss with their own advertising and as such get their PRODUCT not message out not on its merits but on marketing.

    I grow weary of Sun's crybaby antics, other people are content to try and make money from J2EE, which to Sun's credit is a very good concept and technology.
  126. Perhaps the way to for Sun to sell J2EE is to emulate the business model of IBM. They have IT departments that can build the entire solution for the customer -- and wouldn't you know, the solution involves WebSphere.

    The cost of outsourcing a solution has some tax appeal too like being able to depreciate all of the costs.
  127. <quote>
     Perhaps the way to for Sun to sell J2EE is to emulate the business model of IBM. They have IT departments that can build the entire solution for the customer -- and wouldn't you know, the solution involves WebSphere.
    </quote>

    Yes quite. Not just Websphere. But Domino, Tivoli, and a whole bunch of hardware, too. Sometimes they are even technology-agnostic. Often they make more money from the service agreements than the actual hardware and software. Certainly the add-ons. (cost of Websphere pales in comparison to the cost of some of its add ons!)

    I think this is one of the reasons that IBM is willing to support open source. They understand the infrastructure is pretty much becoming commodity and this means it's on a forever-downward cost spiral. They make money off the integration of, and the services that support, the infrastructure.

  128. Yes quite. Not just Websphere. But Domino, Tivoli ...


    In addition:

    I would say their say their solution involves at least one piece of their software. But it also does, sometimes, include software from 'the darkside'. They even have examples on how to do it (ie. ADO with DB2). That combination probably doesn't work as well as say, Java with their products, but they support customers wanting to do what they want (I reeeeally want to say '... do crazy and even stupid things').
  129. JBoss helps J2EE by generating additional interest in the Java/J2EE platform.

    Sun has been pretty active in the Open Source arena:

    1) Sun bought StarOffice. Soon afterwards, Sun created
        http://www.openoffice.org/ (open source office suite)

    2) Sun bought NetBeans. Soon afterward, Sun created
        http://www.netbeans.org/ (open source Java IDE)

    3) Sun donated the Tomcat servlet engine code
        to http://jakarta.apache.org/tomcat/

    4) Sun has engineers contributing code to
        Jakarta projects such as Jakarta Struts and
        Jakarta Commons. These Sun people include
        Craig R. McClanahan and Remy Maucherat.
  130. Thanks to SUN's Business Development Strategy, they(SUN) succeeded in taking themselves out of game and are hell-bent on taking down bright small companies along with them. The only choice they left for upstarts developing J2EE based products is to take the contrarian stance. "Create OPEN SOURCE product and pray for SUPPORT fees."

    The last time I saw they have 37 J2EE licensees, while there are more than thousand companies developing components for J2EE. Doesn't that number sound wrong to CEO. Is their licensing model working? Do you know how much microsoft charges for certifying/licensing .NET? You will be surprised.

    -Stevens
  131. McNealy's argument is that it is advertising that wins markets, not software quality alone. To win against .NET, J2EE needs strong advertising dollars to be spent over the next few years. For such money to be available, commercial J2EE vendors must be able to make sufficient margins. The presence of Open Source J2EE is preventing commercial J2EE vendors from making money, and thus starving J2EE of advertising dollars. Thus, Open Source will be responsible for the victory of .NET over J2EE.

    That argument sounds superficially convincing, but McNealy's assumption doesn't hold true with Open Source. Apache became (and remains) the world's leading webserver without a cent spent on advertising by anyone. Linux may now be promoted by companies like IBM, but it got to the stage where it was good enough to be championed, without either advertising dollars or research dollars being spent on it.

    Open Source is its own R&D and its own advertising. Applying the commercial model to Open Source is simply wrong. McNealy's point may have been relevant to the pre-Open Source world, but everything has changed now.

    The only way for J2EE to beat .NET is through its Open Source implementations, - JBoss today, and Jonas and OpenEJB tomorrow. My fear is that Sun will never understand this, and end up doing something really stupid, such as taking legal action against JBoss to stop its distribution. If that happens, it will be Sun, not Open Source, that will be responsible for the victory of .NET over J2EE.

    Ganesh Prasad
  132. Here I provide Sun CEO the strategy for free. If it works, please write me an email.

    Sun can not sell iPlanet Servers because users do not like it. iPlanet failed does not come from open source J2EE implementations. Why Weblogic was sold out pretty well.

    iPlanet dates back to Netscape Servers and it has many junky codes NOT for J2EE. I meant it is not tidy, compact, and easy to use as Weblogic or OC4J.

    Solution... is Sun need to dump iPlanet Server and buy Weblogic. If Sun can not buy Weblogic because BEA is huge now, Sun still can buy Orion Server and fill in the Webservice stacks.

    Sun still can start from fresh with tidy J2EE server and build up the Portal, Wireless, Intergration and Webservices stacks. In deed the money put on high level frameworks will be much more than the pure J2EE infrastructure. J2EE Server will become commodity stuff even it is very important.

    Best regards,
    Thanh
    thanhdoanchi@yahoo.com
  133. Point #1. Dot Net is better than J2EE the way Sun sells J2EE (read the next point)
    J2EE is sold as must have EJB for DAO.
    As a software engineer, we are to pick best:
    Simpler
    Cheaper
    Faster
    Easier to develop
    Easier to deploy
    More flexible (any sql statement (self join, correlated, outer) and multi row support)

    Only non software engineers strive for complex and expensive and hence slow.

    EJB is very expensive ($15K per license and up) and has bad phone support by vendors.
    Sun is at $4 now and it will go out of business the way Digital has because it does not know how to sell software as good as MS (and it HW is expensive and according to tpc.org slower than cheap Intel machines).

    Companies that did try to deploy EJB in production or maintain it get disgusted. Then comes a 3rd party solution vendor and says…. If you use the cheaper IIS and cheaper SQL ADO, you will develop faster and save money, here let me show you.

    It would be prefers for a manager to chose something slower and more expensive.

    Any version of EJB fairly quickly leads a customer to use .NET and webForm and ADO on their next project!
    I have lost many clients to .NET and I am flaming. Objectively, .NET is better and cheaper if you look at it from a software engineering point of view.
    Also C# and byte code is ECMA open standard.

    EJB is .NET’s best sales tool for converting clients over. Sun plays this in by saying one should use EJB with J2EE. OpenEJB is just as BAD.
    (see www.basebeans.com/bad.jsp for more on EJB)

    Lastly, it would be easy to write MVC in C#

    Point #2. Open source, like Jakrta (Tomcat, Struts, Standard Tags, etc) and SourceForge (basicPortal, Jasper, etc) and others (OpenOffice.org, PostgreSQL.org, Linuxes, Eclipse IDE) are better and cheaper and faster than dot Net.

    This to me means J2EE is great, most of it (Except EJB).

    Tomcat can handle a large load for free, PostgreSQL is very fast under very largeDB and is free, Struts is a great framework etc.
    And each open source has great support for FREE.

    Semi-free is Resin and dbExperts and they have good support.
    My experience is J2EE bad. .Net Good.
    Open Source BEST.

    We are trying to add value and save companies money 2 ways. Finding a way for them to use IT to save money and make money. And do that the fastest and cheapest way.

    If Sun deprecates EJB and standardizes on a DAO interfaces (that could use RowSet or JDO or any other implementation internally - sample DAO interfaces is at CVS of basicportal.sourceforge.net ) that would go a long way.
    Sun might keep giving away the JRE but charge some $ for SDK.
    Sun should stop selling HW or sell that to IBM.
    Sun should do more Linux, as it just announced (and get rid of SlowLaris).

    This sounds like a sales pitch but it’s not since it is open source.
    I tried to implement (and not just say) good practices in web app development to save money in basicPortal (MVC struts, standard tags, fast db access, DAO, event driven, easy).

    I mean…. How hard is it to write “Hello world” using Pet Cemetery? Vs WebForms in ASP?
    BasicPortal shows you how to do this better than .Net or PetStore.
    (Also, JavaServer Faces … if it does not make Struts the reference implementation will make a big mistake).

    Currently Sun makes money on J2EE licensing to BEA, etc. (BEA are losing market share to IBM.)

    Conclusion:
    Open Source is best.
    Sun will go out of business as all of its revenue streams dry up and because it does not know how to sell software, it competes with its 3rd party vendors (MS forbids competition with it’s resellers and consults).

    A good open source vendor is IBM which endorses Linux on Mainframes and elsewhere and it is able to sell Java.
    It will be MS(+Dell) vs IBM in the end.

    Recommendation:
    Please use more standard open source to save your company money and demonstrate it.
    Wasting company money on bad J2EE (EJB)

    Better than webForms is basicPortal and I will continue to improve it.

    I really wish best for Java, and best for it is open source.

    My 2 c.
    Vic












  134. sounds a lot like XDOCLET, which you can get for free:)
  135. Cedric
    >I don't understand this. Computer history in these past twenty
    >years proves that Eric Raymond's statement is simply not
    >happening (operating systems, app servers, end user
    >applications, etc... all these areas are dominated by closed
    >source solutions).
    >
    >Why would that change now?


    I think this is a very very good question. Loads of us will no doubt cry 'cos it works just as well and is free!'

    But that is techie wierdness. What of the business perspective where folks are scared of technology and going with the heard is the best choice you can make? i.e. big market share, well suported etc. If its a rubbish choice then you can point at the other cattle and say they made it too.

    My only real thought is one about age. i.e. the 20 somethings are getting heavily into OS. Even some of us 30 somethings. In another 10 years we will all be more senior and hopefully influence architectural and infrastructure decisions. And so certain proven OS infrastructures will make it.

    However. OS is not magic. The people behind OS will rarely give up their lives for a peice of software. They will move on, or decide money really is a good thing. eg Star Office is no longer free. Linux for most people costs £30 to £90 because they buy a box in a shop. I recon this will continue, the sucessful OS infrastructure projects will start to turn retail... because money is a good thing, and clients want support. eg apache stronghold, shareware stuff like paintshop pro.

    So does that leave us in a bad way? No, it's good. The prices are low (at least initially). Not all OS - some of the tiny stuff will remain free, eg ANT, struts and so on.

    Jonathan
  136. Copy and Paste from the LinuxWorldExpo website:
    http://www.linuxworldexpo.com/linuxworldexpo/v31/index.cvn?ID=10028&p_navid=2

    Scott McNealy, CEO and Chairman of Sun Microsystems, will share his ideas on Linux, and how Sun will participate in the open source community to ensure continued choice, innovation and compatibility, while still managing to make money in the process. Mr. McNealy will talk about new Sun products and services that will expand the company's commitment to supporting the Linux community. Mr. McNealy will also discuss the need for developers to write to a higher order platform like Sun ONE, as opposed to writing to a proprietary application binary interface, which he feels is "so last millennium."
  137. He is speaking to business not techies. As such he is seen as talking rubbish. i.e. talking to business is all about how you say stuff, and nothing to do with reality (loved the links). How many of us have been in sales meetings and had to keep ourselves silent as the sale folks go off on one.

    This isn't to say they are wrong to do so. Without sales people doing a number on clients there would be no money for the techies - even if your product is in its golden moment of brilliance it soon ages and then sales speak is really needed.

    So all this interview shows is an amazing mismatch in the speaker to the audience.

    The background of open source v's big tech companies is fascinating. And it is going to create far more tension over the next few years. The fundamental problem is that a group of 5 programmers (or 2) with no families (ie loads of time) can develop world class systems which wipe out mego co's prize baby. So open source can pull in 100 folks and achieve far more.

    I wonder if tech co's may respond by buying out the key movers. A short term solution that could work. Or sabotage. :) Pay one of the subscribers to an os project to inject loads of bugs... that can't be corp espionage as there is no corp.. That would be fun.

    Jonathan
  138. Business audience or technical audience, he still misses the mark. To the business person, which I am, his statement is EXTREMELY CONTRADICTORY as his company just annouced a FREE J2EE product a few weeks back. So in this case, free closed source, is better than free open source. One helps the revenue model and the other hurts it?

    I am confused....
  139. Well Ben,

    To clear up your confusion :), looking at my latest issue of Linux magazine Sun has taken out a fullpage advertisement on page 9 with the headline "Affectionately known as the one-two punch". Subtitle - "The Sun Fire 280R and BEA Weblogic Enterprise Platform. High performance and scalability that will KO the competition." Oh heh - there's a url at the bottom: www.sun.com/servers/entry/beapromo2

    cheers,
    Markus
  140. <quote>"At the end of the day, open source isn't hurting J2EE. Complexity is hurting J2EE." Jason Weiss</quote>

    J2EE is overly complex and fails to deliver full compatibility between Appservers. Problems such as Classloading hierarchies, differences in EAR deployment and self certification for each app server means you have to know about each appserver even when you write to the J2EE APIs and specs.

    <quote>"Sun can not sell iPlanet Servers because users do not like it." Thanh Doan</quote>

    Having worked with most appservers iPlanet is bottom of my list. It is difficult to install and unpleasant to use.

    Sun really need to do something more constructive than release APIs, they need to run software as a business that they value as much as their hardware ops.
  141. So, potentially you could make an argument that the open

    > source thing is just screwing up all the revenue models

    > the open source community is cutting the legs out from under
    > all the R&D and promotion efforts of all the open interface
    > strategies -- not open implementation, but open interface
    > strategies.

    Right - how must Craig McClanahan be feeling right now? His boss comes right out and says that his work is undermining his employer, that Tomcat undermines the Servlet and JSP specs!

    What about the NetBeans people - are they undermining Forte? Are the OpenOffice people undermining StarOffice? (Note that the OpenOffice community were _very_ pissed off last week when Sun 'announced' that they would be co-operating with Apple over a StarOffice port to OSX - without happening to check with the OpenOffice developers first... Proves they really understand open-source!)

    For a company with such strong ties with open-source efforts, it seems odd that he apparently has not read Eric Raymond's 'The Cathedral and the Bazaar', which states quite clearly that:

    "Perhaps in the end the open-source culture will triumph not because cooperation is morally right or software "hoarding" is morally wrong... [snip] but simply because the closed-source world cannot win an evolutionary arms race with open-source communities that can put orders of magnitude more skilled time into a problem."

    Raymond's point is that open-source, by its very nature will undermine closed-source. So it seems insane for a company that has put so much time and resources into open-source (and this seems to be increasing rather than decreasing) to claim that their work is being undermined by it.

    The only rational explanation is that McNealy differentiates in his own mind between the open-source projects that he believes that Sun controls (Tomcat, NetBeans, OpenOffice) and ones that they palpably don't (JBoss).

    McNealy's approach as a CEO has always been questionable (the ritual bad-mouthing of Microsoft has served more, IMHO, to undermine his credibility rather than MS's), but I really think he has reached the point where he doesn't have a clue how Sun can work with (non-Sun led) open-source, which (again IMHO) casts doubts over their relationship with the whole open-source community.

    /david
  142. <david>
    "Perhaps in the end the open-source culture will triumph not because cooperation is morally right or software "hoarding" is morally wrong... [snip] but simply because the closed-source world cannot win an evolutionary arms race with open-source communities that can put orders of magnitude more skilled time into a problem."

    Raymond's point is that open-source, by its very nature will undermine closed-source. So it seems insane for a company that has put so much time and resources into open-source (and this seems to be increasing rather than decreasing) to claim that their work is being undermined by it.
    </david>

    I don't understand this. Computer history in these past twenty years proves that Eric Raymond's statement is simply not happening (operating systems, app servers, end user applications, etc... all these areas are dominated by closed source solutions).

    Why would that change now?

    --
    Cedric


  143. "I don't understand this. Computer history in these past twenty years proves that Eric Raymond's statement is simply not happening (operating systems, app servers, end user applications, etc... all these areas are dominated by closed source solutions)."

    IMHO the situation is not so clear-cut. How many managers choose gcc, Apache, Linux, without thinking twice?

    Guglielmo
  144. Cedric,

    History is rarely a good economic indicator. I'm not necessarily agreeing with david _or_ you. I think the reality lies somewhere in the middle.

    However, just because something has been done one way for twenty years is often the reasons old economies and companies get left behind. So it's just not a compelling argument. There have to be compelling economic and technical reasons for people to use things, and this applies to open-source and closed.

    -Newt
  145. <Jason>
    History is rarely a good economic indicator. I'm not necessarily agreeing with david _or_ you. I think the reality lies somewhere in the middle.

    However, just because something has been done one way for twenty years is often the reasons old economies and companies get left behind. So it's just not a compelling argument. There have to be compelling economic and technical reasons for people to use things, and this applies to open-source and closed.
    </Jason>

    Note that I wasn't disagreeing with him either. I was just asking him why he thought things were going to be different now.

    I am definitely looking out for signs that the trend is changing, but so far, I haven't seen anything that would indicate that such a radical paradigm shift is about to occur.

    I would go one step further, actually: companies selling closed-source solutions have been showing a pretty good sense at creating incentive for customers to buy their product instead of using open-source ones.

    Open-source proponents often make the mistake to assume that innovation only happens in their world, but they vastly underestimate the power of innovation that upholds capitalism (note that I'm making the assumption that money can only be made with closed source solutions, which is what history has shown us so far).

    --
    Cedric
  146. <Cedric>
    I am definitely looking out for signs that the trend is changing, but so far, I haven't seen anything that would indicate that such a radical paradigm shift is about to occur.
    </Cedric>

    Perhaps not in the app server market. But who's to say it won't. Linux is making in-roads. IIRC, Yahoo uses BSD quite a bit. Apache is the de facto standard for UNIX web servers.
    Obviously, OSS can be relied on for mission-critical solutions. Why won't the app-server market be commodotized in a few years, like we servers are today?

    <Cedric>
    Open-source proponents often make the mistake to assume that innovation only happens in their world, but they vastly underestimate the power of innovation that upholds capitalism.
    </Cedric>

    I think some OSS advocates can come across this way. Obviously, ground breaking software can come from commercial, closed source endevours. However, it IS true that commercial software innovations are limited by the employees involved, no? OSS really has no limits on who can innovate.

    <Cedric>
    I'm making the assumption that money can only be made with closed source solutions, which is what history has shown us so far.
    </Cedric>

    Not true. There probably a FEW people out there making a living admin'ing Apache or Linux, right? Obviously, not much money will be made selling OSS (sort of the whole idea behind it). But there are other ways to make money in software.

    Besides, when people buy software - are they buying the features of the software, or the promise of support for the software? If BEA goes out of business tomorrow, do you think somebody would pay the same price for WebLogic as they would today, even though they are still getting the exact same "software"? People buy support/piece-of mind, NOT the software. This can happen with OSS, too.

    Ryan
  147. Hi Cedric,

    > Note that I wasn't disagreeing with him either. I was just
    > asking him why he thought things were going to be different
    > now.

    Actually, I wasn't intending to make a statement one way or the other about open-source. I was merely pointing out the inconsistency that anyone working with open-source projects should be aware of Raymond's article (given that it is often regarded as the manifesto of open-source), and so complaining that open-source is doing what Raymond said open-source would do is rather bizarre.

    > I was just asking him why he thought things were going to
    > be different now

    Infrastructure, mainly. A readily accessible internet has, and is, bringing more and more people to the open-source table (not just in the western world...).

    > Open-source proponents often make the mistake to assume
    > that innovation only happens in their world

    Interestingly, in Microsoft's Halloween Documents on Linux, they tried to claim that innovation only happens in the closed-source world. Raymond argues that it is individuals that innovate, not any management or development structure, which I cannot find any fault with.

    /david
  148. Sun's Sun CEO McNealy loses the plot[ Go to top ]

    "Computer history in these past twenty years proves that Eric Raymond's statement is simply not happening (operating systems, app servers, end user applications, etc... all these areas are dominated by closed source solutions).

    Why would that change now?

    --
    Cedric"

    What a braindead statement Cedric. Think for a minute. Things take time to evolve. The tools and infrastructure needed to propagate open-source development just wasn't there 20 years ago, not even 10 years ago, (not even 5?). The internet phenomenom is relatively new. SourceForge is new and invaluable to the community. It takes time how to figure out how to manage open-source efforts, since it takes much more than a good coder to have a successful open-source project.

    "dominated by closed source."

    Please...Your domination is a myth. I'm on my 4th client switching to JBoss from BEA.

    Bill
  149. Sun's Sun CEO McNealy loses the plot[ Go to top ]

    <bill>
    What a braindead statement Cedric.
    </bill>

    It wasn't a statement, it was a question (noticed the question mark?).

    Now that we have cleared this up, care to answer it?

    <bill>
    Please...Your domination is a myth.
    </bill>

    What the hell are you talking about?!?

    --
    Cedric

  150. This appears to be a rather short sighted statement. If Mr. McNeely and his counterparts could have found a way to agree to some sort of standard during the UNIX wars, would we even be having such a discussion about M$?

    I am having a very hard time understanding how open-source like JBoss is bad for the J2EE revenue stream, but a free Sun One is not? However, free poop is still free poop. Hence, why JBoss does so well and why Bluestone has gone away and Sun One will go the same route.

    If the Java world is going to take down .NET, a low cost J2EE application server is needed. Which is why I can not understand Mr. McNeely's comments. JBoss and other open source projects are lowering the cost to entry into J2EE. The reason .NET is here in the first place is because of the HUGE costs associated with J2EE from training, to app servers, to tools, etc. The more costs we can eliminate from this platform, the better chance there will be to take down .NET. The more developer centric tools there are the better we will be.

    Microsoft has a new license that is making a lot of their customers unhappy. That with low cost J2EE could be what the Java world needs.

    Maybe the answer for Sun and Mr. McNeely is to let go of Java. If you are not making money b/c there is no revenue model, let others figure it out. Do what you do best, make hardware and a good OS.

  151. did I missed something ? Sun One is open source ? McNealy need to review his definition of an open source project ...
    IMHO, open source community is one (perhaps the only) of the entities that can save J2EE from micro$oft commercial aggressivity.
  152. Pretty hot discussion !!!
    I am sure given a choice Scott will get back his word :-)

    Here are my views on this issue..
    Before that check out this resource
    www.sdmagazine.com/documents/s=746/sdm0003d/0003d.htm

    Note: This might ask you to register...

    1. Open Source serves a good marketing agent for J2EE technologies. Recent polls by Together Soft says "JBoss is the widely used development appserver". The total no of JBoss downloads assures the fact.
    2. Open source products are Innovative. ( though some products might be a simple copy & paste ).
    JBoss is the best appserver with complete MBean architecture.
    ANT is the best build tool in the market.
    Apache is the No-1 appserver.
    3.Open source products aids easy entry to the system.
    I think there are lot of such good things with Open Source..
    The side effect is Yes they eat the bottom line,they make the product a commodity.. If Open Source doesn't do it & then some other company might do it... So plz don't put the blame on Opensource..

    Areas where companies (especially SUN ) can make money..
    a) Consulting... Sun should be the leader in this.. For any microsoft related issue we go to "msdn" & for java related issues where do we go.. ??? javasoft !!!! NO way.. So SUN has to beef up its bottom line to cater to all this. It should make every one to be aware that they are technology experts....
    b) Add value to the tools..Not just Forte can be sold.. but ready made components,framework etc into the system & make people to buy that.. e.g. Struts, BC4J (ORACLE )etc
    c) Training.. I never see SUN doing a agressive work in this..
    d) Never recommend a single solution for every project.
    Come up with Business patterns so that people can choose their application architecture.. eg.
    Architecture - 1: JSP/Servlets -> JDO -> Database
    Architecture - 2: JSP/Servlets -> EJB -> Database
    Help customers to to decide upon the right architecture for their requirement & also help them to transition from one to other...
    e) Segmented solutions.. i mean assemble the stack & have different stack of solutions like
      Stack-1: Linux + Intel/AMD + Tomcat + Apache
      Stack-2: Solaris + Sparc + iplanet
    etc.. & provide a easy navigation path.. If Sun guarantees that it can move any existing application running in stack 1 to stack-2 in 2-weeks who will be not interested ???
    f) Be the leader in implementing your specs !!!
    g) Win the OpenSource support.. I believe the best company in the world is IBM. No company can play the card so well so that can be the buddy of "Microsoft" & "Open source " community at the same time :-)


    The Bottom line is
    Stop complaining.. Time to go back & think what needs to be done to prevent this downslide..Try to become a solution company rather than a H/W company...

    All the Best.. :-)
    ~Murali

  153. Should JBoss be any threat to the j2EE market? Am I crazy or?
    Surely for a large project nobody would think of any other app server than Weblogic/Websphere?

    For a small/middle solution there is Resin - no need for EJB

    What do you need JBoss for???

     
    Regards
    Rolf Tollerud
  154. What do you need JBoss for? I can think of lots of reasons...

    First, the obvious - it's free.

    Second, using open-source software gives you control. If there is a bug that you find, you can fix it. You don't have to wait for a vendor to release a patch.

    Third, what if you don't like Weblogic or Websphere? This gives you another choice. I (am forced to) use Websphere. I don't like it at all. I find a real pain in the ass for the most part to configure [and if you poke around this site, you will find I am NOT the only one :-)].

    Small- and middle-sized solutions can use EJBs, too. Nothing says that because you are not involved in a giant project you should have to write security and transaction control for your application from scrath. EJBs handle a lot of this nicely for you, so you can focus all coding your business requirements. JBoss fits quite nicely in this space.

    I have heard comments and seen benchmarks that claim JBoss does not perform as well as Weblogic or Websphere. You know what - so what!!! Is is fast enough is the question you need to ask. Most of time, the answer will be yes. If you need a solid app server and you don't want/can't spend $?K on one, why NOT use JBoss?

    Ryan

    P.S. Have YOU ever used JBoss?
  155. I think that McNealy and everybody else will gradually learn that people are sick and tired of paying for the same software over and over again.

    Sun should be smart and concentrate on building hardware and consulting services. Machines are expensive to build, and people are expensive to train, but infrastructure software eventually becomes free.

    The problem with Sun is not that it can't sell its software - it's that it needs to sell sparc software in order to keep its hardware market alive. But I think that fighting Intel is futile. Let's face it, the sparc market is not going anywhere. However, there's hope for everyone. Just look at IBM.

    The horse is dead. If McNealy understands that, I think he'll sleep better at night.

    Guglielmo
  156. Companies need this "Open Source" to sell a lot
    of their "Research Lab" products to the dev / user
    community. Makes perfect business sense as

    1. Get a back door entry to any organisation.
    2.
        a. Always puts the competitors off balance in terms of
           pricing ("someone else is offering this free!")
        b. confusing the dev/user community as to which way
           they should be going, thus putting pressure on
           the competitor in the same product group.
  157. I believe JBoss brings J2EE to everyone, not just Fortune 500 companies. There are things the BEA and others do better, so JBoss isn't always the answer, but the fact that you have choices makes J2EE very attractive over .NET .

    Just my opinion though.
  158. One Straight forward question !!!

    Is there any "One" Open Source product which possess a serious threat to any of the commercial product !!!

    I don't see any thing..
    Is JBoss a competitor to Weblogic or Websphere or Sun One server
    ANT a competitor to what ?
    Apache is not eating any one's market.. It has a niche segment..

    Except Linux i don't see there is any thing that competes head to head with many OS but not in all segments..
    In small mail,print,web servers.. linux is a serious threat to commercial OS'es. Big deal .

    Open source products creates new markets yes.. think of this. if you want to create a Enterprise solution & sell it some "5000" customers is it possible to do with commercial ones.. with weblogic or websphere !!!. Licensing cost will kill you !!!

    So the only alternate is to go for M$ solution..But JBoss,Tomcat & other tools help to go in for that..

    Stop whining .... Think smart :-)



  159. Oops...
    On one side the CEO pulls down the open source products & the other side the executive VP supports a Open Source Database...

    http://news.com.com/2008-1082-947510.html?tag=rc_spec


    Moral of the Story:

    Open Source is good as long it doesn't compete with our product !!!
  160. One correction...It's good as long as it's directed aginst MS :)

    Cheers,
    Elango
  161. "Apache is not eating any one's market.. It has a niche segment.."

    Somehow I would hesitate to define Apache's market as a "niche segment". Apache has around %50 of the market. To me that's not a niche segment.

    Apache is the default web server on unix systems. It used to have competitors, like Netscape server, but they are pretty much gone. The only major competitor left is IIS. As it happens, because of IIS' security holes that never seem to end, Gartner Group currently recommends that customers switch to Apache! If that happens, then Apache's "niche" will approach 100% of the market.

    Also, look at Linux. The last I hears, 10% or more of new servers (or was it 25%) ship with Linux. That's a sizable market.

    Look at GCC. Look at sendmail. Look at Perl, or PHP. Look at VI, or Emacs. Do I need to go on?

    Guglielmo
  162. The problem with OSS is that if it is extremenely successful (as the things you mentioned, Xerces, Xalan etc.) people after a while forget that something is an OSS -it just becomes a part of their life. They are estabilished, and by being free, wide-spread (=> good support) and good sw to boot, they ROI is hard to beat by commercial software, so no-one generally tries (people do try though - Sun with iWS, MS with IS).
    The ones that are in the process of estabilishing themselves (i.e. succeeding or failing) are the one that tend to attract attention.
    See things like JBoss, Mono etc.
    Regards,
    Vlad
  163. Please listen to interview on this very site with Marc Fleury, originator of JBoss, wherein he makes the convincing argument that the free open source JBoss is what is going to **save** J2EE against .NET. Emphasis on save is mine.
  164. "...Bill Gates may be sitting up there laughing his butt off because the open source community is cutting the legs out from under all the R&D and promotion efforts of all the open interface strategies..."

    Doubt Bill's got that much of a sense of humor, but who knows? They've been touting Open Source a dangerous cancer and un-American for quite some months now. Poor Sun, it's a tough world out there when you give your product away and still can't compete with BEA, IBM and MS--not to mention JBoss, which appears to be a loose federation of developers with no VC capital and no friends in the industry.

    So anybody know if JBoss is making any serious money? Seem to remember an old Marc Fleury interview on the JBoss site (since has come down) where he says that the JBoss developers are neither wealthy philanthropists nor rabid communists. So how can they afford to be such a thorn in the industry's side?
  165. McNealy must understand three facts. If he can, he can build a business model around Java.

    1. The OS is a commodity. In his case, Java is now the OS. Forget Solaris, the OS for most apps is now, or will soon be, Java or .Net. The OS has been built (J2EE.) It time to really get the tools and apps market going.

    2. Linux is grabbing marketshare so fast into his market he does not even realize it is in the door. Read the news. Companies are switching critical functions from "UNIX" to Linux and getting not only money saving, but also claimed major speed increases. You don't even need to read the tech publications to see this; USA Today had an article today (on their web site at least) today about it. It's time to support Linux on real low to midrange servers, or back up the reasons to use Solaris instead. If Sun offered Linux on their low end (64 bit servers) at a price comparible to their real competition on the low end (Dell/Compaq) they would fly off the shelves. UNIX no longer has the advantage it has enjoyed for years. UNIX is now on the Intel platform. Wake up! Smell the last 12 months! IBM did.

    3. SunOne App Server (iPlanet) sucks. Just read the posts here. People hate it. Get a clue. (I myself have never used it.) If you can't give it away, that should be a hint. My suggestion? Buy out the JBoss Group or Orion, or at least buy themselves commit access to the JBoss cvs. Sell Sun support and "solutions" for JBoss. Compete with IBM...Sun, it's your game, twist the rules a little...make IBM swirm. The money is now in services and applications (IBM gets it.)

    If McNealy can graps those, I think they can find a market to make money in again. As a Java web developer who deploys on Linux (Resin Servler container), I am worried Sun is driving the platform into the ground. We have never given Sun a penny. They don't have a product that we feel is worth paying for. Charge for Java? That's a sure way to shrink the userbase. Their app server? It sucks. Tools? Have you used Forte? I tried it a few months ago and am still waiting for the interface to catch up with what I clicked on a couple months ago. (ok, it's not quite that slow.) I use IDEA. IDEA is fast and I am more productive with it. Beans for either my JSP's or and app server? Do they sell any useful Beans? Your guess is as good as mine. Servers? We have 3 Linux admins. Oh, they run Solaris? Nevermind, we can't support that with staff on hand. Hold on, we just hired an intern...4 linux admins.

    The market is changing. New "UNIX" people are growing up in the lower end world of Linux, not commercial UNIX. I was playing with Linux when I was in school. I use it in my job today and have for the past 4 years. I am sure I am not alone. The first app I need an EJB app server for I will probably develop on JBoss. I have experimented with it before and know how to use it. If it can scale to what I need it for, I will use it. What reasons do I have to try iPlanet even if it is free? Everyone says it is horrible.

    Sun are you listening? There is a whole market willing to pay you if you can provide us value for the money we spend beyond what we currently get with Intel/Linux/Java. From a "low end" standpoint, Sun does everything they can to keep us from having a reason to be their customer.

    I have ranted long enough. I'd love to pay sun for something. They have nothing we have any reason to purchase. I don't even know if I want to buy their stock for the current > $4 price.

    -Pete
  166. <I don't even know if I want to buy their stock for the current $4 price.>

    Hmmm. Perhaps if the Sun stock price goes a little lower, the JBoss Group can buy Sun and the Open Source problem will be solved !!
  167. I think Scott McNealy might have a problem doing this. I vaguely remember his comment about two years back that "SW is not important. Anyone can do SW. It's what it runs on that counts." (not a verbatim quote, couldn't find it anywhere).
    He changed his tack about a year back, that SW matters. Unfortunately, to move even further, to services, is another quantum leap which at the current rate we can expect in about 5 years.
    IMB managed to to it under Lou Gerstner, who rebuilt it from scratch and changed its culture considerably. Can McNealy do it? Personally, I doubt.
  168. Hi,
  169. Bye,


    (Was that the right response? Did I win anything?)
  170. McNealy = Gates... different circus... same clowns...
  171. Difference is Bill is smart enough to buld a muti-billion dollar industry and knows that marketing his companies products is far more productive than attacking his chief competitor. Bill Gates also has the smarts to make sure his company changes with the times. You can see this in how VS eveolved from a desktop application language to a desktop / web applications language, Xbox and games, etc. The company that can go with the flow and make the necessary changes to the timew will always succeed. In the same way that the football team that is able to make adjustments at half and implement those adjustments ususally end up winning.

    If everyone who hates MS so much was so smart why aren't they out there starting their own companies and turning them into the next big business success. Bill Gates is now clown and he turned a hobby he loved into the biggest software company in the world. I'd say the only clown is the one at SUN who would rather attack not only his competition, but also those who have helped keep SUN and Java alive.
  172. Has McNealy forgotten that open source Jakarta Tomcat is the Sun's official reference implementation for Servlets and JSP ???
    Why can't JBoss be the reference for EJBs ?

  173.  Give Scott a break, he's probably getting a lot
    of pressure from Sun's Board. If I was a Sun
    shareholder I'd also be evaluating the return
    on investment that Sun is getting from J2EE,
    and would probably considering selling them...

     The Open Source evangelists often forget a few things :

    - The biggest chunk of the economies attributable
     to open source are going to corporations,
     programmers get the bread crumbs (consulting
     fees and what not). I think of the OSS freaks as
     the Robin Hoods of the Corporations !
     And it makes me laugh to here the 'save the
     world from poorly written software' speeches !

    - Most talented programmers that make valuable
     contributions to OpenSource code have a day
     job at a company that sells, profits, or
     benefit in some way from 'closed' software.

     Now don't get me wrong, I think there's some great
    pieces of code that come out of open source, I love
    and use emacs, posgres, and linux extensively.

     I consider myself as a selfish oportunist,
    I love to use open source software, but will not
    code for free, If every one was like me, there would
    be no open source !

     I think the average Sun share holder
    (the average share holder of any company for that matter)
    Care's much more about the companies profit
    than the contributions to 'good software'.

    Sooner or later Scott will have to communicate
    the the 'Jave Community', and we're just seing the
    begining...
  174. I consider myself as a selfish

    >oportunist, I love to use open
    >source software, but will not
    >code for free, If every one was
    >like me, there would be no open source !

    You still don't get it. The problem is not open source software. The problem is Sun's marketing department's failiure to produce products that people want. Other companies made a tremendous amount of money on J2EE while Sun failied. IBM and BEA also had to compete with open source software, yet they made huge profits. This is proof that open source is not the problem.

    Wake up and stop putting the blame on others.
  175. As reported in the link, the California legislature is considering making a law that mandates the use of Open Source software in government offices over proprietary software from vendors. Yahoo !

                                       http://www.infoworld.com/articles/hn/xml/02/08/09/020809hnrally.xml
  176. Someone who plays golf that well can be excused not to understand his own business too well, right?

    http://www.miami.com/mld/miami/business/3819486.htm

    Cheers,

    Jill.

  177. Very funny, Scott McNealy is the living embodiment of incompetent pointy-haired-management.
  178. Marc Fleury, founder of JBoss, answers Scott McNealy

    http://www.jboss.org/mcnealy.jsp
  179. claiming that we already have Sun One


    Do we already have Sun One? Where?

    --
    jonmartin.solaas@mail.link.no
  180. I do not think that Scott actually said all of that (At least, I hope not). Sun is actually funding some of the Open Source efforts, so it would be very sad if their leader do not get the value of Open Source and how it can help counter MS dominance. At least IBM, BEA, Oracle, ... seems to get it ;). Sun should focus more on their business opportunities & customers rather than on their competitors. Their religion and belief has begun (for a while now) to completely blind them from the real world.