Discussions

News: JBoss closes $10 million in venture financing

  1. JBoss closes $10 million in venture financing (89 messages)

    Well, it seems that some people agree with the vision of "Professional Open Source" that Marc has. JBoss just announced that they secured $10 million in oversubscribed first round financing. The money is to be used for product development, infrastructure build out, marketing, and more.

    David Skok, founder of SilverStream Software, joined the JBoss board of directors.

    Read the press release: JBoss closes $10 million in venture financing

    In related news, read:
    JBoss, Inc. announces Fourth Quarter 2003 highlights

    JBoss investment paves way for flotation

    "In an interview with ZDNet UK, Marc Fleury, founder and president of JBoss, did not rule out the possibility of an IPO, but said the options remain open. "Right now every venture capitalist invests with an exit strategy in mind," he said. "Everybody in the group hopes for an exit strategy, whether it is an IPO or just for the business to continue growing profitably. But we have a great company that is fast growing, and everybody focused on execution. Whatever will be will be."

    Threaded Messages (89)

  2. Congratulations![ Go to top ]

    The JBoss Group now stands beside Red Hat as a company that has turned the power of open-source development into a workable venture that benefits companies and developers.

    Over the next week we will be subject to an army of bilebloggers and other playa haters who wish that their ideas were perceived to be this valuable in a post-dot-com marketplace.

    But hopefully there will be others who finally begin to understand how JBoss benefits open source developers and the Java marketplace as a whole.
  3. Beginning or the end ?[ Go to top ]

    Congratulations ! How do you propose to produce an ROI for the investors ? What is the business plan behind it ?

    Are your investors planning to turn you into RH or SCO ?
  4. Didn't Eazel burn $11 million in 16 month?[ Go to top ]

    Will be interesting to see how long JBoss Group can last.
  5. Congratulations![ Go to top ]

    The JBoss Group now stands beside Red Hat as a company that has turned the power of open-source development into a workable venture that benefits companies and developers.

    >
    Two out of hundreds is hardly a track record that inspires a willingness to duplicate or even confidence.

    No offense to JBoss intended, but holding up 2 examples of profit when for profit,closed source has hundreds more is not the shining example you make it out to be.

    Free software has many advantages but using it, in and of itself as a tool with which to eat, may not be one of them.
  6. Not Yet a BEA Killer[ Go to top ]

    Saw that BEA posted over 1 billion in earnings for 2003.

    It looks like JBoss is probably trying to position itself more like MySQL (without dual licensing) than Red Hat. Their obsession with control over the codebase points to that and probably nixed the non-profit route for J2EE certification.

    Even with help from the founders of Silverstream and Bluestone, Marc's got to be dreaming if he thinks JBoss can win major accounts away from BEA or IBM.
  7. Not Yet a BEA Killer[ Go to top ]

    Even with help from the founders of Silverstream and Bluestone, Marc's got to be dreaming if he thinks JBoss can win major accounts away from BEA or IBM.


    In fact, SilverStream was a really bad product. I hope this doesn't mean that the people who designed SilverStream will get to influence the priorities of JBoss.
  8. Not Yet a BEA Killer[ Go to top ]

    Even with help from the founders of Silverstream and Bluestone, Marc's got to be dreaming if he thinks JBoss can win major accounts away from BEA or IBM.


    BEA has moved on from being just an application server provider. Their product now include content management portal servers among other specialized extensions from the application server, while JBoss is, well, just an application server, but a pretty damn good one!

    So I don't think we're making quite an apple-to-apples comparison here ;)
  9. Not Yet a X Killer[ Go to top ]

    People said the same thing about WebLogic/BEA and dozens of other now successful companies. Plus, they already have displaced both BEA and IBM in Fortune 1000 accounts, which certainly qualify as major. I can't figure out why some people are so [mistakenly] skeptical about JBoss. I have no idea how much money they will make when thye land large accounts, but there is certainly room out there for small and large companies to adopt a free app server that works well instead of paying ridiculous amounts of money for bloatware like WebSphere and platinum gilded, diamond encrusted BEA.
  10. Not Yet a X Killer[ Go to top ]

    \Nicholas Whitehead\
    I can't figure out why some people are so [mistakenly] skeptical about JBoss. I have no idea how much money they will make when thye land large accounts, but there is certainly room out there for small and large companies to adopt a free app server that works well instead of paying ridiculous amounts of money for bloatware like WebSphere and platinum gilded, diamond encrusted BEA.
    \Nicholas Whitehead\

    Many people are skeptical because JBoss does not do everything the commercial offerings do. Want distributed transactions, XA Logging, a solid JMS implementation? Then you go commercial. Marc has publically stated that there are some things that aren't suitable to open source and that you should buy a commercial solution for them. I happen to disagree - I think these things can be done in open source - but that's his view and that's also the JBoss reality today. JBoss does happen to suit the neeeds of many organizations, but there are many situations where it plain doesn't work and the commercial guys have a very clear advantage. Maybe with VC funding that will now change, or maybe not, who knows? And meanwhile there's also Jonas going down the J2EE cert path, and Geronimo still under development.

         -Mike
  11. Not Yet a X Killer[ Go to top ]

    Mike,

    "Want distributed transactions, XA Logging, a solid JMS implementation? Then you go commercial."

    It's always a good practice to stay as objective as can be, as far as choosing a tool is concerned. The J2EE world is still very broad in terms of offering and JBoss is in a niche market (some call it low-end). My personal opinion is that Jonas is quietly gaining attention, credibility and momentum. Another opinion of mine is that many times around, projects only need a servlet/JSP engine and at some point, architects and designers might notice it. Finally, focus is slowly shifting to a higher abstraction level where application servers are merely a commodity: the battlefield is mostly about integration and portal (I think it is where the money is actually). Therefore, JBoss, as well as other contenders such as Orion, Jonas, will have to strive to get market shares, especially knowing that technology is a moving target and a fad, which means that application server companies need a money cushion to afford strategy changes.

    It will not be easy for JBoss to grow on JBoss support/consultancy, but it is not impossible.

                    Yann
  12. Not Yet a X Killer[ Go to top ]


    > Many people are skeptical because JBoss does not do everything the commercial offerings do. Want distributed transactions, XA Logging, a solid JMS implementation? Then you go commercial.

    And if you don't need these things, why are you even installing a full J2EE stack? All you really need is a Servlet container, which is a hell of a lot cheaper, in terms of TCO, than even JBoss. Tomcat is free, if somewhat flaky. Resin is fast and SIMPLE to setup, and pretty cheap. Design your system around a lightweight container to manage components, and you're on your way...

    The bottom line, don't pay for something you don't need, even if it's just time and resources rather than money. Without production quality JMS and JTA transaction manager, I can't see what value JBoss can offer?
  13. Not Yet a X Killer[ Go to top ]

    \Jason Carreira\
    The bottom line, don't pay for something you don't need, even if it's just time and resources rather than money. Without production quality JMS and JTA transaction manager, I can't see what value JBoss can offer?
    \Jason Carreira\

    Umm.....a JMX microkernel and fledgling AOP implementation? Oh, right, you can get those w/out JBoss....

    Let's see, let's see....CMP Entity Beans? Oh, right, the chief software architect says the current implementation is crap and must be re-written from scratch.

    Ah...a strong ISP base due to the awesome power of the LGPL license? No?

    Harumph...eh, how about JNDI? Yeah, you do get JNDI! That's it!

    Other than that, er, OK, I give up. There must be something though, right?

        -Mike
  14. Not Yet a X Killer[ Go to top ]

    Let's see, let's see....CMP Entity Beans? Oh, right, the chief software architect says the current implementation is crap and must be re-written from scratch.


    :-)
  15. Not Yet a X Killer[ Go to top ]


    > > Many people are skeptical because JBoss does not do everything the commercial offerings do. Want distributed transactions, XA Logging, a solid JMS implementation? Then you go commercial.
    >
    > And if you don't need these things, why are you even installing a full J2EE stack? All you really need is a Servlet container, which is a hell of a lot cheaper, in terms of TCO, than even JBoss. Tomcat is free, if somewhat flaky. Resin is fast and SIMPLE to setup, and pretty cheap. Design your system around a lightweight container to manage components, and you're on your way...
    >

    It seems from the press release that JBoss Inc. is more than just JBoss. Hibernate, Tomcat, JBoss Nukes. How can JBoss Inc. claim Hibernate and Tomcat?

    > The bottom line, don't pay for something you don't need, even if it's just time and resources rather than money. Without production quality JMS and JTA transaction manager, I can't see what value JBoss can offer?

    JMS performs quite well in production with the more recent versions of JBoss. I've even heard some rumblings of JMS clustering being available in later JBOss releases. Anybody have any more info on that?

    As far as distributed tx and recovery and logging goes, yes, that is a weakpoint, but how often do apps require that kind of functionality. My guess the number is in the minority. I guess what I am saying is that the JBoss J2EE stack is quite complete minus distributed tx and recovery based from what I've seen.

    /Joe
  16. Not Yet a X Killer[ Go to top ]

    As far as distributed tx and recovery and logging goes, yes, that is a weakpoint, but how often do apps require that kind of functionality.


    If you have a queue and a database, you need that functionality, no?
  17. Not Yet a X Killer[ Go to top ]

    \Joe Murray\
    JMS performs quite well in production with the more recent versions of JBoss. I've even heard some rumblings of JMS clustering being available in later JBOss releases. Anybody have any more info on that?
    \Joe Murray\

    Even JBoss people say that their JMS implementation isn't very good. They're doing a complete re-write from scratch of the implementation even as I type.

    \Joe Murray\
    As far as distributed tx and recovery and logging goes, yes, that is a weakpoint, but how often do apps require that kind of functionality. My guess the number is in the minority. I guess what I am saying is that the JBoss J2EE stack is quite complete minus distributed tx and recovery based from what I've seen.
    \Joe Murray\

    You'd be surprised at the growing number of people who do need these things. XA in general used to be a tiny niche thing, but with JCA and JMS it's becoming increasingly common to use multiple resources in a single transaction - and for that you need XA.

    JBoss can handle the XA protocol stuff allright, but w/out transaction logging, if a failure happens it can leave dangling transactions hanging in the breeze. Many people don't think this is a big deal - until they realize that an in-doubt transaction in your database can be holding locks preventing anything else in your app from doing useful work. And frankly, there are many places I've seen where they plain don't realize that there's any potential for that - they don't really understand XA, don't know that in-doubts can lock up your database, and don't know that JBoss doesn't do recovery.

    As for distributed transactions, that may indeed be less common. But if you need it, you need it, period.

    In all I think Jason's comments are valid - JBoss employees are saying CMP their CMP is bad, JMS is bad, we don't do distributed transactions or XA recovery, plus many other more minor things they don't do. Strip all those things away from the equation, and is what's left really better than the alternatives? Might it not make sense to go to a servlet container, try out one of the new emerging IoC containers, maybe look at Jonas, or generally go down a different road altogether?

         -Mike
  18. Not Yet a X Killer[ Go to top ]


    > As far as distributed tx and recovery and logging goes, yes, that is a weakpoint, but how often do apps require that kind of functionality. My guess the number is in the minority. I guess what I am saying is that the JBoss J2EE stack is quite complete minus distributed tx and recovery based from what I've seen.
    >

    Yes, it is a minority that need distributed transactions... It's just the ones that actually benefit from a full J2EE stack. :-) So, again, if JBoss can't support the MOST BASIC requirement for an Enterprise system (remember, J2EE has the word "Enterprise" in it), then what value does it add? JMS working well or not is moot for enterprise systems if you can't consume messages in a JTA transaction and be confident in the recoverability of the system.
  19. JMS[ Go to top ]

    Want distributed transactions, XA Logging, a solid JMS implementation? Then you go commercial.

    Really? That's pretty funny, Mike, because my recollection was that when you needed these things you grafted your code into an existing open source project to get them.

    Incidentally, my JBoss server (3.2.2) processes over one million messages each day from fourteen queues, under heavily concurrent loads. Extremely robust, plus I get failover, but not clustering. I'm not going to break any world records with these kinds of numbers, but it seems sufficient to qualify as a 'solid JMS implementation'.
  20. JMS[ Go to top ]

    \Corby Page\
    Want distributed transactions, XA Logging, a solid JMS implementation? Then you go commercial.

    Really? That's pretty funny, Mike, because my recollection was that when you needed these things you grafted your code into an existing open source project to get them.
    \Corby Page\

    Read my full comments - I believe you _can_ do these things in open source. But in the specific JBoss case, you won't find them.

    In my own case, the point of the exercise was not to graft code onto an open source implementation. The point was to integrate alot of legacy and internal proprietary messaging and monitoring stuff with Java clients through JMS, and an open source (BSD licensed) product turned out to be the best and shortest path for where we needed to go. 18 months of several people's time, and we're still working on that - and I somehow don't think that most people interested in JMS are interested in spending man-years of effort.

    That was then, and this is now. The Geronimo folks have lit a fire under several other projects as well, and distributability and reliability are among the requirements that are becoming important in open source development.

    \Corby Page\
    Incidentally, my JBoss server (3.2.2) processes over one million messages each day from fourteen queues, under heavily concurrent loads. Extremely robust, plus I get failover, but not clustering. I'm not going to break any world records with these kinds of numbers, but it seems sufficient to qualify as a 'solid JMS implementation'.
    \Corby Page\

    Corby, go talk to Bill Burke about this, not me. He's bitched about the JBoss JMS implementation in public for quite a long time now. Go talk to the guy who's re-writing JMS from scratch for JBoss. I'm happy that JBossMQ has worked out well for you - for many other people, including the Chief Software Architect for JBoss, it's been a big giant let down.

         -Mike
  21. JMS[ Go to top ]

    Wasn't the original JBoss JMS (not the revamped stuff) written by one of the guys who left for Geronimo?
  22. JMS[ Go to top ]

    It could be, I don't remember anymore. My fuzzy memory seems to remember that JBoss MQ came from SpyderMQ (and you still see Spy* references throughout the code). I think Geronimo will be using OpenJMS initially for their JMS provider. And OpenJMS is moving towards a full XA implementation - last time I checked it wasn't complete, but alot of new code has been added in to address XA.

    The Spyder-based (JBossMQ) stuff looks like it's being replaced with "JBoss JMS". See the JMS Development forum on jboss.org, and the CVS tree at http://cvs.sourceforge.net/viewcvs.py/jboss/jboss-jms. This is seperate and distinct from JBoss MQ (http://cvs.sourceforge.net/viewcvs.py/jboss/jbossmq/).

         -Mike
  23. JMS[ Go to top ]

    Corby, go talk to Bill Burke about this, not me.

    If Bill posts that the JBoss 3.2.x implementation is not suitable for production use under large loads, I'll be happy to make my points to him. And I have talked to Marc and Sasha in the past about what I am doing with JBossMQ.

    But Bill didn't make these comments, you did. Now be man enough to back them up.

    Go talk to the guy who's re-writing JMS from scratch for JBoss.

    I assume you are talking about Nathan. He never slammed the JBossMQ implementation the way you have, although he recognized that it needs distributed transaction support. The primary motivation for rewriting JMS from scratch is to integrate it into the AOP framework so that it is using the same system of invokers, proxies, and AOP support as the other server components.
  24. JMS[ Go to top ]

    Corby, wake up - JBoss JMS has been slammed by just about everyone but the pope.

    See some of Bill Burke's fine comments on some of the JBoss forums. Also, feel free to check out:

        http://www.jboss.org/developers/projects/jboss/jms/project-plan

    I quote from the opening paragraph:

    "The current JMS implementation (JBossMQ) is one of the weak points of the JBoss J2EE stack. While it is functional, and better than many JMS implementations out there, it is not the same level of quality and functionality that JBoss users expect or demand. Furthermore, it does not use many of the patterns and system level services formalized and made available by the core JBoss server. We have, therefore, elected to embark on a complete rewrite of the code."

    FYI, Adrian Brock is the lead of the re-write, as indicated on the project plan page and all of the code checkins in CVS for the new re-write codebase.

        -Mike
  25. P.S.[ Go to top ]

    Was that man enough for you Corby, or do you want me to drag out every negative comment JBoss employees have made about the current JMS stack? It might take me weeks to collate that much data, but it'd be worth it to prove that I'm a man to you. :-)

       -Mike
  26. JMS[ Go to top ]

    Corby, wake up - JBoss JMS has been slammed by just about everyone but the pope.

    >
    > See some of Bill Burke's fine comments on some of the JBoss forums. Also, feel free to check out:
    >
    > http://www.jboss.org/developers/projects/jboss/jms/project-plan
    >
    > I quote from the opening paragraph:
    >
    > "The current JMS implementation (JBossMQ) is one of the weak points of the JBoss J2EE stack. While it is functional, and better than many JMS implementations out there, it is not the same level of quality and functionality that JBoss users expect or demand. Furthermore, it does not use many of the patterns and system level services formalized and made available by the core JBoss server. We have, therefore, elected to embark on a complete rewrite of the code."
    >
    > FYI, Adrian Brock is the lead of the re-write, as indicated on the project plan page and all of the code checkins in CVS for the new re-write codebase.
    >
    > -Mike

    Like I said. JMS has seen dramatic improvements over the past 6 months, JBoss 3.2.2 and beyond. This plan looks like it was written last July. Same could be said for the CMP implementation. Vast improvements over past 6 months. So, if JBoss employees stated that JMS and CMP sucked awhile ago, it seems that they have worked to address the situation.

    /Joe
  27. JMS[ Go to top ]

    Joe, FYI the plan was for JBoss 4.0 and the new JMS code base. Not JBossMQ. Some work has been done to JBossMQ, most visibly the "HA" JMS stuff, because people didn't want to wait for JBoss 4.0 (I can hardly blame them for that). But, if you read the docs and the forum comments and the code, even the HA is only partial, and 4.0 and the new JMS is where the effort's going.

    Just so you understand - the plan is for new-JMS and JBoss 4.0, not JBoss 3.2.x and JBoss MQ.

    As I said initially - maybe JBoss MQ is good enough for you guys. If so, congratulations. But it's not good enough for many other people, though - a thought shared by JBoss developers themselves.

    As for CMP - within the past month JBossians were still making very forceful comments about the existing CMP codebase, and showed a great deal of excitement over re-writing that to use Hibernate. You and Corby obviously believe the JBoss as it is right now is just fabulous, but I don't see alot of others sharing your view, including the Chief Software Architect and the CEO who control the software.

        -Mike
  28. JMS[ Go to top ]

    Joe, FYI the plan was for JBoss 4.0 and the new JMS code base. Not JBossMQ. Some work has been done to JBossMQ, most visibly the "HA" JMS stuff, because people didn't want to wait for JBoss 4.0 (I can hardly blame them for that). But, if you read the docs and the forum comments and the code, even the HA is only partial, and 4.0 and the new JMS is where the effort's going.

    >
    > Just so you understand - the plan is for new-JMS and JBoss 4.0, not JBoss 3.2.x and JBoss MQ.

    Really doesn't matter whether JBoss 3.2 JMS is being rewritten in JBoss 4. It is obvious to me at least that the JBossians are putting resources towards improving the existing 3.2 base as well as any rewrite.

    /Joe
  29. JMS[ Go to top ]

    \Joe Murray\
    Really doesn't matter whether JBoss 3.2 JMS is being rewritten in JBoss 4. It is obvious to me at least that the JBossians are putting resources towards improving the existing 3.2 base as well as any rewrite.
    \Joe Murray\

    I've commented on this in the past. Since JBoss is open source, you can divine quite a bit about what's going on with the development just by watching the CVS code base. The stark reality is that they only have 10 or so active developers, and that number of people simply can't service the 3.x code base and be engaged in the 4.0 re-write simultaneously. What I've seen in the CVS tree is a ping-pong-rest, ping-pong-rest, ping-pong-rest pattern. Consider the "ping" the 3.x code base, the "pong" 4.0 work, and "rest" being engaged in other work (consulting, symposiums, and the like). From individuals you can see a flurry of work in a ping phase, and then another flurry on pong, and then the usual rest period.

    Some work was done on JBossMQ for 3.x, you can't dispute that. You also can't really dispute that it's not a full-bore effort - even the "JMS HA" stuff is only partial, and that's because of the 4.0 code base hovering in the background. The JMS base may have improved in 3.2, but at the same time it doesn't look to me like it's improved all that substantially. And the JBoss developers seem to share that view - hence the re-write.

    Going back to the original points - JMS may be better under 3.2, but that's a self-comparison. How does it compare to other offerings on the market? From where I sit, nothing has been done to significantly changes its status in comparison to the commercial offerings. Based on what's been done, no one's going to re-evalute JBossMQ now with an eye towards throwing away BEA, or IBM MQ or the like.

    A less kind way of putting this is that JBossMQ had no where to go but up...

        -Mike
  30. JMS[ Go to top ]

    \Joe Murray\

    > Really doesn't matter whether JBoss 3.2 JMS is being rewritten in JBoss 4. It is obvious to me at least that the JBossians are putting resources towards improving the existing 3.2 base as well as any rewrite.
    > \Joe Murray\
    >
    > I've commented on this in the past. Since JBoss is open source, you can divine quite a bit about what's going on with the development just by watching the CVS code base. The stark reality is that they only have 10 or so active developers, and that number of people simply can't service the 3.x code base and be engaged in the 4.0 re-write simultaneously. What I've seen in the CVS tree is a ping-pong-rest, ping-pong-rest, ping-pong-rest pattern. Consider the "ping" the 3.x code base, the "pong" 4.0 work, and "rest" being engaged in other work (consulting, symposiums, and the like). From individuals you can see a flurry of work in a ping phase, and then another flurry on pong, and then the usual rest period.
    >

    You would actually see this same ping-pong activity at a regular product company. With flurries of CVS commits near release time and usually a lull afterwards. The ping being a maitenance release, and the pong work being done for the future. Product companies usually do not have two teams working on maintaining the current release and future next.0 releases. They are usually the same team. This flurry is probably more common in open source projects as people contribute in their spare time? I don't know, never contributed anything to open source, or watched CVS commits of any project as carefully as you seem to do. So, what I am saying is that this flurry is probably a normal thing.

    As far as JBossMQ going nowhere but up, hey, I'm happy with it and I don't pay a dime to JBoss Group, BEA, IBM, or Sonic.

    /Joe
  31. JMS[ Go to top ]

    \Joe Murray\
    You would actually see this same ping-pong activity at a regular product company. With flurries of CVS commits near release time and usually a lull afterwards. The ping being a maitenance release, and the pong work being done for the future. Product companies usually do not have two teams working on maintaining the current release and future next.0 releases. They are usually the same team. This flurry is probably more common in open source projects as people contribute in their spare time? I don't know, never contributed anything to open source, or watched CVS commits of any project as carefully as you seem to do. So, what I am saying is that this flurry is probably a normal thing.
    \Joe Murray\

    There's something to what you say, but the difference is the number of people involved in all aspects of the development. JBoss seems to have around 10 people involved in serious development work on the code base, according to publically available information. Bill Burke and Marc Fleury have also stated in public that those 10 people are doing all of the essential work (with the usual disses to the poor unwashed masses :-).

    How many people do you think are involved in Weblogic development or Websphere development? Do you honesty think that number is anywhere near 10? Factor in things like support, pre-sales support, documentation as well.

    If you compare the two, I think a ping-pong effect will be much less noticable in the commercial efforts.

    Note that this isn't a slam on open source in general. JBoss seems rather, ah, unique in its approach. Other projects may share a ping-pong problem to some extent, but there's less of a corporate pull, and/or they have a larger base of coders. Loss of many prominent coders to the Geronimo effort can only have exacerbated this effect, and CVS again bears this out.

        -Mike
  32. JMS[ Go to top ]

    Corby, wake up - JBoss JMS has been slammed by just about everyone but the pope.

    Dude, if it's good enough for the Pope, it's good enough for me.

    See some of Bill Burke's fine comments on some of the JBoss forums. Also, feel free to check out:

    "The current JMS implementation (JBossMQ) is one of the weak points of the JBoss J2EE stack...."


    All right, since you won't stand behind your remarks, I will argue with Bill. JBoss Group has a disturbing tendency to praise code contributions made by employees, only to trash those same contributions once those people leave the company. JBossMQ and JBossCMP had serious problems 18 months ago (note: I am speaking from experience, not just repeating something I heard from a blogger), but in the 3.2.x line both are extremely usable under high loads.

    Was that man enough for you Corby?

    Or adult enough, if you prefer. No, frankly, I'm disappointed. You are smart enough to be able to back up your own statements (assuming you are not just trolling) about JBossMQ. Hiding behing Bill of all people, rather than defending your statements, is not very adult.
  33. JMS[ Go to top ]

    Corby, man, WTF? Understand what's going on - _JBoss is abandoning the 3.2.x JMS codebase and doing a ground-up re-write for 4.0_. The JBoss web site says "The current JMS implementation (JBossMQ) is one of the weak points of the JBoss J2EE stack".

    So, OK, I'll admit it. We're all wrong - Adrian Brock is wrong for writing a new JMS code base. Bill's wrong for every negative thing he's posted about JBossMQ. I'm wrong about my own criticisms of the JBoss codebase. We're all wrong, and you're right. I'll hang my head in shame, go home to my wife and tell her that Corby says I'm not a man anymore. Adrian and Bill and I suppose even Marc Fleury will likewise do the same with their significant others. Every single poster in the main JMS forums and the JMS development forum will likewise do the same. We're all wrong, you're right.

    Even so, JBossMQ is _still_ going the way of the Dodo bird.

         -Mike
  34. JMS[ Go to top ]

    I am very glad that JBossMQ (and JBossCMP, for that matter) is being re-written from the ground up and replaced. I still stand by my three points:

    1) The current version of JBossMQ is quite usable in production environments, assuming you don't want distributed transactions.

    2) JBoss tends to put themselves in an inconsistent and confusing position when they slam the contributions of ex-employees and those contributions are part of their current server product. You have noted this yourself before, so I suspect you agree.

    3) When pressed to defend your statement, it seems very disingenuous of you to rely on 'Well, Bill said it so it must be true.' You have repeatedly questioned Bill's character and motives in the past, so when this is your only defense it reinforces my impression that you make broad negative assertions without thinking about them.
  35. JMS[ Go to top ]

    Want distributed transactions, XA Logging, a solid JMS implementation? Then you go commercial.

    >
    > Really? That's pretty funny, Mike, because my recollection was that when you needed these things you grafted your code into an existing open source project to get them.
    >
    > Incidentally, my JBoss server (3.2.2) processes over one million messages each day from fourteen queues, under heavily concurrent loads. Extremely robust, plus I get failover, but not clustering. I'm not going to break any world records with these kinds of numbers, but it seems sufficient to qualify as a 'solid JMS implementation'.

    These are my impressions as well Corby. JMS and CMP performance improved dramatically with JBoss 3.2.2 and beyond. Seems they got their act together finally on these components over the past 6 months.

    /Joe
  36. Not Yet a X Killer[ Go to top ]

    Point taken. I have not done distributed transactions, and I am not sure what XA Logging is. We get by with the JMS, and it sure has improved of recent, but I'll 'fess to the issues that are still there.

    Having said that, BEA's JMS implementation has never been more than an idle curiosity that has zero standing amongst dedicated commercial (and perhaps open source) JMS vendors. It is a convenience to people who have WebLogic and may use JMS for non-critical tasks and because they need it to qualify under the Sun specs.

    At the end of the day, I think it is an excellent stack that has much to recommend it. Those parts that need help are supported by an architecture that allows you to easily plug in your own JMS, Transaction Manager, Web Server or what ever.
  37. Not Yet a X Killer[ Go to top ]

    I have not done distributed transactions, and I am not sure what XA Logging is.


    Think banking, ATMs, stock trading, billing, etc. If you're interested, I suggest that you start with a google search for "ACID".

    > Having said that, BEA's JMS implementation has never been more than an idle curiosity that has zero standing amongst dedicated commercial (and perhaps open source) JMS vendors. It is a convenience to people who have WebLogic and may use JMS for non-critical tasks and because they need it to qualify under the Sun specs.

    Categorically false.

    I work for BEA, and so can't reveal how customers leverage our products out of respect for their privacy, BUT I think it is safe to say that a very large proportion of our J2EE customers use BEA's JMS extensively, and a large number of them do so in mission critical ways (financial, telecom, transportation, defense, and government - Asia, Europe, and the Americas).

    And this is no accident: I think it also is fair to say that BEA's transaction and messaging team is broadly experienced in their field - having helped write some of the very standards you use every day (JMS for instance, SQL is another, XA is another) for example - and having developed multiple queuing products between them.

    For a BEA customer list, see
    http://www.bea.com

    For extensive info on BEA JMS, see
    http://dev2dev.bea.com/technologies/jms/index.jsp

    Full disclosure: I am a BEA employee and the above opinions are my own. (If strongly held.)
  38. Not Yet a X Killer[ Go to top ]

    Thanks Tom. I get ACID, and transactions. But the specific term "XA Logging" is a cloud.

    I also know people using BEA and they claim the JMS is sub so they integrated vendor X instead. I myself have not used it since 5 and it was highly sketchy then, but thas was a while ago, so I concede on that. This stuff gets fairly subjective and annecdotal, so let's dispense with it. However, scan the big JMS players like Fiorano, Sonic and IBM and they are benchmarking each other up, down and sideways. I have yet to see anyone benchmark against WebLogic JMS. I suppose there are two opposite reasons why that might be, but my conjecture is that its not worth the effort since there is no competetitive threat.

    What ever happened to DEC Message Queue that BEA bought ?
  39. Not Yet a X Killer[ Go to top ]

    We are way way off topic! To answer your questions:

    -- I wouldn't characterize Sonic or Fiorano as competing in the same space as BEA JMS (same story for JBoss). As for IBM MQ: yes/maybe/depends. I don't mean to disparage any of the above, its a matter of perspective.

    -- Java, JVMs, J2EE, and WebLogic have all come a long way since WL 5.1 days. WL 6.0 JMS was a rewrite of WL 5.1 JMS. And WL is now at version 8.1.

    -- DEC MessageQ is now branded BEA MessageQ.

    -- As for benchmarks, BEA does fairly well on a level playing field - win most, lose some. But, as you likely know, benchmarks are twisty easily biased things (and I'm biased myself) -- we always strongly encourage folks to write their own. If you're truly interested, I suggest reading BEA's "JMS Performance Guide" white-paper (dev2dev.bea.com).
  40. Not Yet a X Killer[ Go to top ]

    \Nicholas Whitehead\
    But the specific term "XA Logging" is a cloud.
    \Nicholas Whitehead\

    In the simplest, stripped-down terms, here's what XA Logging means/implies:

       - The app server generally controls the transactions we'll call the TM (transaction Manager).
       - Everybody else is a "Resource" e.g. database, JMS, JCA adapters, etc.
       - When a transaction starts going into the commit phase, the TM logs this fact before telling all the Resources to commit.
       - If the TM crashes/fails/whatever, the TM can use the transaction log to figure out what work may have been only partially completed at crash time.

    So - if a TM crashes for whatever reason, it uses the XA log to figure out what pieces it left in inconsistent state, and then fixes it on restart.

    TMs which don't have transaction logs don't do anything if they crash and restart. They pretend everything's OK. Meanwhile, your Resources may be in an inconsisent state (e.g. JMS commited, Oracle didn't), or just as bad, Resources like a database may be holding a lock that prevents other work from being done.

    People who worry about correctness and failure/recovery scenarios very, very much want a TM that does transaction logging and proper recovery. You can live without it if the data/system isn't all that critical, or if you're the world's biggest optimist and don't think your system's ever going to crash.

        -Mike
  41. Not Yet a X Killer[ Go to top ]

    Having said that, BEA's JMS implementation has never been more than an idle curiosity that has zero standing amongst dedicated commercial (and perhaps open source) JMS vendors. It is a convenience to people who have WebLogic and may use JMS for non-critical tasks and because they need it to qualify under the Sun specs.


    In my company we do use BEA's JMS implementation for mission-critical applications.
  42. Not Yet a X Killer[ Go to top ]

    Many people are skeptical because JBoss does not do

    > everything the commercial offerings do. Want distributed
    > transactions, XA Logging, a solid JMS implementation? Then
    > you go commercial. Marc has publically stated that there are
    > some things that aren't suitable to open source and that you
    > should buy a commercial solution for them.

    Does the JBoss Group recommend Arjuna+JBoss for mission
    critical enterprise applications?

    http://www.theserverside.com/news/thread.jsp?thread_id=23585

    http://www.arjuna.com/products/jboss/
  43. Not Yet a X Killer[ Go to top ]

    Does the JBoss Group recommend Arjuna+JBoss for mission

    > critical enterprise applications?

    It is certainly not mandatory for mcea. Based on the links you provided, it looks like it gives the JBoss environment distributed 2PC and fully distributed and transactional JMS support for those mceas that need that functionality.

    So, considering the two companies anounced a partnership, and assuming a specific customer needed D2PC etc. etc., I guess they would.

    The absence of distributed two phase commit heterogeneous resource message based transactions in system does not mean its not a mission critical enterprise application, does it ?
  44. good news. is there any plan to spend some of this money for development(coding) of JBoss? or is it mostly for marketing?
  45. Funny...[ Go to top ]

    Hasn't Marc Fleury been repeatedly crowing for the past year about _not_ using VC money, and about how evil it all is? What's changed? Why was VC money evil 3 months ago but not now? ;-)

        -Mike
  46. Ahahaha![ Go to top ]

    This is great news! It means that the VC's own his ass/soul, and will expect and demand a return on their investment. This will often involve screwing a variety of people. VC's aren't stupid, they'll want more than some siily Frenchman's soul for that sort of money.

    It also means that JBoss could not sustain itself and grow to the levels that the clan keeps bragging of, despite all the big words and furiously flailing limbs of their fearless leaders. Although, it'll be fun seeing Gavin pull out his polite-yet-serious-and-lets-all-get-along-but-actually-youre-all-idiots lines below at some point. I also look forward to a number of exclamation marks in a row by the original Fleury, and perhaps a 'I'm sick of you all' rantorama by Berk.
  47. Ahahaha![ Go to top ]

    Hani: It also means that JBoss could not sustain itself and grow to the levels that the clan keeps bragging of, despite all the big words and furiously flailing limbs of their fearless leaders.

    That's not necessarily true. If $10MM now makes them grow to $20MM run rate that much faster, then it's worth it. ;-)

    Time is money.

    Peace,

    Cameron Purdy
    Tangosol, Inc.
    Coherence: Clustered JCache for Grid Computing!
  48. Ahahaha![ Go to top ]

    Hani,

    I think you need to understand more about financial management before you make comments like "It also means that JBoss could not sustain itself and grow to the levels that the clan keeps bragging of, despite all the big words and furiously flailing limbs of their fearless leaders." This growth is about operating growth not just how many people use JBoss, and how widespread it is. It's a different thing entirely.

    Plenty of companies are profitable but go out of business do to cash-flow. At the stage that the JBoss group is in, Cash Flow is about 10 times more important than there income statement. When I do valuation work for startup companies I usually do a "value" based on discounted cash flows, since it is often more indicative than earnings. If you don't understand the difference, go get a finance book.

    More importantly, operating cash flow, revenues, earnings and retained earnings only allow you to grow at a certain rates based on assets and equity/debt (look up internal and sustainable growth rate). These guys obviously want to grow a lot faster than that, and should.

    Congrats JBoss.

    Jason McKerr
    The Open Source Lab
    "Open Minds. Open Doors. Open Source."
  49. Good stuff[ Go to top ]

    Well played Marc.
    You (and your team) deserve it.

    Thierry Janaudy
  50. Ahahahah[ Go to top ]

    Hani,

    >
    > I think you need to understand more about financial management before you make comments like "It also means that JBoss could not sustain itself and grow to the levels that the clan keeps bragging of, despite all the big words and furiously flailing limbs of their fearless leaders." This growth is about operating growth not just how many people use JBoss, and how widespread it is. It's a different thing entirely.
    >
    > Plenty of companies are profitable but go out of business do to cash-flow. At the stage that the JBoss group is in, Cash Flow is about 10 times more important than there income statement. When I do valuation work for startup companies I usually do a "value" based on discounted cash flows, since it is often more indicative than earnings. If you don't understand the difference, go get a finance book.
    >
    > More importantly, operating cash flow, revenues, earnings and retained earnings only allow you to grow at a certain rates based on assets and equity/debt (look up internal and sustainable growth rate). These guys obviously want to grow a lot faster than that, and should.
    >
    > Congrats JBoss.
    >
    > Jason McKerr
    > The Open Source Lab
    > "Open Minds. Open Doors. Open Source."

    Haha, you want Hani to lookup various things and you can't spell worth a damn?
    'do to cash-flow.', ' than there income ', 'I do valuation'. Good work.
  51. Ahahahah[ Go to top ]

    Patrik: Haha, you want Hani to lookup various things and you can't spell worth a damn? 'do to cash-flow.', ' than there income ', 'I do valuation'. Good work.

    Well, I for one welcome the $10MM investment in JBoss, particularly since it's not my money. Maybe it'll get Hani off his rump and blogging again. As for Jason, he probably does know his stuff, since he started (one of?) the first university open source labs in the States.

    It's not whether the $10MM is 'good' or not, it's how well it will be used. Time will tell.

    Peace,

    Cameron Purdy
    Tangosol, Inc.
    Coherence: Clustered JCache for Grid Computing!
  52. Ahahahah[ Go to top ]

    It's not whether the $10MM is 'good' or not, it's

    > how well it will be used. Time will tell.

    There are too many people who thinks that JBoss is a real business. They make only noisy like a real business. It's a hot air balloon which will burst in about 2 years. I bet they will never reach J2EE compatibility. They are too far behind. If they ever reach 1.4 compatibility, J2EE is at 1.6.

    Fleury doesn't care about JBoss. Come on! His big challenge now is how he can legally move as much as possible from the 10MM onto his private account. Legally! That's the point. That's the good'ol dot com story. May be he will invest a bit, yes. But not in a real business. Only in an IPO to get even more money and then exit. Legally. Fleury's dream is to have 1 more Ferrari than Mr. Chuang from BEA. His dream is not to stay in this f*cking OS business any longer than needed.

    Yawn.
  53. Ahahahah[ Go to top ]

    There are too many people who thinks that JBoss is a real business.


    I never doubted that it's a real business. The problem I have is whether it's a "real" open-source project. I have to give Marc credit though. He stands right up there as the anti-Richard Stallman :-)

    > They make only noisy like a real business. It's a hot air balloon which will burst in about 2 years. I bet they will never reach J2EE compatibility. They are too far behind. If they ever reach 1.4 compatibility, J2EE is at 1.6.

    My impression is that it has already burst, and that's why the finally are selling off stakes in the business. And now that they are owned by real business people certification will probably be priority one.
  54. Ahahahah[ Go to top ]

    Sheesh. Tuff kritiks. I uhpologise for poor grammer and speling.

    By the way, what's wrong with the phrase, "I do valuation work..." It looks normal to me. I help "value" non-public, startup companies, hence "valuation."

    Valuation: the act or process of valuing; specifically : appraisal of property

    Anyway, the issues are still important. Growth of the operations and finances of the group as opposed to the growth of the technology, the Open Source project or adoption of the application server.

    Jason McKerr
    The Open Source Lab
    "Open Minds. Open Doors. Open Source."
  55. Ahahahah[ Go to top ]

    It is interesting to see that JBoss Group LLC is now JBoss Inc., a corporation. And that apparently the JBoss (tm) Trademark is now owned by JBoss Inc. and not Marc Fleury any longer (at least it appears to be that way now).

    \Jason McKerr\
    Anyway, the issues are still important. Growth of the operations and finances of the group as opposed to the growth of the technology, the Open Source project or adoption of the application server.
    \Jason McKerr\

    I think it's interesting that many people, including some JBossians, seem to suffer under the illusion that JBoss is a product's company, and are comparing it to RedHat. In fact, JBoss is nothing more than a services company that happens to let people bang out free code :-)

    You can go out and buy Red Hat Linux, in various flavors - as almost a side effect, you can also get it for free, and they also do services. JBoss Inc., by comparison, doesn't sell any product. They don't "package" JBoss - developers on source forge just happen to give links to tarballs on a regular basis. Services != products.

    It should be interesting to see how a VC affects this equation. Will they push for JBoss to become more Red Hat-ian, and package and sell JBoss? Will they foster more of a "product" image? Or will they push the services side - which may or may not have a backlash effect on actual development.

    Indeed - how will the VCs react to the fact that most of the employees are "telecommutting"? Will the ugly phrase of "utilization rates" suddenly become a hot topic around the (virtual) water cooler? How well will Marc and Bill Burke's insistence that "we don't do consulting" fly with the men with the big bags of money?

    As you sort-of said, it's important to keep JBoss, the project, and JBoss Inc seperate in your head. They're obviously related, but what's good for one is _not_ necessarily good for the other.

         -Mike
  56. Ahahahah[ Go to top ]

    Mike: It is interesting to see that JBoss Group LLC is now JBoss Inc., a corporation. And that apparently the JBoss (tm) Trademark is now owned by JBoss Inc. and not Marc Fleury any longer (at least it appears to be that way now).

    No one is going to invest in a company that doesn't own its own trademarks ;-) .. and the incorporation is probably as a C class corporation, which is not at all unusual when soliciting an investment.

    Mike: I think it's interesting that many people, including some JBossians, seem to suffer under the illusion that JBoss is a product's company, and are comparing it to RedHat. In fact, JBoss is nothing more than a services company that happens to let people bang out free code :-)

    RedHat and JBoss are not that dissimilar. In fact, JBG employees probably wrote a much higher percentage of JBoss than RedHat employees wrote of RedHat. And RedHat doesn't sell a product, they simply charge for distribution and support. (JBG simply hasn't figured out how to sell a CD yet, but that isn't too lucrative anyway, compared to support. Sorry Mike ;-)

    Peace,

    Cameron Purdy
    Tangosol, Inc.
    Coherence: Clustered JCache for Grid Computing!
  57. Ahahahah[ Go to top ]

    \Cameron Purdy\
    No one is going to invest in a company that doesn't own its own trademarks ;-) .. and the incorporation is probably as a C class corporation, which is not at all unusual when soliciting an investment.
    \Cameron Purdy\

    Agreed and agreed. I didn't say this was bad or made no sense. It's obvious biznessification, and probably the right thing for the company. It also happens to fly right in the face of much of what the CEO has been saying for the past year :-)

    \Cameron Purdy\
    RedHat and JBoss are not that dissimilar. In fact, JBG employees probably wrote a much higher percentage of JBoss than RedHat employees wrote of RedHat. And RedHat doesn't sell a product, they simply charge for distribution and support. (JBG simply hasn't figured out how to sell a CD yet, but that isn't too lucrative anyway, compared to support. Sorry Mike ;-)
    \Cameron Purdy\

    I'm afraid I'm going to have to disagree strenuously, Cameron. I don't know Red Hat's financial breakdown, but they quack and waddle and shed water like a product duck. Go to the RedHat web site - PRODUCTS are primary. The first three icons on the web site - "Red Hat Enterprise Linux", "Red Hat Network", "Applications" - click on any of those buttons and see "Buy now buy now buy now".

    Services is the last icon, far on the right.

    How much effort has RedHat expended on packagers, installers, working on various pieces of PC hardware? I agree that RedHat hasn't written most of the Linux code, probably not even a meaningful fraction of it - but they _have_ productized it. Open source or not, they have QA teams and product managers. I walked into J&R Computer World the other day, and there were Red Hat packages prominently displayed in the OS section.

    This does not mean that Red Hat makes most of their money off of product revenue. It does mean that they've made a commitment to walk and talk and quack like a product company, with all the infrastructure that that implies. They may make a boatload of money off of services, but I'm willing to bet that much of that revenue comes from a very real image of being a product company that really knows what's what.

    JBoss does _not_ walk, talk, or quack like a product company. In any sense of the word. Their "product", literally, is a tarball that you can download, and then fiddle with various config files in your favorite editor. They do not have a QA person, let alone a QA department. There are no product managers. They don't - really, they can't - go out and "sell" JBoss the product, because there really is nothing there to sell _but_ a free tarball download.

    The difference seems silly but it's very, very real. When all you've got is a tarball on sourceforge.net, even senior management people manage to sense this and react to it.

         -Mike
  58. Employees are expensive[ Go to top ]

    No one has yet mentioned that JBG has employed a bunch of developers for coding/support and training. It cost money to build up an organization like that and if you dont want to bootstrap it, you get VC funding.

    The biggest change we will probably see is that Marc wont be able to become a 5-year old on the news groups anymore :)
  59. Employees are expensive[ Go to top ]

    No one has yet mentioned that JBG has employed a bunch of developers for coding/support and training. It cost money to build up an organization like that and if you dont want to bootstrap it, you get VC funding.

    >

    Yeah, no one has also mentioned that JBoss is profitable right now. At least, that is what the press announcement says. Having worked for a few startups in the past, getting up to what, 20 employees without VC fundings is quite a feat. The real test will be to see if the VC's force them to burn that money.

    > The biggest change we will probably see is that Marc wont be able to become a 5-year old on the news groups anymore :)

    Yeah, its eerily quiet on this post from JBossians. Where are the toddlers at now?

    /Joe
  60. Employees are expensive[ Go to top ]

    Yeah, no one has also mentioned that JBoss is profitable right now. At least, that is what the press announcement says. Having worked for a few startups in the past, getting up to what, 20 employees without VC fundings is quite a feat. The real test will be to see if the VC's force them to burn that money.


    Although the business model is what it is, I think Marc here should take the credit for the company being profitable. The man is a marketing machine. He keeps such a storm of controversy in the open source world that everyone knows about jboss. It's fascinating that bad publicity can make you money also.
  61. Ahahahah[ Go to top ]

    Agreed and agreed. I didn't say this was bad or made no sense. It's obvious biznessification, and probably the right thing for the company. It also happens to fly right in the face of much of what the CEO has been saying for the past year :-)


    I think one of the press releases above (or some other article I found) actually says that they changed to a corporation so they can go public.
  62. Why HE loves "Open Source"[ Go to top ]

    Part I - The bubble years

    Traffic light. Hot. Hot temperatures, Hot IT marked, money pouring everywhere. Damn SUN, makes me sweat. Money, action but ... I can get no satisfaction.

    Eureka: I have an idea: Develop (my own) App server to host (and charge for) clients' applications. Find VC willing to in bed with me. Nobody likes me (: People laughing at me… they'll see!

    No money, no clients, no prospects. What to do?

    I have (some) app server. I will call it an "open source project".
    Lure ("we will take on the big guys") developers to do the work for me.

    All in the name of Open Source.

    Part II - cold nuclear winter
     
    2000 …1 BAMMM!

     The bubble's burst. Poisonous gases cloud the IT continent. All VCs are dead.
    But, shhhht - do you hear- deep underneath the Earth surface, the Jboss group is … coding.

    HE can’t wipe off the smirk of his hace - "Ha, this Open Source stuff is great, I have people, effectively working for me for free. The Net is the great Equalizer - I chant, they work, Its my Jboss, they work. I know how to keep them happy - will chant few slogans. Let me think:

     " …Screw the VCs." This is a good one, shows I am not in for money".

    There is another one:

    "The most talented people work for me (oops, pardon me), read that "work for Jboss". Memo to self - careful in the future.
    “

    2001 - still cold

    But the gases are gone. We can go up there and be seen.

    PART III

    2002 - "JBossional Open Source"

    The legend grows big - there are 24 trillion downloads of Jboss.

    Java wannabes write "Hello Jboss" applications everywhere.

    Hopefull developers deploy, test and report bugs. For free.

    Architects and managers break off the chains of the "commercial middleware tyranny" and rush to the savior. Some of them, totally Jbossed, tearfully part with the 10 big ones salvaged from BEA and IBM licenses (you see, they only claimed three CPUs on their server) and sign up for support.

    Money! It is a trickle. But it is money!



    "JBossional Open Source" comes to life.

    Everybody sees the light - from now on all IT development will be based on Jboss.

    2002 - rebellion

    Great - bootcamps mushrooming everywhere, things are going great, more money, more professional open source. Gotta give some to the coders - hey kiddo come, come here... there - fetch $500. We are "professional" now. Screw the VCs.

    Damn, some idiots criticize… and that kiddo wouldn't keep his mouth shut. "Jboss is not what HE says it is" They all say that.

    Gotta do something, Make more noise.
    …..In the news: "Every person in the world have downloaded Jboss at least three times."


    2003

    And this just in … "Jboss will AOP everything from Bean to Router."

    Huaaa, who would believe it, more contracts, I better send everybody consulting, Oops, we don't consult, I repeat - WE DO NOT CONSULT. WE SUPPORT.

    Damn, these disco colors on the Jboss web site.

    Damn, this EJB stuff sucks, I'd better Hibernate.

    More contracts - more money.

    Hmmm, money, money, money, must be funny in the rich's men world
     
    Gotta do something. Gotta tell My story - "hey people, I love open source , I made tons of money before, I just don't need anymore, just open , just open, open …"


    2004

    Ohh, can't help it - money is the problem now - its never the right amount you have. Its either too much or not enough..

    "AOP, EJB, we are the champions, BEA is dying, no time for losers, IBM is lagging, we re the greatest…. Blah, blah blaaaah".

    Gosh , I am tired of chanting

    Wait, what is that - a VC on the hook, I'd better reel in carefully this time, if what to make it big…

    Yea, that’s a big one, look at the gills,
    Yes, yes - the most talented VCs work for us. The Net is the great IPO-aliser. Screw those naïve open source hidalgos, They work - I get rich, Yeah screw the developers, screw work, Charles Barkley said it right on Leno – “Work is overestimated” … find someone to work for you.”


    Yeaaah – I love open source.
  63. Total Market Value?[ Go to top ]

    I would like to see and estimate the current market value of JBossGroup might be,
    to understand who is going to control development from now on.

    For all the fancy ideas Marc has about doing only 3rd-level support, or whatever he calls it, now that they are a regular old startup it's definitely going to affect how they operate. They are probably just going to end up doing training and consulting ..
  64. I still can't believe a project of this size still uses sourceforge. Their anon CVS servers have been such crap for the past year that competent projects have taken to releasing tarballs of their source with every release. Maybe we can even get some basic documentation out of them. Just a description of all the config settings and how to use the JMX interface.
  65. I don't really see what's wrong with JBoss Group getting $10 millions... This is just a company following/adapting its strategy to survive/live in a tough competition. Marc Fleury is certainly provocative, very much in the vein of Larry Ellison, and most people tend to believe that his attitude is plain arrogance. Whatever... This is all a show. :) I don't think we should take all that quacking-around too seriously, or too personally. Politics is just a game. Look beyond that... If you like the product and service for your needs, just use them. This is not like supporting a tyranic regime. We should avoid resorting to pigeonholing people or products out of shallow impressions.

    Whatever one thinks of Marc Fleury, at least, he has worked hard, taken risks, shown initiative, which are things the vast majority of us who toughly criticize him did not do. I am happy to see that tough work can still lead to some form of awarding. I just hope that they do the best with this money and I wish JBoss well in this very tough market.

    In Mr Purdy's very wording: peace! :)

                    Yann
  66. Well said Yann, I agree with what you have written.

    Marc and company deserves much more credit than some people give them. As a heavy JBoss user, I have certainly been a beneficiary of the contributions of the JBoss team, without having paid a single cent other than purchasing documentation that costs even less than books in the bookstore.

    As someone who have set up businesses, I've experienced first hand how businesses cost money to set up and run. Without outside investment, all that will have to come out of the founders' own pocket, and if don't have fat saving accounts, that could really hurt.

    These guys are talented folks and could've been making good money elsewhere, and I'd rather them be working on JBoss rather than anything else ;)
  67. Take out the magic words "JBoss", "Open source", and "Marc Fleury" and just look at the business for a moment. Frankly, it's rather mundane. Someone started up a little IT services company and grew it to around 20 people - wow, amazing! Astounding! Last time I checked there were several thousand such companies in the U.S. alone.

    Too many people buy into Marc's fiery rhetoric and get blinded to what JBoss Group actually was, and some of the comparisons are really funny. A Red Hat or BEA probably each employ more technical documentation people than are employed by the entire JBoss organization. One large contract in such a company probably equals JBoss' 2003 total yearly gross.

    It's nice that JBoss has gotten a fat capital injection, and I'm sure everybody employed therein are quite happy right now. Hey, good for them.

    What'll be interesting to see is how they use that money, and what sort of control the VCs assert over the company. In these sorts of deals everything is not always as it seems, and capital injection can be a good thing, or it can be disastrous. A company I worked for was private and organically grown with mostly their own money over the course of around a decade, and then suddenly got a $50MM shot in the arm by some very, very big financiers. Explosive growth followed that, and then mass hemorraging of money and the eventual massive contraction. They were even hot to IPO, but withdrew it when the bubble burst. The kicker we all got was from the SEC filing - everybody in the company got to find out from there that the partners took home about $43MM of the original investment - that money paid to buy their shares out - and that very little was invested in the company itself.

    It just goes to show that not everything in private company financing is what it seems. It should be very, very interesting to see how JBoss Inc. differs and diverges from JBoss Group LLC. It could very well be a good thing. Or not.

         -Mike
  68. Ars Digita Revisted[ Go to top ]

    This reminds me of good old ars digita. A eloquent and arrogant company owner with a sun sustainable open source model taking money from VCs. They changed technology, business model and management. I wouldn't be much surprised if it were to end the same way as well, maybe JBoss becoming NBoss and selling inferior .NET client side components.....

    Regards, Karl
  69. this guy is unbelievable, I tell you. I do not think I have ever met anyone quite as bull headed as this guy. Even if he is dead wrong, he believes he is right. Perhaps, Mr. Spille should be in the legal profession instead of living his glorious life in a cubicle. However, he does have a lot of hate deep down hate for JBoss. So, I am in the process of doing a little experiement. First off, I am going to call his so-called employer and see if there really is a Mike Spille there. If not, I have a strage suspicion that he could be one of the defectors of JBoss just can't grasp with the fact that the project has continued after their fork. My hypothesis could be wrong, but curiousity is killing me. I have a hard time seeing people so hateful in the world.

    I am not saying the JBoss guys have never done anything wrong or that they are saints by no stretch of the imagination. That fleury has definitely put his foot in his mouth on more than one occasion. He has been known to be hateful and revengeful. However, this Spille is consistent with his hate for JBoss and being curious I am going to find out if he really exists...
  70. First report...[ Go to top ]

    There is no Mike Spille at SIAC, which is the company he said he worked for in a previous post a while back. He said he worked for the company that developed the technology for the NY Stock Exchange. There is a chance that he is not a subscriber to this system. Now the second test, calling information for Spille in NYC...
  71. First report...[ Go to top ]

    \Arun Patel\
    There is no Mike Spille at SIAC, which is the company he said he worked for in a previous post a while back. He said he worked for the company that developed the technology for the NY Stock Exchange. There is a chance that he is not a subscriber to this system. Now the second test, calling information for Spille in NYC...
    \Arun Patel\

    This is interesting on several levels, Arun. First, I most assuredly do work for the company you mentioned, and I let it slip once by accident in a posting (which I'm not supposed to do). Feel free to e-mail me, it's 'mspille' sans quotes. Sorry, but I'm not going to give out my private business number to the likes of you.

    Secondly, I really wonder how you determined anything about who works where on a Sunday. Who exactly did you call that would be in on a Sunday?

    Thirdly, most companies in NYC will not give out employee information over the phone.

    Fourthly, several people on TSS have had business dealings with me. Since the proverbial cat is out of the bag, people please feel free to tell people who I am and who I work for.

         -Mike
  72. \Arun Patel\
    this guy is unbelievable, I tell you. I do not think I have ever met anyone quite as bull headed as this guy. Even if he is dead wrong, he believes he is right. Perhaps, Mr. Spille should be in the legal profession instead of living his glorious life in a cubicle. However, he does have a lot of hate deep down hate for JBoss.
    \Arun Patel\

    I re-read my posts on this thread a couple of times, and I didn't detect any hate, and certainly don't feel any. Just an attempt to inject a bit of reality.

    \Arun Patel\
    So, I am in the process of doing a little experiement. First off, I am going to call his so-called employer and see if there really is a Mike Spille there. If not, I have a strage suspicion that he could be one of the defectors of JBoss just can't grasp with the fact that the project has continued after their fork. My hypothesis could be wrong, but curiousity is killing me. I have a hard time seeing people so hateful in the world.
    \Arun Patel\

    My resume and background are freely available on the web, via my Blog and also my own web site. Feel free to read it, I assure you it's accurate.

    As for my current employer - several regular readers of TSS have had business dealings with me at my place of work, and have already vouched for my identity on another thread.

        -Mike
  73. Yes, sir I have seen your resume and website, but I still do not buy who are who you say you are. Who are these people that can vouch for you. I have run a search for technovision and found nothing....
  74. WTF?[ Go to top ]

    Technovision's web site is here:

        http://www.etechnovision.com/

          -Mike
  75. Off topic and bs[ Go to top ]

    I can personally vouch for Mike Spille. He *is* who he says he is. Now, Arun Patel on the other hand I am not sure about. Where do you work Arun? Can I call them and verify who you are?

    Don't cry wolf.

    Dion
  76. Off topic and bs[ Go to top ]

    I work for wipro in Bangalore, India.
  77. Off topic and bs[ Go to top ]

    Dion: I can personally vouch for Mike Spille. He *is* who he says he is. Now, Arun Patel on the other hand I am not sure about. Where do you work Arun? Can I call them and verify who you are?

    I know Mike. Rob Misek, who's one of our sales guys (covers NYC) has talked to Mike a couple times (at Mike's place of employment, which I have agreed not to disclose.) Mike even stood me up for lunch, which proves he's a real New Yorker ;-)

    However, no one has actually ever seen Mike Spille and Rickard at the same time, so it's possible there's some Clark Kent/Superman thing going on.

    As for Arun, I've point-blank asked him before who he was without response. Whatever ..

    Peace,

    Cameron Purdy
    Tangosol, Inc.
    Coherence: Clustered JCache for Grid Computing!
  78. Off topic and bs[ Go to top ]

    Thanks Dion, Cameron. It seems odd to thank someone for merely acknowledging your existence, but it's still appreciated.

    We _will_ get lunch or another meal-like function one of these days, even if I have to take the extreme and dangerous trek to Boston to do it.....

    As for the rest...linguistical analysis of my posting patterns will prove that I do not boast any scandanavian based language as my mother tongue. Simply put, my English is too poor.

         -Mike
  79. Off topic and bs[ Go to top ]

    I know Mike. Rob Misek, who's one of our sales guys (covers NYC) has talked to Mike a couple times (at Mike's place of employment, which I have agreed not to disclose.) Mike even stood me up for lunch, which proves he's a real New Yorker ;-)


    Easy there :-)
  80. Second Test...[ Go to top ]

    Called information in New York city for Mike Spille and there is no listing. Once again, it is possible he is not listed, but even if he is not listed I would have received a recording that says, at the customers request, this number is not listed. But they could not find a Mike Spille within 30 minutes of NYC. According to his blog, that is where he lives and works.

    Now I am really starting to get suspicious.
  81. so sorry to mike...[ Go to top ]

    I apologize to Mike. I was just suspicious, as it seems he always has an agenda against JBoss. Sorry for taking this off topic, althogh it seems as it already was as we were talking about JMS.
  82. so sorry to mike...[ Go to top ]

    No problem, Arun. I am interested though in how you so conclusively determined where I do or don't work on a Sunday. Doubly so in a phone search of NYC - given that I have a cousin on the upper east side of NYC with the name Michael Spille (Michael G. Spille, IIRC) who _is_ listed in the phone book (FYI my number is unlisted and was registered by my wife).

    Please give me an e-mail at work tomorrow, I'd just _love_ to chat about this in detail.

        -Mike
  83. so sorry to mike...[ Go to top ]

    It is monday for me Mike...
  84. so sorry to mike...[ Go to top ]

    Arun Patel:
    > It is monday for me Mike...

    It might be Monday for you but this still doesn't explain how you managed to call someone in the US on a Sunday who would tell you that Mike doesn't work where he claims he works.

    --
    Cedric
  85. so sorry to mike...[ Go to top ]

    voice mail system, I guess they do not have these in France...
  86. so sorry to mike...[ Go to top ]

    Arun:
    > voice mail system, I guess they do not have these in France...

    Now that was very funny, congratulations. But let's get back to the point.

    By all means, please give us the phone number of that voice mail you called that will tell us that Mike Spille doesn't work at the company he claims he works for.

    Or did you make these phone calls in your imagination?

    --
    Cedric
  87. so sorry to mike...[ Go to top ]

    I want to get this straight - you called information in NYC from Bangalore, India trying to find my home phone number? And used "voice mail" to find out if I worked for a particular company - I'm curious, what number did you call (again, all the way from Bangalore, India)? To the best of my knowledge our voice mail system does not have a directory service (I work there and ought to know).

        -Mike
  88. so sorry to mike...[ Go to top ]

    Did anyone ever hear of VOIP, long distance calling is dead, there is not cost. you think just because I live in a country like India we can not call NY on a sunday? it is not hard to dial +1 212-555-1212 and ask for a listing. In india we also have things like phone mail attendents, and yes we did learn this in call center 101. You can dial by name.
  89. Arun Patel is a phony[ Go to top ]

    Hey Mr Patel:

    I don't think Azim Premji would be too happy to hear his employees spending Wipro's $$$ to make personal international directory enquiry calls. All your posts so far have been pure BS (it may be Monday for you, but it could still be Sunday in many parts of the world, including NYC). Didn't they teach you that in Call Center 101?

    It doesn't take long for us to call Wipro's HR and alert them to your actions (assuming you actually work there as you say you do and are not holed up in a motel somewhere in NJ).
  90. so sorry to mike...[ Go to top ]

    It is indeed Monday for you Arun - at the time of that post it was around 7am on Monday for India, right? Wow, and your earlier posts in this thread were at like 3am in India. Working the night shift?

        -Mike