Opinion: Sun / JBoss Cold War Heats Up

Discussions

News: Opinion: Sun / JBoss Cold War Heats Up

  1. Opinion: Sun / JBoss Cold War Heats Up (66 messages)

    Alex Rupp discusses the rift between the JBoss and Sun camps. Will Sun allow JBoss into the TCK game? (given that they are not a "qualified" non-profit), will they try to sue JBoss to bleed their finances?, and would JBoss developers care about CORBA support?

    The last line of the article also catches the eye:

    "So, while reading Fleury's rhetoric, we ask you to remember one thing—this is a cold war for cold hard cash. Open source has nothing to do with it"

    Read: Sun / JBoss Cold War heats up as JavaOne conference nears

    What are your thoughts on the issues?

    Threaded Messages (66)

  2. From the article: "Suppose [JBoss does] implement CORBA support and complete the TCK. Name one other server that will be J2EE and .NET compliant in 2004. We can't."

    So I'm reading into this statement that they believe JBoss is pursuing .NET compliance as well. Has then been mentioned in any other article, or any reliable source? It seems to me that keeping up with the J2EE spec, innovating on top of J2EE, and wrestling with Sun is a large enough effort without jumping into the .NET fray, fighting a compatability war that is not winnable, and wrestling with Microsoft.

    The difference is that Sun has to keep the J2EE spec manageable, at a rate that app server vendors can keep up with it, and thus, so can JBoss. With .NET, Microsoft can turn on a dime and introduce waves of new features with no advance notice. Microsoft isn't concerned about Mono because Mono's existence depends on some public specs and a bunch of gentlemen's agreements for Microsoft not to sue for infringement. Microsoft likely thinks that they can always keep Mono one step behind, so it won't really hurt their bottom line, and if it ever does, they have a handful of lawsuits waiting in the wings. Meanwhile, the existence of Mono is a PR boon to Microsoft, as they can pretend to be community-friendly. Furthermore, ANY Microsoft competitor will face the tools challenge - the advantage of .NET over J2EE isn't the runtime environment, it's the tools, such as Visual Studio. Unless you want to develop a VS.NET clone as well, you have no chance going toe to toe with MS.
  3. Good one[ Go to top ]

    Meanwhile, the existence of Mono is a PR boon to Microsoft, as they can pretend to be community-friendly


    That was a good one.
  4. I'd like to refute some of the claims of this article:

    Even if Sun did place the TCK under the LGPL, the JBoss developers probably wouldn't work on it, or at least not on all of it.

    This is total bullshit. This is what Sun and other JBoss competitors want you to think. Its just not true. Rick Saletta from Sun, publicly stated two separate spec compliance problems JBoss had. Within 2 hours of knowing this, these problems were fixed and committed into the JBoss core. Our developers are itching to make JBoss totally, absolutely compliant. Reason? We want to be the best. Period. This isn't a hobby for us, this is a mission.

    CORBA support would be the first thing to go (and no one would ever notice).

    I am so sick of hearing this. This is FUD spread by individuals who haven't even gone to our site. We have had CORBA support for over a year. JBoss CORBA support has been actively supported by one of our core developers, Francisco Reverbel.

    JBoss.org has created a public forum dedicated to J2EE compliance.

    Best Regards,

    Bill Burke
  5. I feel like Sun Microsystems is using the same technique that used by Micorsoft all the time. "Java" will become more than a trademark of Sun Microsystems. I hope Sun will not kill "Java" itself.
  6. Hopefully JBoss will win this "Cold War". JBoss is more developer friendly than anybody else out there. Sun is failing us developers in favor of big business.

    Sun can't seem to make up its mind on open source. On one end they seem to see it as a way to combat MS and on the other end they don't want anybody else doing it except for them. You can't have your cake and eat it too. Sun should embrace open source with its arms wide open, that is the only way they will win developer support. I will have to say I won't recommend any of suns products to my clients while this childish and immature behaviour continues.

    tx

    Matt
  7. True enough... That's would not be good for us :)

    --
    David Smoth
    dave at prokmu dot com
    Jsp Hosting PROKMU
  8. TSS / JBoss FAQ[ Go to top ]

    Q: I think Marc Fleury should turn over the JBoss project to a third-party.

    A: Uh, why?

    Q: Because he says he loves open source. If he really loves it, he should trash the company he built up for years and give up all control of the project.

    A: I love rollerblading, but I didn't stop working so I could do it all day.

    Q: I found a shocking secret about the JBoss Group. They aren't working on JBoss to promote world peace; they actually want to MAKE MONEY.

    A: Yeah. Hold the phones.

    Q: That is so slimy! If I had the courage to start my own business around a LGPL codebase, I would never try to make money off of it!

    A: Yes, well, we'll never know, will we?

    Q: I don't trust JBoss!

    A: What's not to trust? If you like their server, you are free to download it and use it as you please.

    Q: Yes, but as soon as I do that they will claim the code is no longer open source, and try to charge me license fees!

    A: Well, actually they can't do that under the LGPL license.

    Q: But what if they do? I'm SCARED!

    A: Don't be scared, sweetie. It's OK.

    Q: I don't like Marc Fleury personally.

    A: OK.

    Q: I refuse to use the JBoss server because I don't like Marc.

    A: That works out nicely. There are a lot of other Java application servers to choose from.

    Q: I hate JBoss because if I were running JBoss Group, I would do things differently.

    A: Have a party. Fork the JBoss source, start your own company, and do things the way you would like. It's perfectly legal.
  9. TSS / JBoss FAQ[ Go to top ]

    Congratulations Corby on your Q&A session! It was funny. But I doubt it quiets down the "scared" people who don't have the guts to do anything on their own (+ Gunter / Tollerud troll division). ;-)

    /T
  10. I'd like to focus on some of the details skimmed over in the following quote.

    "I am so sick of hearing this. This is FUD spread by individuals who haven't even gone to our site."

    Didn't (s)he link to your site? I saw two links. WHITE only recently came out, which suggests that this response was fairly prompt on their reply. I'm only guessing, but this person frequents your homepage (god help them, the place comes off like a porn site).

    "We have had CORBA support for over a year. JBoss CORBA support has been actively supported by one of our core developers, Francisco Reverbel."

    We appreciate that detail, but you know very well there's a HUGE difference between CORBA support and CORBA compliance. It's obvious in the paper that he was talking about compliance. Of course there's CORBA support, it's been there forever. But there is no CORBA compliance.

    I don't think whoever wrote this is trying to spread Fear, Uncertainty or Doubt. What, are you going to tell us that Sun is really in control of the Inquirer? Last I checked they wanted Magee's head on a pike, so don't tell me there's some conspiracy going on over there. This person obviously did their homework or has been observing for quite some time.

    And if that's the only thing you can dispute about the article, we're left to assume the rest is true?

    --Will Tabat
  11. Not impressing[ Go to top ]

    As far as I am concerned, then unless someone specifically state
    how JBoss is not CORBA comliant, then I find that claim strongly
    suspicious.

    This is supposed to be a technical forum where people
    know some of all the acronyms, so let us get a few on the
    table.
  12. Red Herring[ Go to top ]

    The corba thing is a red herring. It is one technical point in a very long article. Where is the response from JBoss on the rest of the article?

    I followed the link Bill Burke posted and found these two points on the page:

    * Currently JBoss/IIOP only provides basic IIOP access to enterprise beans. Transactions and **security** are not supported yet.

    * Status: JBoss/IIOP is currently in alpha state.

    So I cracked open the EJB 2.0 spec and found the following:

    19.6 Transaction Interoperability (optional)

    19.8 Security Interoperability (NOT optional)

    So JBoss definitely does not have interoperability and is definitely not compliant in the Corba area, but who cares? Last time I saw corba was the early 90's.

    I'd like to see a response to the whole article instead of this clever pick at the technical points PR tactic.
  13. About Norman Alex Rupp[ Go to top ]

    He is indeed no stranger to Jboss site or to its internals. As a matter of fact he uses JBoss for his open source project NewsFighter(he is the founder, lead developer). He works closely with Jboss members. So, saying that he doesn't know anything about JBoss or is spreading FUD is kinda illogical at this point. Go read, if you want to see how Pro-Jboss he is..(as a matter of fact , he doesn't even bother to develop for any other application server, because he thinks the rest don't push the limits.)

    http://sourceforge.net/mailarchive/forum.php?thread_id=1869247&forum_id=32023

    this is the NewsFighter site.
  14. About Norman Alex Rupp[ Go to top ]

    Below are his thoughts on webwork/xwork. Who would actually work on a project with this guy as the lead. Must be some desparate people out there.




    "Thanks for your input on the forking issue. We're not going to fork their
      entire codebase, but we are going to steal a *bit* of their code, and I am
      going to try to get in as a core developer on their project. I don't know
      if it'll happen yet, but I've inquired about it. Apparently they're
    rolling
      out a version 2 pretty soon. Right now they're doing documentation.

      That's like cooking for six hours and then doing all the dishes at once.
      Yuck.

      Shocks is tied in with JBoss and also with Webwork, but neither of them
      should rely on Shocks. That's where we draw the line. I don't know that
      our JMX additions really belong with WW but if they let me help add
    comments
      to their code and clean up some minor nicks and burrs I'm sure it'll
    help us
      both out. We can let them keep driving WW and we'll pick up where they
      leave off. ; )"

    "XWork adds interceptors (good idea) to the framework. It also breaks the
      Action model out of WW (so it can be used in non-web-based
    applications) and
      WW2 (forgive the abbreviation--I didn't name the damn thing) adds
    lifecycle
      support and interceptors to WW. We're going to use the JBoss service
      lifecycles, so this doesn't buy us much. Neither of the two forks leaves
      much breathing room for our JMX additions.

      And they've definitively said "no" to my ideas about JBoss management
      ties--they're standards compliant and won't tie in with JBoss, which is
      fine. I see their work moving unneccessarily away from the service model.
      They've got good ideas, but I'm not convinced yet that they're good for us
      or our developers or users.

      Not that some of their ideas aren't pretty cool. They're going to provide
      XML configuration support now. We won't need that for our ManagedAction
      configurations, though. They're also doing something similar to a
      DynamicAction (think DynamicMBean), but our model circumvents this by
    taking
      it to the service level.

      So. . . .

      I might pick up their 1.3 release and run with it on Shocks, just wish
    them
      well and consider future versions of their products. XWork and WW2 are
      nowhere near ready for us to use, and I don't want to tie us up for six
      months.

      Plus, they still don't document their code!
    "

    " Webwork has its serious ups and downs. On the plus side, it has a good
      architecture.
      On the negative side, NOTHING IS DOCUMENTED. It's beginning to wear me
      down. I might have to actually work on Newsfighter tomorrow, just to
      maintain my
      sanity.

      Believe me, I'm looking forward to actually working on *our* project
    again.

      I spent the evening dissecting the dispatcher package and adding
    much-needed
      API
      comments, tightening the access modifiers for fields and the like. My
      thoughts
      after tonight's coding session are that good design does NOT mean good
      implementation. I have no intention of completely redoing WW--not that we
      couldn't,
      but it's such an unbelievably complex pain in the neck. We might need to
      completely
      redo it, though. The thought does not excite me.

      I mean, we're going to have to implement Actions and they don't even
    have a
      DTD for
      their Action.xml file available. Their XMLActionConfiguration needs a
      complete
      overhaul and has *zero* lines of API documentation. It's unbelievable how
      little
      documentation there is on this. I can't expect any of you to do the
    digging
      I'm doing
      just so that you understand how to write an Action. It's ridiculous, and
      it's sucking me
      in. Almost makes me wish I'd extended Struts, but I know that'd be worse
      : )

      So I'm going to continue working my way through it, adding API docs as
    I go.
      If you
      don't hear from me for a few days, you'll know why. Also, be sure to
    update
      the
      shocks module if you want to see what's going on.
    "
  15. About Norman Alex Rupp[ Go to top ]

    Well, you might be right about his personality.
    However, we are talking about his relationship with JBoss not how he arrogantly criticized WebWork.
    We can't bark at everybody who expresses a different point of view. This attitude always reminds me of the republicans. No matter, what you say, they call you "Unpatriotic". So, hearing FUD on everything that is written against JBoss is getting very tiring personally. I want to hear everybody's point of view. JBoss has an excellent code base. But, what is the average company Joe is thinking? That's what i want to know. So, you guys, stop bashing everybody who utters even a word against JBoss. All opinions are welcomed.. It is a boring world when everybody agrees or forced to agree, or we are afraid to express what we really think because of the bullies like you..
    And please, please, stop saying FUD!! We are all so sick of this word being used, whenever there is not a logical argument about the subject..Be factual, and express yourself clearly, why you are against and for, without calling it FUD..
    Thank you.
  16. About Norman Alex Rupp[ Go to top ]

    <quote> Well, you might be right about his personality.

    > However, we are talking about his relationship with JBoss not how he arrogantly criticized WebWork.</quote>
    I have seen no evidence that he is associated with JBoss. Where did you get that? I would like to know.

    > <quote>We can't bark at everybody who expresses a different point of view. This attitude always reminds me of the republicans. No matter, what you say, they call you "Unpatriotic". So, hearing FUD on everything that is written against JBoss is getting very tiring personally. I want to hear everybody's point of view. JBoss has an excellent code base. But, what is the average company Joe is thinking? That's what i want to know.</quote>

    As far as what i think as an average developer. Here it is. JBoss lets me develop my applications according to J2EE. We enjoy the developer friendly features that JBoss supports. Their new IDE is icing on the cake. With XDoclet we are able to port are applications to other app servers with ease. We give our clients options as to how our applications run for them. Some of our clients strictly use Oracle so we port our app to Oracle and they pick up the bill and are very satisfied. Other clients take our advice and run our stuff on JBoss saving them money. We try not to push other vendors besides Oracle, BEA, IBM, and of course JBoss. We have not been impressed with Sun's app server offerings'. Using JBoss significantly lowers the cost of our products and wins us more contracts. These are my own words.



     <quote>So, you guys, stop bashing everybody who utters even a word against JBoss. All opinions are welcomed.. It is a boring world when everybody agrees or forced to agree, or we are afraid to express what we really think because of the bullies like you.. </quote>

    I wasn't barking, I was just letting the guy use his own words to discredit himself and he seems to do a good job of that.

    ><quote> And please, please, stop saying FUD!! We are all so sick of this word being used, whenever there is not a logical argument about the subject..Be factual, and express yourself clearly, why you are against and for, without calling it FUD..
    > Thank you.</quote>

    As far as my other post, it was about FUD. To say JBoss has not contributed to the JCP and should be criticized for it is most definitely FUD. Do you like the words Uniformed or Untrue better than Fear Uncertainty and Doubt? Should I be quiet win someone is outright lying. Logical argument. How is that not a logical argument. He said one thing, I gave references to show logically he was spreading FUD.


    by the way

    I am a republican.


    tx

    Matt
  17. What's wrong with that?[ Go to top ]

    <quote>...we are talking about his relationship with JBoss not how he arrogantly criticized WebWork.</quote>

    What's wrong with that? I gripe about technology all the time with my friends at work. Why are you changing the subject? I spent about ten minutes looking at his list and I don't see a problem. He doesn't like how webwork's API isn't documented. He's forking it because they're moving on to version 2, 1.3 can't handle JMX and they won't let him tie it in with JBossMX. Seems reasonable enough to me.

    I guess if I were a volunteer project manager and I needed to convince my team to rebuild an entire MVC system I'd probably talk tough too. It's a small team anyway, less than two months old so who cares?

    Besides, I was way more interested in his predictions for the future of the JBoss Group. I notice nobody's talking about that. It's much easier to kill the messenger, right?

    Will Tabat
  18. What's wrong with that?[ Go to top ]

    William,
    You are absolutely right about your statement. I didn't have to call it arrogant. And that wasn't the point of my post. I have been following WebWork for a while and I really like it. I didn't appreciate his comments against the WebWork team and Rickard, that's all. These guys did a brilliant job and really working hard to get it to completion with documentation and all.

    My intention was to point out that he wasn't a stranger to JBoss as people claimed. He does seem to work with Dain and is very pro JBoss.

    I only wanted to point that out. If it came out differently, it is my bad.
  19. About Norman Alex Rupp[ Go to top ]

    I don't think he's arrogantly criticizing webwork. He's going out of his way to compliment it, even when it doesn't quite fit his needs. Is he arrogant because his project isn't 100% suited for webwork? That seems to be an arrogant supposition, if you ask me; arrogant in that it assumes that webwork is the be-all, end-all of web frameworks. I don't think "they have a lot of cool ideas" is really cutting on the webwork team, is it?

    Besides, some of his criticisms of webwork are dead on. I love webwork and am in the process of migrating several struts projects over to it. But he's right that the documentation isn't as good as it could be. It's getting better but it's still a bit thin, which is unfortunate. I know that the team recognizes this because they keeps saying how high a priority it is to improve the documentation. But these people apparently have day jobs, so it's not going to happen overnight.

    As for the "cooking for six hours and then washing the dishes" comment about saving documentation for last, I agree 100%. Mostly because I've been burned by that myself when saving my own documentation until the end. Or my own dishes.
  20. transaction interoperability[ Go to top ]

    A point on EJB: the requirements for transaction interoperability aren't exactly optional; you may omit support for context propagation, but you still need to properly encode IORs to be compliant. In other words, you have to provide a signal to the other server environment that you can't deal with the infection of the transaction context they have created. This is actually important if, for example, you expect work on a remote bean with a Required transaction demarcation to be included in a single unit of work, rather than in two transactions, one bounded in execution by the other.
  21. Open source != evolution[ Go to top ]

    "Evolution cannot be stopped, and nobody pushes the evolution of software better or faster than the Open Source Movement."

    This myth has to be de-bunked...

    JBOSS dilutes the standards and pollutes the J2EE market since it doesn't contribute to the standards process (JBOSS Group, nor Mark Fleury, nor Bill Burke are members of the JCP) but they [pretend to]use the standards to compete with the companies that invest in the development of the standards.

    Matt
  22. Open source != evolution[ Go to top ]

    "Evolution cannot be stopped, and nobody pushes the evolution of software better or faster than the Open Source Movement."

    >
    > This myth has to be de-bunked...
    >
    > JBOSS dilutes the standards and pollutes the J2EE market since it doesn't contribute to the standards process (JBOSS Group, nor Mark Fleury, nor Bill Burke are members of the JCP) but they [pretend to]use the standards to compete with the companies that invest in the development of the standards.
    >
    > Matt

    pure FUD

    some of JBoss Group are on the JSR committees, take a look at JMX for instance. they piooneered much of that work. It would be nice if you actually knew what you were talking about, but then again all you are spreading is FUD.

    http://www.jcp.org/en/participation/members/L
    L is for Lindfors i.e. juha

    tx

    Matt
  23. Open source != evolution[ Go to top ]

    "Evolution cannot be stopped, and nobody pushes the evolution of software better or faster than the Open Source Movement."

    > >
    > > This myth has to be de-bunked...
    > >
    > > JBOSS dilutes the standards and pollutes the J2EE market since it doesn't contribute to the standards process (JBOSS Group, nor Mark Fleury, nor Bill Burke are members of the JCP) but they [pretend to]use the standards to compete with the companies that invest in the development of the standards.
    > >
    > > Matt
    >
    > pure FUD
    >
    > some of JBoss Group are on the JSR committees, take a look at JMX for instance. they piooneered much of that work. It would be nice if you actually knew what you were talking about, but then again all you are spreading is FUD.
    >
    > http://www.jcp.org/en/participation/members/L
    > L is for Lindfors i.e. juha
    here are some more need I go on
    http://www.jcp.org/en/jsr/detail?id=160

    Tom Elrod and Juha Lindfors

    even more to debunk your FUD, lies, and damn lies

    tx

    Matt
  24. Good find![ Go to top ]

    I applaud your investigative abilities!

    They seem to be better than mine -Perhaps you can publish a list of JBOSS/JCP members and the specs they contribute to.
     
    Who knows, maybe we'll find that JBOSS invented Java, J2EE, etc. and SUN stole it?

    thanks,
    Matt
  25. shut up[ Go to top ]

    Considering that they have just proven that you have posted false information,
    then I suggest you either apologize or at least shut up.
  26. shut up????[ Go to top ]

    Censorship at TSS...hmm???

    I definitely will stand behind my posts.
    Go to http://www.jcp.org/en/participation/members/J
    look for the names I mentioned:
     JBOSS GROUP, Marc Fleury, Bill Burke, even Rickard Oberg.

    They are still not there!
    -even after the post that supposedly proved me wrong ;-}

    One or two guys contributing to 1-2 of the 200 or so JSRs does not weaken my point. Which is that if we relied on Open Source's "rapid evolution" we would still be coding in C.
    Matt
  27. shut up????[ Go to top ]

    Open Source is, agreeably, pretty bad at actually coming up with new stuff. JBoss, with its model, seems to actually be getting out of that rut and doing some new things.

    I don't really see JBoss as being services vs. someone like BEA being product. Both sell a stack of things, a product (free/open source or not), services (expensive consultants), and things that may or may not directly be under a JSR. (WebLogic Portal, etc., or AOP in JBoss's case.)

    I don't see what the big deal is.

    Besides, if we relied on straight open source's rapid evolution, we'd all be coding in PHP. ;-) Juuust kidding.

    You can stand by your JCP statements all day, but JMX is one of the cooler projects and you have to give props to JBoss for their implementation. Like any other vendor, they have things that are JCP-direct and offerings that extend the "standards."

    Just because they have a different management model doesn't mean they're of the devil. There are a lot of people, and a lot of managers, who don't see the point in paying for licenses for infrastructure software, when a free/open-source product is available. They'd be stupid not to at least investigate it.

    Steve
  28. Cheap retoric[ Go to top ]

    The fact is that JBoss developers are heavily involved
    in the JSR's.

    And then your point is false.

    That the JBoss group CEO is not a JCP member is rather irrelevant.

    [I have a feeling that if I lookup IBM/BEA/SUN CEO or even
    the relevant VP for app-servers, then they are not either]

    And you are probably rigth that IBM, BEA and SUN has many more
    people into the JSR's.

    But guess what - they are also bigger companies.

    I would not be surprised if JBoss was the company of those 4 that
    had the highest percentage of JSR participants.
  29. Dilution?[ Go to top ]

    Matt Gunter,

    Exactly how is it diluting the standards? Are you referring to trademark dilution? Value dilution (to the customers)? Economic dilution to the industry?

    And who cares if they're on JCP anyway? Even if their jobs were _solely_ to develop a J2EE compliant server, why does being involved in JCP make a difference to that endeavor?

    I think though, that you really need to explain the dilution comment. You throw stuff like that out there, and it doesn't even seem that you understand what the term dilution means in this context, let alone how someone is doing it. Sounds like more FUD to me.

    Explain yourself!

    Jason McKerr
    Northwest Alliance for Computational Science and Engineering
  30. pee in the pool[ Go to top ]

    McFleury explains it best:
    "Open source is like pee, you can put it in the swimming pool but you can't take it out".

    ...i.e. pollution and dilution... and pretty soon we're all swimming in Marc's warm pee.
    Matt
  31. Pollution and Dilution?[ Go to top ]

    "...i.e. pollution and dilution... and pretty soon we're all swimming in Marc's warm pee."

    What's with all the talk about 'purity', Matt? You're beginning to sound like a 20th century German dictator. ;)

    --Will Tabat
  32. Open source != evolution[ Go to top ]

    Matt G -

    I understand that you are tied at the hip with BEA but your statements make no sense, even in that context. Show exactly how JBoss dilutes the standards and pollutes the J2EE market. When I write a J2EE application, I find that I am more than able to deploy it on JBoss or Weblogic or OC4J or Orion or EA Server or.... I notice no pollution or dilution. All of them, including JBoss, have various extensions and additional components. So what? That's neither pollution nor dilution. By the way, there are JBoss developers on the JCP. Go look it up. Though it doesn't matter if they are on it or not, really.

    If you are going to make statements like that, explain exactly what you mean. I am assuming you are the same Matt Gunter who is a software engineer at BEA - so I am also assuming that you might know how to back up your statements. Please do so, with examples.

    Cheers
    Ray - User of many application servers
  33. Open source != evolution[ Go to top ]

    Hi Ray,
    responding to:
    "If you are going to make statements like that, explain exactly what you mean."

    Dilution:
    There is a lot of effort/investment that goes into bringing a product to market and then creating a market for it. Standards development, product definition, product packaging, partner product coordination, marketing, evangelizing, even selling are all necessary.

    Open Source does not incur any of these costs, nor does it pass any on to the customer, thereby short-circuiting the return owed on all of that invesment.

    So the first question is: would JBOSS exist without capitalizing on SUN, BEA, Oracle and IBM's efforts in the first place?

    The second question is: in a future dominated by Free Open Source "counterfeits" of commercial products, what company is going to continue to design the next generation of products, work on standards bodies, hold conferences, proactively promote technology, advertise on the serverside, etc.?


    Pollution:
    Not being certified but claiming J2EE status.
    Likely infringing on Sun's J2EE License and therefore impacts all licensees.
    Promotes a damaging belief that software IP is "valueless".
    Promotes a damaging belief that Services are equivalent to or superior to Product.
    and finally, back to the original - the damaging belief that "...nobody pushes the evolution of software better or faster than the Open Source Movement."

    The combined negative impact these both have on the industry is significant and must either be corrected or will lead to an industry with no investment, no technical improvement, and no choice in technologies.
    (see thread http://www.theserverside.com/home/thread.jsp?thread_id=18933 for more discussion on this.)

    regards,
    Matt
  34. Open source != evolution[ Go to top ]

    "There is a lot of effort/investment that goes into bringing a product to market and then creating a market for it. Standards development, product definition, product packaging, partner product coordination, marketing, evangelizing, even selling are all necessary. "

    And JBoss hasn't taken the effort or made that investment? That's silly. JBoss container developers have done just as much work as any "real vendor" container developers. As everyone knows by now, they too do their product packaging, partnering, marketing, and more than enough evangelizing.

    JBoss just didn't sprout up from thin air as you seem to suggest. Money "owed" can be decided by a container developer to be taken in services, instead of licenses. What's wrong with that? It's a free country. Hell, Marc Fleury would be getting RICHER if he actually sold licenses like everyone else. But they've decided to take a different path. That doesn't mean they're not self-interested, but neither does it mean they're evil moneymongers ruining it for everyone else.

    "The combined negative impact these both have on the industry is significant and must either be corrected or will lead to an industry with no investment, no technical improvement, and no choice in technologies."

    I know you want to return to the glory of the .com days, but non-software development companies are really looking at their bottom lines, and no longer feel the need to support commodities at exorbitant rates.

    Steve
  35. Open source != evolution[ Go to top ]

    Matt,


    There you go, spouting economic nonsense about investment, economic drivers, and whole host of business and economic issues that you clearly don't understand. As I mentioned in the other post, differentiation is what sells, even if only for a market penetration strategy. And companies will continue to differentiate. Differentiation makes money. If you can't differentiate from the competition, you're out. That's true both against proprietary and open-source.

    It's clear that you either ignored my post in that thread, or just plain don't understand much at all about economics. But you sure do claim to have a lot of knowledge in that area.

    You are the one spreading this economic FUD.

    Take off your blinders and read up on this stuff. A lot of people are successfully making money with open-source (me included since it pays my salary indirectly).

    Jason McKerr
    Northwest Alliance for Computational Science and Engineering
  36. Matt,

    reply to "Dilution" : You claim that it takes a lot of effort and investment to create standards. I agree. If you feel treated unfair because someone uses a standard that you've invested in, my suggestion is : Don't participate in standards. Standards tend to benefit a lot of people that didn't invest in it. I think participating in JCP is absolutely no requirement to produce a good app-server. In my humble opinion, JBoss has contributed a lot more for the massive adoption of J2EE then you might imagine.

    reply to "Pollution" : Claiming that JBoss is not certified doesn't mean anything to me. If you could give an example of a violation against the J2EE-spec, that would be interesting ! I haven't heard a lot of those so I am very convinced of that JBoss adheres to the J2EE spec. Especially since there are big companies that would benefit from exposing such incompatibilities. JBoss has convinced me they really want to (-and do-) comply with the spec.

    In my opinion, Sun's J2EE certification strategy gives a false reassurance to managers that decide on which app-server to choose. I don't understand that you attack JBoss on certification while other commercial vendors (that pay for the TCK), may self-report the results (correct me if I'm wrong). In these circumstances it just doesn't seem right to attack with the certification-argument.

    Tom, an independant consultant with no ties to any of the app-servers.
  37. Matt,

    >
    > reply to "Dilution" : You claim that it takes a lot of effort and investment to create standards. I agree. If you feel treated unfair because someone uses a standard that you've invested in, my suggestion is : Don't participate in standards. Standards tend to benefit a lot of people that didn't invest in it.

    I'd like to support Matt's opinion. What other industry offers things for free? Name one(no Costco freesamples, please). Until now the only "killing" Linux realised was SCO(and maybe AIX). I find it ridiculous that the fact that Jboss charges for services is considered dubious and blamable. I'd like to see all OS chraging for services. The good parallel would be with name/generic drugs.
    But they are protected by patents...
  38. Nonsense[ Go to top ]

    To my best knowledge open source has shown a definate capability
    to create new capabilities.

    JBoss, Apache, Linux etc.etc..

    There are no reason to belive that open source dominance would
    reduce the progress in products.

    I can not take such an assumption seriously. The real world
    tells a different story.

    In fact I think that it is often the commercial products that
    steal the good ideas from the open source products.

    You can argue that you are worried about f.ex. lack of documentation
    for open source. It is usually behind that for commercial products.
  39. Open source != evolution[ Go to top ]

    Matt -
    Everyone knows that there is a lot of work that goes into bringing a product to market. J2EE is a standard. It is published. Anyone can create a product based on those standards - that includes BEA, JBoss and anyone in between. As a consumer of both Weblogic and JBoss, I and my clients are glad to have the choice of both.

    As you know, JBoss developers do participate in standards development as members of the JCP. They invest the most precious resource there is: time. And of course, JBoss Group exists to make cash on top of the JBoss code-base and does a tremendous amount of marketing, evangelizing and selling as well as product definition and packaging, if you are in to that sort of thing.


    <q>
    So the first question is: would JBOSS exist without capitalizing on SUN, BEA, Oracle and IBM's efforts in the first place?
    </q>


    Open Source in general exists because of larger commercial vehicles (though this isn't always the case). So the answer to your question is probably "no". And open source doesn't spend a lot of effort in creating the market initially - that is typically the role of a few companies (in this case IBM, BEA, et al).

    But fortunately, creating the market doesn't beget market ownership. Since J2EE is an open standard, any ol' body can write software to those standards and create a product. There is no rule or law that says they have to participate in that market in the same way. There is no rule or law that says that if BEA spends $X on developing those standards or creating a market, someone owes them something. The only thing that counts in the end is the marketplace. As long as BEA has customers that want to pay them money (and to my knowledge, they still have plenty), then that's about all they can ask for. Those who have the best product and best market footprint win. If you spend one hundred billion dollars developing your product and I spend one dollar, but my product is better (or at least comparable) and I get more customers, I win.

    The only thing that I can see open source diluting in the J2EE marketplace is price.

    <q>
    The second question is: in a future dominated by Free Open Source "counterfeits" of commercial products, what company is going to continue to design the next generation of products, work on standards bodies, hold conferences, proactively promote technology, advertise on the serverside, etc.?
    </q>

    I know of no open source counterfeits. I can't think of a single O/S product that is a copy of Weblogic, for instance. mySQL and PostGreSQL aren't copies of Oracle. Hibernate isn't a copy of TopLink. But as to the rest of your question, it's not realistic to assume that the future will all be open source. I think the world needs both types of products, though perhaps I am naive. It will probably be much as it is now - with good ol' BEA out there holding conferences and advertising on TSS.

    <q>
    Not being certified but claiming J2EE status.
    </q>
    They can claim J2EE compatibility all they want. I've not had a problem with them not supporting any of the J2EE standards that I use.

    <q>
    Likely infringing on Sun's J2EE License and therefore impacts all licensees.
    <q>
    I'm not qualified to make any guesses like that.

    <q>
    Promotes a damaging belief that software IP is "valueless".
    </q>
    In what way?

    <q>
    Promotes a damaging belief that Services are equivalent to or superior to Product
    </q>

    Actually, I believe the JBoss team would assert that their product is at least as high quality as Weblogic. They just have a different business model - that's all.

    <q>
    ...nobody pushes the evolution of software better or faster than the Open Source Movement."
    </q>

    I don't know that such a statement is entirely true, but I believe that o/s does have a hand in shaping the market and the direction of products just by the nature of how o/s works. They are without many of the constraints that hinder folks in large companies - rigid management, rigid investors, and the like. It does, in fact promote somewhat greater creativity - or at least has the potential to do so.

    I think your dire predictions are entirely unfounded. If BEA can't survive in the marketplace, that is their own fault. The fact is that o/s is here to stay. It will continue to shape the marketplace.

    Cheers
    Ray
  40. Open source != evolution[ Go to top ]

    <Matt>
    JBOSS dilutes the standards and pollutes the J2EE market since it doesn't contribute to the standards process.
    </Matt>

    So are you saying that only those who contribute to the standards process are allowed to develop software that adheres to the standards that are craeted through this process? That hardly seems fair.

    Ryan
  41. Meanwhile...[ Go to top ]

    ... JBoss continues to be a reliable, affordable server that happens to run J2EE applications really well.

    J2EE certification sure would be nice. But let's face it, JBoss hasn't suffered for lack of it. Not with the 2 million downloads they boast.

    With any luck the certification suites will pick up some incompatabilities between JBoss and the spec. In fact, I'm sure JBoss will probably find a few things that they'll consider a bug in the spec! So even if this process doesn't end in certification, I'm sure the both J2EE and JBoss will be the better off for it.

    Peter @ http://rimuhosting.com - JSP and EJB (on JBoss, of course) Hosting Specialists
  42. Hi,

    Just to say that its becoming really painful to see that every little article that has the word JBoss on it produces this long list with many nonsense and fud comments... TSS is becoming day by day less interesting...
    When I see this, I really turn happy for using JONAS... Really nice, kicks JBoss ass when we talk about speed, has everything that I need and the most important thing: Doesn’t have an arrogant community (has nice and helpfull developers) and does not have Marc Fleury.
    Sorry, it was just a personal opinion.

    Pedro Costa
  43. Really tired of reading you Jonas users whine on every JBoss thread. In case you did not notice, the title clearly indicated this thread is about Sun/JBoss relationship. As a further clue, when you see JBoss in the thread header, you DO NOT have to click it and read the thread if it bothers you so much.

    When Jonas becomes a real production-quality app server, maybe we get Jonas articles too on TSS.

    /T
  44. Good response to the "arrogant developers" comment.
  45. A Time for Everything[ Go to top ]

    I just want to propose the following scenario:

    1. A spec is created.

    2. Vendors make money off of it by selling implementations thereof for, say, 5-10 years. Let's call this the spec's "commercial period."

    3. After five years, the commercial implementations are replaced by a single open-source implementation.

    4. Repeat cycle with a new spec.

    This has happened with many tools. It certainly has happened with the C compiler (who needs anything other than gcc?), it's happening with Linux.

    It should have happened with SQL, but unfortunately SQL is not a spec: it's way to vague - read C.J.Date's whole book on it, called "A Guide to the SQL Standard".

    It has happened with HTTP - who needs anything other than Apache for a web server?

    You are watching phase 2 of the cycle for enterprise Java.

    To BEA: be smart and open-source the pure J2EE part of WebLogic server, and start selling other stuff. I am sure you can come up with something. IBM did the same thing with Linux. Think outside the box, invent something.

    Guglielmo
  46. A Time for Everything[ Go to top ]

    <guglielmo>
    4. Repeat cycle with a new spec.

    This has happened with many tools. It certainly has happened with the C compiler (who needs anything other than gcc?), it's happening with Linux.
    </guglielmo>

    "Certainly"?

    You are not seriously comparing the {market|mind}share of Visual C++ with gcc, are you? Or Windows with Linux?

    <guglielmo>
    It should have happened with SQL, but unfortunately SQL is not a spec: it's way to vague
    </guglielmo>

    Well, that's a personal opinion. Here is a fact: SQL *is* a spec (with several iterations actually).

    <guglielmo>
    To BEA: be smart and open-source the pure J2EE part of WebLogic server
    </guglielmo>

    Yeah, this strategy certainly worked great for Netscape :-)

    This recurring statement never ceases to amaze me. Where did this fallacy that "open-sourcing a piece of software is suddenly going to make it take over the world" come from?

    --
    Cedric
    http://beust.com/weblog
  47. A Time for Everything[ Go to top ]

    Netscape isn't a good example, just because a browser isn't something you can charge services for, enough to keep the codebase going at a decent pace. Netscape had zero services to sell once the product was free. They tried to parlay their browser knowledge into other products, but no one wanted them enough, it seems.

    I personally don't think BEA should open source their container. They will offer products/services more to business people and less to development shops, which is fine.

    Steve
  48. A Time for Everything[ Go to top ]

    Steve writes:

    > Netscape isn't a good example, just because a browser isn't something you can charge services for,

    That's not what I was trying to say. I was just pointing out that the practice of opening the source of a software won't magically make it succeed where it failed.

    Actually, if the Netscape experience is any indication, nobody will contribute to it except the original developers.

    --
    Cedric
    http://beust.com/weblog
  49. A Time for Everything[ Go to top ]

    Yeah, I can agree with that. If opening up source is deemed to be an overtly political move (like Netscape was) then no self-respecting open-source weenie will join the mission.

    Steve
  50. A Time for Everything[ Go to top ]

    That's not what I was trying to say. I was just pointing out that the practice of opening the source of a software won't magically make it succeed where it failed.


    I didn't claim that. If I recall, I said "invent something". And I can't say BEA isn't already doing this. For example, I really like the Workflow Management system they sell. There's no Java standard for that, so that can definitely generate revenue.

    And I don't think that J2EE is dried up yet. It's pretty plain that it isn't. But the time will come. Maybe not because of JBoss (which sounds more like a new Microsoft in the making, thanks to Marc), but there's Jonas, and if Jonas goes south another server will surely take its place.

    > Actually, if the Netscape experience is any indication, nobody will contribute
    > to it except the original developers.

    But that doesn't mean anything. It's good if the code doesn't change. It means that it was the right time to open-source it.

    Look at Apache: when 2.0 was released people just yawned, because they didn't immediately need the enhancements.(Incidentally, the Apache Portable Runtime I believe was derive from a Netscape code base ..)

    As far as GCC goes, it is the industry standard C compiler on UNIX. It doesn't stack up againsta VC++ because of the different feature set (the pretty pictures, basically - the GUI). But VC++ doesn't work on UNIX.

    Again, I know BEA *is* building things on top of J2EE which aren't standard. And I am saying that you shouldn't try to artificially keep J2EE commercial, because you will just get wiped out if you are not ready. Look at Sun. They finally gave in to Linux. IBM is giving up on AIX.

    So, good luck with your strategy.

    Guglielmo
  51. A Time for Everything[ Go to top ]

    Netscape isn't a good example, just because a browser isn't something you can

    > > charge services for,


    Well you can, just make them plugins for the browser. The Reiser file system team (http://www.namesys.com/) makes money by open sourcing the file system, but charging for support and donations from sponsor companies like SuSE who want to see specific high-level features in the file system.

    >
    > That's not what I was trying to say. I was just pointing out that the
    > practice of opening the source of a software won't magically make it succeed
    > where it failed.

    Agreed, but the problem with Netscape was a project management issue, not really an open sourcing one.

    The Netscape team threw out the old version and started from scratch and lost too much time and market share during the re-write. Netscape could have gotten where it is now, faster, without losing market share by using staged releases and refactoring. (1. Version 4.6: Separate front-end from GUI, 2. Version 4.7: Rewrite the GUI. Version 4.8: Rewrite the back-end).

    Such a migration is possible in Open Source. Linus Torvalds used this staged approach with Linux (it was originally designed for the 80386) and was able to get Linux ported to the large number of platforms it currently enjoys. If Linus had taken the Netscape approach of "rewrite from scratch", Linux today would have been as finish as the GNU Hurd.


    > Actually, if the Netscape experience is any indication, nobody will contribute
    > to it except the original developers.

    I don't know the level of outside contributions, but I do know that Red Hat has made several major contributions, OEone (http://oeone.com/) has contributed, non-Netscape individuals and companies and individuals have ported Mozilla to other OSes (http://www.mozilla.org/ports/), and individual non-Netscape developers have submitted bug fixes (which are checked in by Netscape developers).

    I also suspect that something like Phoenix (a.k.a. Mozilla Firebird Web Browser) would not have been started if Mozilla was not open-sourced.
  52. A Time for Everything[ Go to top ]

    I agree BEA should not open source their server, I see no benefit in that. Then again, BEA could follow HP and give their server away for free and focus on servicing the customer ala JBoss. BEA surely would not suffer the consequences of HP as Weblogic is well respected. This is always an option. Of course, it would probably follow the pattern of lowering prices to a point where the license fee was negligible. They have options. They aren't going away.

    JBoss has not taken off just because it is open source, look at the Jonas thread. Some of the reasons it has taken off are innovation, developer support, all followed by a recession. In BEA's growth period they took off because of time to market, innovation, all during a dotcom boom, ie bull market. Economics has played a big factor in how this market has been shaping up since its inception.

    tx


    Matt
  53. A Time for Everything[ Go to top ]

    I agree BEA should not open source their server, I see no benefit in that.


    There is no financial benefit, that's for sure. I think there *are* benefits to the company itself, but exactly what they are is another discussion.

    The reason I was advocating was for the greater benefit of technological progress. At any given time, there is a baseline of technology which is accessible to all companies. When this baseline advances it spurs growth.

    So I am just saying that after IBM, BEA, and ORACLE have made enough from *competing* in market A they should start *cooperating* and start competing in market B.

    I wasn't trying to say that one vendor should try to kill the other by going open-source. I was just saying that it's a waste of human effort to have three equivalent, industrial-strength implementations of J2EE. But, again, this should kick in after vendors have gotten their money back and made a nice premium.

    Guglielmo
  54. A Time for Everything[ Go to top ]

    I agree BEA should not open source their server, I see no benefit in that.

    >
    > There is no financial benefit, that's for sure. I think there *are* benefits to the company itself, but exactly what they are is another discussion.
    >
    > The reason I was advocating was for the greater benefit of technological progress. At any given time, there is a baseline of technology which is accessible to all companies. When this baseline advances it spurs growth.
    >
    > So I am just saying that after IBM, BEA, and ORACLE have made enough from *competing* in market A they should start *cooperating* and start competing in market B.
    >
    > I wasn't trying to say that one vendor should try to kill the other by going open-source. I was just saying that it's a waste of human effort to have three equivalent, industrial-strength implementations of J2EE. But, again, this should kick in after vendors have gotten their money back and made a nice premium.
    >
    > Guglielmo

    spot on Guglielmo!



    Matt
  55. A Time for competition[ Go to top ]

    So I am just saying that after IBM, BEA, and ORACLE have made enough from >*competing* in market A they should start *cooperating* and start competing in >market B.


    Without competition there is no choice. No matter how efficient it would be to have "one industrial-strength implementation of J2EE", fair competition is actually more ideal...

    Matt
  56. Well, that's a personal opinion. Here is a fact: SQL *is* a spec (with several iterations actually).


    The devil is in the details. It's a spec as far as the marketing people are concerned. But if you actually use SQL in an application (who doesn't ..) then you will almost bind your code to a specific RDBMS.

    Incidentally, the later iterations (after SQL92, or even 89, maybe?) are off most developers' radar screen. It's as if they did not exist.

    Again, C.J.Date, pp 427-453 are chock-full of "outstanding issues" in the SQL standard.

    It just doesn't compare with the J2EE spec. The reason basically is that SQL was an attempt to standardize after vendors already had their products out. That's what killed the quality of the standard.

    Guglielmo
  57. A Time for Everything[ Go to top ]

    This is a very reasonable proposal -similar to patents/copyrights but addresses the unique aspects of the software industry.

    Shouldn't SUN be able to enforce this through their J2EE license? Maybe they are but are unwilling to do it for some reason (M$ maybe?).

    As software development techniques and tools improve they will inevitably speed up the rate at which functionality can be replicated (on top of a different design/architecture -to avoid patent/license problems). So, as rapid functionality replication is combined with free open source distribution, Commercial Software becomes increasingly endangered.

    Open source folks believe this is merely the efficiency of the Internet and the "magic" of open source. Unfortunately it is not clear at all how much of the "magic" is efficiency and how much is parasitic - My hope is that it is not the early stages of a terminal disease.

    Matt
  58. A Time for Everything[ Go to top ]

    This is a very reasonable proposal -similar to patents/copyrights but addresses the unique aspects of the software industry.


    I guess it is a bit like patents.

    > As software development techniques and tools improve they will inevitably
    > speed up the rate at which functionality can be replicated (on top of a
    > different design/architecture -to avoid patent/license problems). So, as
    > rapid functionality replication is combined with free open source
    > distribution, Commercial Software becomes increasingly endangered.

    I think existing commercial products become increasingly endangered with time, but new products can always be made. But you can't put too many constraints on a company, or they won't be able to get into those new markets. For example, Oracle RAC is a great idea, but it would not be possible without the right hardware. Hence the partnership with Dell. IBM has had a similar solution on mainframes, and they make both.

    I think the issue of staying profitability is an issue of focus, and automatically also diversification. I think if one focuses on the problem being solved for the customer as opposed to the products to be sold, one can continually adjust one's business model to avoid getting boxed in like this.

    I remember when WebLogic was the name of a company, and the product was brand new and really cool. It was on the bleeding edge of enterprise development (when many companies wouldn't give Java the time of day). Perhaps now it can happen all over again in a market that open source developers have no clue about.

    Guglielmo
  59. A Time for Everything[ Go to top ]

    "I think existing commercial products become increasingly endangered with time, but new products can always be made."

    Sure, but my point was that the increasing rate-of-replication will cause the "window of differentiation" to approach zero. With normal commercial product competition, this is OK because the feature still has market value -i.e generates income. But when competing with free, it results in the "market value" of the new (but replicated) feature to drop to near-zero.

    Matt
  60. A Time for Everything[ Go to top ]

    Sure, but my point was that the increasing rate-of-replication will cause the "window of differentiation" to approach zero.


    I wasn't thinking about it as "differentiation", I was really thinking about totally new markets. For example, BEA could write a .NET implementation. Then you are dealing with a whole new space.
  61. old friend?[ Go to top ]

    Matt Gunter, is this the Matt Gunter that Graduated in 1985 from Churchill County High School? If so, please e-mail me, thanks, if not, sorry for the intrusion. mikewardok@hotmail.com
  62. Historical perspectives[ Go to top ]

    <Guglielmo>
    1. A spec is created.
     
    2. Vendors make money off of it by selling implementations thereof for, say, 5-10 years. Let's call this the spec's "commercial period."
     
    3. After five years, the commercial implementations are replaced by a single open-source implementation.
     
    4. Repeat cycle with a new spec.
    </Guglielmo>

    Actually, the first two steps have typically occurred in reverse in the Java community--especially the Java enterprise community. Server-side Java was pioneered by companies implementing and marketing highly proprietary solutions (this includes WebLogic with its Tengah server, BroadVision, Blue Martini, etc.). The initial enterprise specs were attempts by Sun to standardize server-side technologies in an effort to keep the Java market from splintering (and spinning out of its control)--a laudable objective, since a fractured marketplace would only play into the hands of competitive platforms (read: MSFT).

    JNDI standardized a Java-specific answer to the emerging LDAP standard. EJB had forerunners and analogs prior to its specification, as well (who can forget IBM's San Francisco framework?). And JMS mandated a unified approach to Java-based MOM architectures, with early vendors such as IBM, Vitria, and others pioneering the way.

    Given this pattern, it is understandable that a commercial, closed-source vendor who helped to pioneer the technologies now embodied in the specification might feel trepidation at the prospect of upstart open-source communities eroding their share of that market. This does not mean, however, that it's a bad thing.

    Probably the biggest risk in opting for an open-source solution is the commitment/activity of the community associated with that solution. If the project can't maintain enough excitement and commitment (from both consumers and developers), it will languish. But that's nothing new. Surprise, surprise, the same thing happens to commercial vendors. Competitors introduce new features into their products or launch more effective marketing campaigns or improve their customer service, resulting in an erosion of market share. Suddenly, your employees begin to drift away to other, more exciting or more challenging opportunities, management becomes unfocused--in short, the company loses its competitive edge.

    The only difference, really, is that open-source developers typically have no financial incentive for working on a project. Instead, they must volunteer their time, unless they're lucky enough to have found a position where they can get paid to devote their full attention to their projects.

    Anyway, I don't know that I have an ultimate, over-arching point here. I just find it interesting that the only reason we're really even have these types of conversations is because of the ease of assembling a virtual community around a given technology project. Ten years ago, there were probably less than a dozen well-known and truly-distributed open-source or free software projects. Linux was still in its infancy, little more than just a beta kernel. Apache wasn't even an idea. The top three were probably emacs, gcc, and perl.

    Of course, I'm omitting probably the most successful of open organizations in the history of technology: the IETF, without which we wouldn't have the Internet or the web.

    Apologies for rambling a bit off topic ... :)

    cramer
  63. Historical perspectives[ Go to top ]


     Anyway, I don't know that I have an ultimate, over-arching point here. I just find it interesting that the only reason we're really even have these types of conversations is because of the ease of assembling a virtual community around a given technology project. Ten years ago, there were
    >

    Assembling open source communities sounds like one of the most difficult things to do, assembling proprietary R&D labs is easier as you have money, command employees etc. In open source you don't control the people, how can assembling them be "easy". I don't buy the argument.
  64. Historical perspectives[ Go to top ]

    <zahid raman>
    Assembling open source communities sounds like one of the most difficult things to do, assembling proprietary R&D labs is easier as you have money, command employees etc. In open source you don't control the people, how can assembling them be "easy". I don't buy the argument.
    </zahid raman>

    Actually, the comparison I was shooting for was with the (relative) difficulty of assembling such a community in the past, given the tools available. You're absolutely right that commercial ventures have a lot of advantages over open-source communities when it comes to assembling teams, not the least of which is money. But these days, with such supporting infrastructures as SourceForge and such community rallying points as TSS, if you've got a decent idea and a good start at a design and implementation, you can launch an open-source project and publicize it rather easily.

    Obviously, there's no guarantee that the masses will flock to your aid, but it's certainly easier to get started today than with the email/Usenet mechanisms predominant ten years ago (not to mention sheer size of the audience today, numbering in the millions of developers on the web today versus maybe tens of thousands on Usenet ten years ago). These improvements are really the only reason that open-source communities like JBoss, Apache, and probably even Linux can even try to be competitive with commercial software vendors.

    cramer
  65. What's the big deal?[ Go to top ]

    Sun wants everyone who's making money off J2EE to pay them.
    Marc Fleury wants to make money off the J2EE but not pay Sun.
    Why is everyone getting so worked up about this? How does this personally affect you in any way? This looks like a business negotiation problem between Sun and the JBoss Group.
    If you like JBoss, use it.
    If you don't like JBoss or Marc Fleury, don't use JBoss.
  66. Censorship in this thread[ Go to top ]

    Dion,

       Next time you delete my post or anyone else, you should state so. And btw, I thought you worked for The Middleware Company? Why are you deleting posts in here? Isn't that a conflict of interest since TMC is owned by M$?
  67. Why are you spreading FUD?[ Go to top ]

    n n -

    I am getting frustrated with the FUD that you are putting in these discussions.
    My job is architect for TSS. I care a lot about this community, as well as J2EE. If you do some research that I have written several articles, taken part in many events.... and generally been active in the J2EE community for a long time. I do not mind when someone accused me of something that I have done, but I have NO interest in "M$". All of the posts that I make are trying to get info to the J2EE community. Heathy debate is important. Knowing about your competition, and what they say about you is important.

    NOTE: I am marking both this at your post as NOISE, as they are totally off topic.

    Dion