JBoss: A modern day plague?

Discussions

News: JBoss: A modern day plague?

  1. JBoss: A modern day plague? (181 messages)

    Hani has gone off on an interesting rant about JBoss. He discusses the leader, core developers, various claims, the spat with Sun, and the "fanatics" (many of whom probably don't like this post :).

    Hani's JBoss Rant

    Rickard points to this rant, and discusses the AOP side of JBoss 4, mentioning some weeknesses that he sees.

    Rickard's AOP comments

    ***Editorial Apology/Comment*** - Posted June 18th, 11am EST
    TSS would like to apologise for the flame bait that was linked to in this post (Hani's rant). It violated our own rules, and was unnecessary.
    We have no interest in personal attacks, and due to haste the whole rant wasn't even fully read. The purpose of this post was to try to get a discussion going on JBoss AOP, and AOP in general: i.e. a technical discussion.
    Again, we apologise and will try not to make mistakes like this in the future.

    Threaded Messages (181)

  2. JBoss: A modern day plague?[ Go to top ]

    If you look at Hani's other blog entries, it seems that he is a professional ranter. :)

    Floyd
  3. Standards?[ Go to top ]

    Speaking of, they ought to apply to journalism as well. Posting this crap is a new low, even for TSS, the news sluts. What's next, Chiara on your editorial board? Of course, she'd probably have to submit a picture. Hani, alias Zoolander, alias the technology stud has replaced the metric of "not quite J2EE" with "not quite beautiful enough."
  4. Thank you MS, Cisco, ... -> No!!
    Thank you for Weblogic, Websphere - > No!

    THANK YOU for JBOSS, thank for the hard work!!!!!!!!!!!!!!!!!!

    I don´t know Rickard, what it is he doing? Why is he so stupid. There are so many ugly people on this planet .... be happy Rickard that someone is doing something. JBoss does a lot. you are not fair.

    JBoss: GREAT WORK! Rickard make it better if you can. Rickard do something and especially do not annoy.
  5. Professional ranting?[ Go to top ]

    I'm just amazed how low this rant is. Describing peoples appearances and calling JBoss users sluts is waaaaaay out of line. If this guy lives in the States, then I've have heard of something called Anger Management, but if he lives in Europe, he'll get his a** kicked in any pub with this attitude :))
  6. Professional ranting?[ Go to top ]

    Rikard seems to be quite bitter.
  7. Professional ranting?[ Go to top ]

    Judging by the dates on some of his posts, I don't even think he gets to see the inside of a pub! :)
  8. Professional ranting?[ Go to top ]

    Judging by the dates on some of his posts, I don't even think he gets to see the inside of a pub! :)

    >

    Errh, who, when? Timezone etc...
  9. Blog entries[ Go to top ]

    I think it is a good idea that TSS is linking to personal blogs to get some discussion started. There have been good discussions in the past, e.g. when Cedric was linked after he started the AOP challenge.

    Hani's blog, the BileBlog is of course more controversial, but nevertheless there could be productive results coming out of this (use 'Mark noisy'!!)

    As for TSS editors, I wish you'd not link directly to freeroller but just paste your blog or link to javablogs...Freeroller has been down to often in the last days and I fear the TSS traffic has something to do with it. This also applies to roller users themselves: I know you'd like to be on top of the index.jsp, but what's it worth when the app goes down? Thank you.
  10. Noisy[ Go to top ]

    As for TSS editors, I wish you'd not link directly to freeroller but just paste your blog or link to javablogs...Freeroller has been down to often in the last days and I fear the TSS traffic has something to do with it.

    You mean freeroller has been theserversided?

    --
    Cedric
    http://beust.com/weblog
  11. JBoss: A modern day plague?[ Go to top ]

    So you're asking us to respond in a non-fanatical way to what is accurately described as a "rant?" :)

    Here's my argument: We're JBoss customers. It suits our purposes very well. We have been very pleased with the level of support. It's free. We can cluster as many machines as we want without worrying about license fees. We can open up data centers in Europe and Asia without spending a million dollars on license fees.

    I have some more where that came from.

    Steve
  12. I asked a lot[ Go to top ]

    I know that I asked a lot (since hani's comments were built to make people come out swinging) but you stepped up to the challenge... didn't shout, and just told everyone the facts. Great job.
  13. JBoss: A modern day plague?[ Go to top ]

    I suspect him of working either at BEA or IBM as he says 'Finally, you TSS sluts need to realise that not everything is about JBoss. Stop trying to drop its name into every damn conversation.'. Looks like his job is in danger.

    When reading 'Describing JBoss as a J2EE application server is like describing a bathtub as a bathroom.' makes me think he isn't worth my time. Yes, JBoss is a J2EE Application Server, either give some proof that it is not or quit shouting dumb things.

    It looks like he only want to step on people, pretty pittyful. I am more interested in his, jet to appear, second rant about the technical negative aspects of JBoss. Then we can judge him better, all I can say about the guy now is that he his a bit of a hateful person.
  14. JBoss: A modern day plague?[ Go to top ]

    I suspect him of working either at BEA or IBM


    Actually it looks like he is an avid Orion fan.

    >as he says 'Finally, you TSS sluts need to realise
    >that not everything is about JBoss. Stop trying to
    >drop its name into every damn conversation.'.

    Funny enough, useless rants like these is what keeps JBoss in the headlines here on TSS. It is kind of obvious when you think about it. But then reading Hani's poorly argumented rant proves that 'thinking' is not an activity he spends too much time on.

    I do find it sad though, and kind of pathetic, that there is a group of people out there who waste their time and energy to such a negative activity. I would think it would be better to channel that same energy to something more productive, use it to create something new, and then maybe get a little bit of success for yourself too.

    But as it stands, jealousness is such an ugly trait in human character it makes you do a lot of useless things.

    Anyway, I'm rather comfortable with using JBoss, it follows the specs, it assumes you know and understand the spec but it works as expected. And it is free of charge which for me is a big deal. I would not pay for an J2EE application server ever. And with JBoss I don't need to. Thank you JBoss.

    /T
  15. Not one dig was made against the product. Only an analogy to a bathroom with no crapper. So I guess I have to agree that the jboss team is ugly, they dress funny have a crappy website, and their marketing slogan is trash.

    Perhaps if this character bothered to comment on the product I would have more to agree with. Why does tss even link to this there is not a real argument anywhere on the page. When I post vile flame bait, at least I bother to make a few valid points and speak my mind in between the mad raving.

    It takes all the fun out of the reply, when you agree with everthing that was said, but it makes no difference, because nothing important was said.
  16. Sorry, i found part 2, and see that it does comment on how hard jboss is to use. Can't say i disagree, but your own ignorance is no excuse to hate a product.
    Bad error messages and runtime rather then load/deploy time errors do suck... So I guess I agree with most of these complains, but again there are not many, and they are not very strong.
    But **** it.. at least he tries, and I respect anyone with a negative opinion.
  17. Instead of being a jboss fanatic she is a jboss basher. What a waste of time.
  18. Jboss[ Go to top ]

    I don´t know what´s so wrong about Jboss. Why is it not a J2EE Server? I don´t know if it is only for me, but it works nice. Even without the docs (at least while I didn´t have classloader issues, nor ejb issues).
    Technically, I think it's a full fledged J2EE server. It´s not the fastest around (I think Orion/Oracle is), but it can handle the job.

    However, I have to agree about Mr. Fleury and his staff behavior. They do not inspire confidence with its infantile behavior.
  19. Jboss[ Go to top ]

    I don´t know what´s so wrong about Jboss. Why is it not a J2EE Server?


    Let's turn that one around. How do you know that it is a J2EE server?

    /Rickard
  20. Wrong question[ Go to top ]

    I don´t know what´s so wrong about Jboss. Why is it not a J2EE Server?


    >Let's turn that one around. How do you know that it is a J2EE server?

    Why do I care?

    Really interesting questions are: Can I build nice applications with JBoss? Will it be easier/harder than with WebSphere/WebLogic?

    P.S. lets see some photos of pretty-boy Hani, for comparison.....
  21. Wrong question[ Go to top ]

    Let's turn that one around. How do you know that it is a J2EE server?

    >
    > Why do I care?

    Well, let's say (hypothetically) that you are a developer who want to deploy J2EE applications. In that case you might be interested.

    > Really interesting questions are: Can I build nice applications with JBoss?

    Excellent question! And it is kind of relevant too! The answer is "depends". If you want to develop "JBoss applications", then that will work just fine. If you want to develop "J2EE applications", and run them on JBoss, then the answer is "maybe".
    If you happen to only use the J2EE subset that JBoss implements you're ok, but if you want the whole J2EE stack, then you're out of luck. Maybe. Why maybe? Because noone really knows exactly what subset of J2EE JBoss implements. Why? Because noone has tried to certify it yet.

    /Rickard
  22. good answer[ Go to top ]

    Let's turn that one around. How do you know that it is a J2EE server?


    > > Why do I care?

    >Well, let's say (hypothetically) that you are a developer who want to deploy J2EE >applications. In that case you might be interested.

    Cool. Well, thats not me :) _I_ merely want to deploy applications.

    Applications that implement *user* requirements, not Sun requirements.

    Never used JBoss - but with all the hate being directed toward the JBoss team, they must be doing *something* right :) So I'm looking forward to a chance to use it in a real project!
  23. good answer[ Go to top ]

    Ahh Gavin,

    That may be just wonderful for you, and you could probably use .NET too. But one of MY consistent customer requirements is that my systems work an multiple different app-servers. Every product out of NACSE is tested on at least 5.

    To some of us, and many of our clients, choice really *is* important. And if I can offer my customers more choices....

    Even more, I had a large government project deploy on on server (open-source). They found that the server only marginally met their needs. Switched to a commercial server in no time (a day or two), and everything settled in nicely.

    Jason McKerr
    Northwest Alliance for Computational Science and Engineering.
  24. Ahh Jason[ Go to top ]

    That may be just wonderful for you, and you could probably use .NET too.



    Certainly; and I would if it .NET was open source.

    Unfortunately the .NET initiative is controlled by a vendor only marginally better able to produce good software than Sun.


    >> But one of MY consistent customer requirements is that my systems work an multiple different app-servers.


    The way _I_ achieve this is by using an infrastructure layer that abstracts my business logic from the underlying (and rather badly designed) J2EE APIs. I'll develop using open source toolkits and components like Struts, Maverick, Hibernate, log4j, etc.
    I have a command invocation infrastructure that lets business logic run transparently inside a session bean, or asynchronously using JMS. My application code can even run *outside* an application server, in a batch process, for example. (Now THAT is what I call platform-agnostic.)

    Of course, my application is completely portable between different J2EE-ish application servers with minimal changes.

    >> Even more, I had a large government project deploy on on server (open-source). They found that the server only marginally met their needs. Switched to a commercial server in no time (a day or two), <<

    So, all this alleged non-spec-compliance caused your project a setback of a whole *two* days porting time.

    Hmmm. I'm sorry, but from *your* experience it should be quite clear why I'm not taking this whole "JBoss is not J2EE" meme very seriously.
    I am quite surprised that serious people like Rickard are jumping on this particular bandwagon. This stuff is for marketing people, not developers.

    Gavin King
    http://hibernate.sf.net
  25. Ahh Jason[ Go to top ]

    Gavin,

    That sounds like an interesting solution, and I especially like the batch process idea. I also try to use components for the API's that are badly designed, in fact most of the ones you mentioned.

    However, the batch process and session beans sounds like focusing on the wrong problems to me.

    My focus should be on the business problem (or in my case the research problem). I shouldn't have to have to create work-arounds for poorly designed application servers, especially when they claim to be J2EE compliant. If they make that claim, it should be a non issue. If they don't make the claim, that's a different story.

    It is a waste of my time (and my clients' money) to create externalities that make up for app-server problems. I'll just find another app-server. If the damn thing runs fine on 5 or 6 app-servers, but not on one, no biggy. I've still got the others (and I try to make sure that at least one of the open-source one's is open-source).

    Jason McKerr
    Northwest Alliance for Computational Science and Engineering
  26. Ahh Jason[ Go to top ]

    Yeah, that least sentence of mine made a lot of sense.

    Jason
  27. Wrong question[ Go to top ]

    Rick,

       I do agree with your points on use of AOP. It makes things not portable(which may not be a problem if you stick with only JBoss. But the real life is not so!!).

    And JBOSS is not J2EE compatible. Yes, but even if you port applications between two J2EE servers, the life is not going to be as easy as expected. (say, WebSphere and WebLogic).

    But still, i could not agree with Hani saying JBoss is a plague. (yes, its disturbing all the revenue models, and would give bad name for J2EE servers among customers, if they fed up with JBoss, and could end up accusing all the J2EE servers).

    But remember, nobody is just going to jump on something, just b'cos its free. 90% of the time, a serious consideration is taking place (financials, technical $$$).

    And nobody is going to jump on JBoss b'cos it got AOP,JMX, etc,etc.

    The question is simple. How reliable (techical, support, financial strength of the vendor, integ. with legacy systems,platform support, consultant availability, and finally support for CMA).

    The choice depends on the company.

    We have to note a serious point.
    Despite JBOSS is free, and is being around for years, still WebLogic and WebSphere is leading the market.

    Its not good to say JBoss is a plague, just b'cos its not J2EE compatible. After all its written in Java.

    Just my 2 pence.
  28. Jboss[ Go to top ]

    I deploying our j2ee apps everyday
  29. Jboss[ Go to top ]

    Hmm. Could you possibly put together a list of refactorings instead of this warmer/colder stuff. I assume you are critical of the implementation and not the concept. Let's also assume this is a 1.0 release - or even beta. FUD is not constructive.

    Jboss is clearly taking a chance with AOP but I would call this leadership instead of architectural suicide. Hell, if they went through the JCP how long would this have taken.

    cheers
  30. Jboss[ Go to top ]

    Hmm. Could you possibly put together a list of refactorings instead of this

    > warmer/colder stuff.

    I'm sorry, I don't understand the above. Is JBoss a J2EE server? If yes, then how do you know so?

    > I assume you are critical of the implementation and not the concept.

    I'm just wondering whether the alleged OpenSource J2EE server JBoss actually is a J2EE server. That's all. (I'm trying to keep it simple here)

    > Let's also
    > assume this is a 1.0 release - or even beta.

    JBoss is now up at 4.0. That's not a beta, or even an early release.

    > FUD is not constructive.

    What in my post was FUD? (to be precise, please explain what in it is Fear, what is Uncertainty, and what is Doubt)

    > Jboss is clearly taking a chance with AOP but I would call this leadership
    > instead of architectural suicide.

    What does this have to do with the question I posed?

    > Hell, if they went through the JCP how long would this have taken.

    How is this relevant to the question you are responding to?
  31. Is JBoss 4.0 Beta?[ Go to top ]

    JBoss is now up at 4.0. That's not a beta, or even an early release.


    It's called "Developer Release 1". I am not quite sure what that is supposed to mean, but I could not find any statements on their web site saying it's recommended for production, in contrast 3.2 is called "Production Release".

    From what I read, I consider 4.0 to be beta and not meant for production environments yet.

    I think they were just pressed to release some code because of JavaOne.
  32. Jboss[ Go to top ]

    Sorry, I replied to the wrong message. You have so many today. My post was in response to message: 86306. In that message you suggest people download and trace JBoss AOP calls. If you have suggestions for improving JBoss AOP please post them. I have no motive in favor of JBoss or AOP but I don't see where you are going with your reasoning.

    As for JBoss's J2EE-ness.

    We can bend the fork when we realize it is not there. In the same vein, you can do whatever you want with JBoss when you realize J2EE certification is not there.

    Business perspective:
    Is it certified? - No. Is it trying to pass certification? - Fleury says yes if someone else pays for it - (so let's assume No). CDN doesn't seem to have a published position nor venture capital to pay for it.

    Technical perspective:
    Does it adhere to the J2EE spec. I don't hear anyone complaining. Can I deploy all EJB types to it - yes. J2EE services - yes. Performance - acceptable.

    User perspective:
    Value - if I use the formula standards+performance+maintainibility/$$$ then my calculator blows when I divide by zero.
  33. JBoss[ Go to top ]

    Let's turn that one around. How do you know that it is a J2EE server?


    Because when I take my J2EE app and deploy it on JBoss it works. Duh. That's the only part I care about.

    /T
  34. Jboss[ Go to top ]

    You are correct Rick.

    If somebody claims they got a J2EE server, i would expect them all the things J2EE mandates.

    Omitting a or b, may not give a problem when u r sticking to JBoss. But it could break everything, when u r *deploying* in some other server.

    And JBoss AOP.. no comments. I could not afford to have a very slow response, or a lot of objects being created. (In java, object creation is a costly operation, as for as I know.)
    I dont understand what this buys me anyway.
  35. We know it's not a J2EE app server because J2EE app servers are licensed from Sun and tested against the J2EE certification suite. JBoss hasn't been licensed or tested. It's J2EE-compatible for the most part, not "J2EE compatible."
  36. JBoss isn't a J2EE server.[ Go to top ]

    We know it's not a J2EE app server because J2EE app servers are licensed from

    > Sun and tested against the J2EE certification suite. JBoss hasn't been licensed
    > or tested. It's J2EE-compatible for the most part, not "J2EE compatible."

    Precisely. So, why is everyone calling it "J2EE server", when it really should be "for the most part J2EE-compatible server"?

    I think I can sum it up in two words: wishful thinking.
  37. JBoss isn't a J2EE server.[ Go to top ]

    Sometimes it's advantage...
  38. hmm J2ee server huh?[ Go to top ]

    There are several J2ee servers that are certified and do not repeat do not or cannot run Sun' sown exampel sout of box..so what is your freaking point?

    let me remind everyone the J2ee spec is the following:

    J2ee spec fomr jcp.org+ vedor Implementatiosn=J2ee true spec..

    Anyone who does not understand this needs to do more J2ee deployments then get back to us on this issue!
  39. hmm J2ee server huh?[ Go to top ]

    let me remind everyone the J2ee spec is the following:

    >
    > J2ee spec fomr jcp.org+ vedor Implementatiosn=J2ee true spec..

    No, it's not. The "J2EE spec" is a document that describes the behaviour of application servers implementing it, and also refers to a number of other specifications that has to be properly implemented. For anyone to be able to call a server "a J2EE server" they have to:
    1) Implement the J2EE spec
    2) Implement all referenced specs
    3) Validate through certification that 1 and 2 have been accomplished

    > Anyone who does not understand this needs to do more J2ee deployments then get
    > back to us on this issue!

    I'm afraid you're the one who got it all mixed up.
  40. hmm J2ee server huh?[ Go to top ]

    Rickard,

    I agree with your JBossAOP statements. I do not know the J2EE spec. and JBoss implementation as much as you do. Can you please list couple of JBoss' incorrect or none-spec J2EE implementations?

    thanks.
  41. hmm J2ee server huh?[ Go to top ]

    I don't think that Rickard said that JBoss was "out of spec compliance". What he was saying, and stop me if I misunderstood, is that there is really no way to be 100% sure that JBoss is 100% percent J2EE-compliant. Not without the certification.

    Which isn't necessarily a big deal *depending* on what your needs are. I think CIOs of large companies are not going to accept JBoss' validity in the enterprise space without the certification, but there are thousands of IT managers and developers out there who need a solid "J2EE-compatible" app server that performs decently and doesn't show up as a 4 or 5 figure number on their budget reports. JBoss fits the bill perfectly.

    Is JBoss' AOP implementation sub-standard? Well, what IS the standard? In any case, AFAIK, you are not forced to use any of it.

    Sandeep
  42. Enough[ Go to top ]

    Resin is $500 and works.
    Orion is $1500 and works.
    Jboss is $0 without documentation BUT _doesn't work_.

    What is there to debate over? How much does _doesn't work_ cost? You gimps keep forgetting that. Free license != free. Let me say that again, no wait, read that again! Once more! No, read it again! Get it now?

    How much money do development companies spend on br0ked assed app servers these days? How many can say that Tomcrapola or JBoss is free? Sure, I get a handful of gravel for free too, but turning it into a nice house would cost a bunch!

    I know you JBoss/Tomcrapola zealots don't understand that but it feels better now that I've said it.
  43. Enough[ Go to top ]

    <quote>
    > Jboss is $0 without documentation BUT _doesn't work_.
    </quote>

    Just so that we're all clear about this, what does "_doesn't work_" mean? Does it mean that
    a) You could not start up the server?
    b) You could not deploy your application and get it running?
    c) You ran into chronic, non-replicable (!) errors that baffle you and your IT team?
    d) You ran into subtle problems with your deployment that stymied the JBoss developers?
    e) Something else?

    IMHO, a statement like "Jboss is $0 without documentation BUT _doesn't work_." doesn't even qualify as FUD. Its just silly without some sort of relevant substantiation.

    Sandeep
  44. I have a doubt kinds.[ Go to top ]

    I'd say: all of the above. Did you read fate's rant? Are you implying he makes that up?

    kinds.
  45. I have a doubt kinds.[ Go to top ]

    I'd say: all of the above.


    Really? Hmmmm. OK. Riiiight.

    > Did you read fate's rant? Are you implying he makes that up?

    God forbid that I imply that "fate" makes up anything. But are YOU implying that your posts were solely based on "technical facts" gleamed from reading his "rant"?

    Sandeep
  46. Doubts![ Go to top ]

    And again, where are _your_ facts? I had a few points, which is it that you refute?

    -> Free license == free.
    -> Jboss doesn't work
    -> Orion works
    -> Resin works
    -> Free gravel != Free house.
  47. None of the above[ Go to top ]

    I have built a commercial shipping J2EE based product with JBOSS. I had none of the problems and I got nothing to complain about JBOSS.

    Pranab
  48. I have a doubt kinds.[ Go to top ]

    I'd say: all of the above. Did you read fate's rant? Are you

    > implying he makes that up?

    So you're basing your opinion on someone else's rant (who is fate?).

    Way to build credibility there, buddy! You made a believer out of me in no time!

    /T
  49. I have a doubt kinds.[ Go to top ]

    so, actually, you're saying:

    f) You don't know.

    Do you even know what an application server is? Or J2EE?
  50. Enough[ Go to top ]

    Patrik - I noticed your response was that you have had issues with a-e in this post. You should be specific with your own experiences.

    I use JBoss and Orion for the most part. For my applications on JBoss I can:
    * Start the server
    * Deploy and run my J2EE applications.

    I do not:
    * Run into chronic, non-replicable errors that baffle me and my IT department
    * Run into problems that stymie the jboss developers

    I don't run into problems. Are our applications major 24x7 heavy duty clustered billion transactions/sec apps? No - but we interact with a wide array of data sources, employ EJB (CMP, SLSB, SFSB), etc. The apps have to perform pretty well too.

    I don't have it in for JBoss, nor do I care about Marc Fleury - I've never interacted with him and doubt much that I ever will. If it didn't work, I wouldn't use it. Is AOP the way to go? Is it too much overhead? Maybe - in which case they will either address those issues or people won't use the product.

    Personally, Orion is by far my favorite. JBoss works so I use it.



    > <quote>
    > > Jboss is $0 without documentation BUT _doesn't work_.
    > </quote>
    >
    > Just so that we're all clear about this, what does "_doesn't work_" mean? Does it mean that
    > a) You could not start up the server?
    > b) You could not deploy your application and get it running?
    > c) You ran into chronic, non-replicable (!) errors that baffle you and your IT team?
    > d) You ran into subtle problems with your deployment that stymied the JBoss developers?
    > e) Something else?
    >
    > IMHO, a statement like "Jboss is $0 without documentation BUT _doesn't work_." doesn't even qualify as FUD. Its just silly without some sort of relevant substantiation.
    >
    > Sandeep
  51. Sandy man, you are so deeeeeeeeeeeeeeeeeeeeeeeep into this JBoos hoopla, you seem to be unable to discern fact from fiction. I can only pity you and hope you realise that you've been taken for a ride all along. Enjoy it ;-)
  52. Actually, that's funny. Thanks! 8-) Sigh. It takes all kinds to make a world.

    I first used JBoss in 2000, even before 2.0 was announced. I've had mostly good experiences with it. As far as I am concerned, its plusses outweigh its minusses. I think its a pretty good piece of software. I honestly believe that JBoss is one of the best J2EE open source projects of all time.

    So, if you want me to apologize for liking the JBoss project, forget it. I think that if you to criticize an open source project, you had better be a participant in another project of equal stature, or you should let your name and credentials speak for themselves. Else, a) pipe down, and b) stop using the software, or start speaking up for it.

    Sandeep.
  53. you sound frustrated[ Go to top ]

    It's funny how often "It doesn't work", really means, "I can't figure it out in 3.5 seconds so it's crap."

    It's obvious you haven't worked much with JBoss. You might want to try the newb forum.

    CP

    > <quote>
    > > Jboss is $0 without documentation BUT _doesn't work_.
    > </quote>
    >
    > Just so that we're all clear about this, what does "_doesn't work_" mean? Does it mean that
    > a) You could not start up the server?
    > b) You could not deploy your application and get it running?
    > c) You ran into chronic, non-replicable (!) errors that baffle you and your IT team?
    > d) You ran into subtle problems with your deployment that stymied the JBoss developers?
    > e) Something else?
    >
    > IMHO, a statement like "Jboss is $0 without documentation BUT _doesn't work_." doesn't even qualify as FUD. Its just silly without some sort of relevant substantiation.
    >
    > Sandeep
  54. Thanks to JBoos.

    Christian, you da man! Thanks for figuring this out. I hear JBoos has a few openings for folks of your caliber.
  55. Resin is $500 and works.

    > Orion is $1500 and works.
    > Jboss is $0 without documentation BUT _doesn't work_.

    If the free docs and the O'Reilly workbook are not sufficient for you, we can change your summary to:

    JBoss is $30 and works.
  56. Nick Whippersnapper: Sorry, JBoos might be only $30, but it still doesn't work.
  57. Come on Cary, come up with your own names, at least.

    You're not too sharp, but you are good amusement for your betters.
  58. Are Cary Bloom and Rolf Tollerud the same person? Think about it:

    - They are both chock-full of bile - Cary hates "Marc Flowery" and "JBoos". Rolf hates.... well, everything Java.

    - Neither has anything meaningful to add to any discussion. Rolf throws out empty .NET stats. Cary...well, Cary makes fun of people's names and that is really about it.

    - Notice that there has been nary a Rolf sighting lately, while Cary has been trolling with a vengeance recently.

    C'mon, Hairy Doom...er...Scary Tomb...er...Very Soon...er...Jerry Seinbloom...er...Dairy Gloom...er...Berry Bufoon, fess up - are you a Rolf in another troll's clothing?

    Anxiously awaiting Cary's clever twist on *my* name.

    Ryan
  59. I think I know who Cary is ... Rolf isn't the only one who suddenly got quiet. ;-)

    Peace,

    Cameron Purdy
    Tangosol, Inc.
    Coherence: Easily share live data across a cluster!
  60. Cameron == Cameron?[ Go to top ]

    Cameron, you are an idiot as usual. I quit at TSS because the discussions had sunk to such a low level - to which you have contributed a not insignificant amount. How does it feel, to have contributed to that the discussions now are like the current one?

    You are welcome to go on with your "Slashdot" bickering. I am out.

    Regards
    Rolf Tollerud
  61. Ryan Broadband?[ Go to top ]

    Shouldn't you be working at Broderbund? Or are you related to Bo Derick? ooh ... Jboos user Broadband Rider?
  62. lets get this straight[ Go to top ]

    There is a *world* of difference between compliance and certification. JBoss Group has always claimed compliance for the JBoss server.

    The big argument is about certification - which is a totally different thing.

    Certification does not imply compliance.
    Compliance does not imply certification.

    The whole issue is a big red herring, since even compliance to the spec (and for J2EE 1.3 I thing JBoss comes as close as anything to compliance) isn't nearly enough to write an application.

    Frankly the spec is there to make is easier to write the business logic of applications. It's deliberately vague on everything else because it's designed to promote vendor value-add.

    I'm really bored of this same discussion going on over and over again.

    Chris
  63. hmm J2ee server huh?[ Go to top ]

    there is really no way to be 100% sure that JBoss is 100% percent J2EE-compliant. Not without the certification


    Of course, certification might come close to assuring 75% compliance, since the test suites have known (and definitely unknown) bugs themselves. And anyone who has ever written test suites knows that there is no way to 100% test something.

    Except, of course, that certification is only an honor system. You pay for the certification, receive the tests, and declare yourself compliant.

    JBoss is J2EE "compatible". They shouldn't be allowed to say they are J2EE "compliant" , or "certified", or even that they have a "J2EE" server. Of course, Sun could challenge the compatibility claim if they could find a case where it is false.
  64. hmm J2ee server huh?[ Go to top ]

    For anyone to be able to call a server "a J2EE server"

    > they have to:
    > 1) Implement the J2EE spec
    > 2) Implement all referenced specs
    > 3) Validate through certification that 1 and 2 have been accomplished

    Since us mere earthlings have to live in the 3D world of what we call "the reality" rather than on a distant start in a galaxy far, far away, maybe you could elaborate what parts of numbers #1 or #2 do you think JBoss is lacking? You see, for us mere earth beings the numbers #1 and #2 are the only ones that matter.

    You seem to be implying that JBoss does not, in fact, fulfill the points #1 and #2. If this is so, please provide the facts showing so. If you do not have these facts then what you are doing, in effect, is what we earth IT people call "Fear, Uncertainty and Doubt".

    Looking forward to be educated by your higher density.

    /T
  65. Tracing[ Go to top ]

    I'd recommend that people download the JBoss AOP package and do this: trace a call to an AOP object, from beginning to end, and jot down:
    *) How many objects are created
    *) How many hashtable lookups are done
    *) How many synchronized blocks are passed
    *) How many API inconsistencies you can find on the way
    Etc.

    When you're done, consider what would happen if you had an app that had lots of objects, each having lots of interceptors, each using metadata triggering and uses the response attachment facility. Interesting, eh?
  66. Tracing[ Go to top ]

    I'd recommend that people download the JBoss AOP package

    > and do this: trace a call to an AOP object, from beginning
    > to end, and jot down:
    > *) How many objects are created
    > *) How many hashtable lookups are done
    > *) How many synchronized blocks are passed
    > *) How many API inconsistencies you can find on the way
    > Etc.

    Extremely valid points. I have no doubt that the initial implementation of JBoss AOP is not as elegant as what Rickard has put together. There just aren't a whole bunch of Rickards in the world.

    But I am not too worried about the future of JBoss 4. If you remember, JBoss 3.0.0 totally sucked donkey in terms of performance. They put out a release that concentrated on features and functionality, and started doing the performance optimizations later.
    By 3.0.2, they worked the bad kinks out, and by JBoss 3.2 they had a server that rocks from the performance standpoint, and is used to deploy great product like Sitevision.

    I don't have any problems with the JBoss folks avoiding premature optimizations in a 4.0 alpha release. But their approach is proven to be successful, and the results speak for themselves. I have no doubt that a year from now Sitevision will be running on 4.x.
  67. spot on[ Go to top ]

    But I am not too worried about the future of JBoss 4. If you remember, JBoss 3.0.0 totally sucked donkey in terms of performance. They put out a release that concentrated on features and functionality, and started doing the performance optimizations later <

    Precisely. Isn't "release early, release often" supposed to be axiomatic for OSS projects? And for agile teams in general.

    And Rickard should very well know that counting "How many objects are created" in the debugger is not the correct way to test performance. The ONLY reliable approach to answering performance questions is proper S&V testing. So if Rickard has some actual numbers to share with us?

    Looks like the JBoss team were following The Rules:

    Rule 1: don't do it
    Rule 2: do it later
  68. spot on[ Go to top ]

    Precisely. Isn't "release early, release often" supposed

    > to be axiomatic for OSS projects? And for agile teams
    > in general.

    Oh dear. So on one hand the JBoss team should be releasing code early and often, giving the community the chance to participate as early as possible, getting a cooperative development started (agile style) yet at the same time Rickard is calling for a rock solid, non-changeable API, with a complete up-front design including all performance optimizations (waterfall style).

    This is most disconcerting! I sense a rift in the Matrix. And who is that man outside my window taking my picture?! Ayiee!! I will be right back!

    /T
  69. Tracing[ Go to top ]

    I'd recommend that people download the JBoss AOP package

    > > and do this: trace a call to an AOP object, from beginning
    > > to end, and jot down:
    > > *) How many objects are created
    > > *) How many hashtable lookups are done
    > > *) How many synchronized blocks are passed
    > > *) How many API inconsistencies you can find on the way
    > > Etc.
    >
    > Extremely valid points. I have no doubt that the initial implementation of
    > JBoss AOP is not as elegant as what Rickard has put together. There just
    > aren't a whole bunch of Rickards in the world.

    Well, if you listen to the JBoss folks (Marc&Bill) you'd believe that I don't really have a clue, and that JBoss AOP is the greatest thing since sliced bread. That's what they've been saying for quite some time now.

    Now that it's actually available you hear the most imaginative excuses for it's inferiority, such as "it's only a beta" (as if the API is going to change significally), and "we're only doing a minimal framework" (except there's 89 unused methods and classes in what's been released), and so on. I'm amazed, but even more amazed that people actually buy it.

    It's a typical "eat the cake and have it scenario". On the one hand it's JBoss, so it's great (because Marc says so), and on the other hand it's really not that good, but only because it's it's a beta and they didn't focus on any of the important things, such as performance.
    And again, I've been exposed to this reasoning from JBoss dev enough so I'm not amazed. I am, however, surprised that people buy it.

    > But I am not too worried about the future of JBoss 4. If you remember, JBoss
    > 3.0.0 totally sucked donkey in terms of performance. They put out a release
    > that concentrated on features and functionality, and started doing the
    > performance optimizations later. By 3.0.2, they worked the bad kinks out, and
    > by JBoss 3.2 they had a server that rocks from the performance standpoint, and
    > is used to deploy great product like Sitevision.

    Actually, the only stuff we use right now is the deployment and datasource stuff, so we're looking at ditching JBoss if possible in the next release, and only use Tomcat.

    You are also missing one major point here: JBoss 3.2 was about implementing EJB, i.e. a given spec and API. Sure, you can fix performance in later releases, and that's ok. However, with JBoss AOP they are *creating an API*,
    which needs to be well thought out and don't change much. This is much harder to fix later on, especially since it's gonna need a whole lotsa fixin' to be anywhere usable (IMHO of course).

    > I don't have any problems with the JBoss folks avoiding premature
    > optimizations in a 4.0 alpha release.

    That is so not what I'm talking about. If the architecture was optimizable that'd be one thing, but currently it is written in such a "everything is dynamic" way that there's little room for making such optimizations.

    > But their approach is proven to be
    > successful, and the results speak for themselves.

    See above, you're missing the point that this time they have to invent an API, which is much harder, and which at least DR1 has failed miserably at.

    > I have no doubt that a year
    > from now Sitevision will be running on 4.x.

    I would be very surprised if that's the case. It is much more likely that we're using either plain Tomcat or JBoss 3.2, since we're not using much anyway from JBoss. Heck, we could even use Oracle since they have a "embed for free" deal. That's an interesting idea...

    /Rickard
  70. rickards humble opinion[ Go to top ]

    <snip>
    This is much harder to fix later on, especially since it's gonna need a whole lotsa fixin' to be anywhere usable (IMHO of course).
    </snip>

    nothing about your opinions is ever humble
  71. rickards humble opinion[ Go to top ]

    I think I'm gonna blow chunks the next time I see someone verbally wax their pole to this guy.
  72. Tracing[ Go to top ]

    I have no doubt that the initial implementation of

    > JBoss AOP is not as elegant as what Rickard has put
    > together.

    But then, we will never know, will we? Rickard's AOP framework is NOT open source and we'll never be able to see its pure "elegancy". See, all Rickard has is a big mouth and a blog. No code.

    /T
  73. Tracing[ Go to top ]

    "See, all Rickard has is a big mouth and a blog. No code."

    To his defense: http://www.dreambean.com
  74. Tracing[ Go to top ]

    "See, all Rickard has is a big mouth and a blog. No code."

    >
    > To his defense: http://www.dreambean.com

    Wasn't most of that work funded by Marc Fleury while Rickard was on his payroll? In fact, apart from the PetStore, isn't all of his work paid by the very same people he likes to disparage (Marc Fleury, TheServerSide, etc)?

    Some integrity you've got there Rickard. Maybe, just maybe, you should take a good hard look in the mirror and think about your own actions. Or in other words -- time to grow up, kiddo!

    /T
  75. Performance[ Go to top ]

    Markus Kohler did some quick test to check Rickards statement JBoss AOP is slow:

    http://www.javablogs.com/ViewEntry.jspa?id=37639

    "Rickard is claiming that the first Release of the Jboss AOP implementation is less then perfect and probably inefficient.

    I did a small test and used the simple example fromt the onjava article to measure function call speed. Actually 10000000 function calls take around 2200000 milliseconds, with the Interceptor turned on, and doing some basic stuff. Without the interceptor it takes only 20 ms (YES). Argh ! That's a huge difference. And less than 500 function calls on a Duron 1200 isn't very good at all. I hope they will improve this soon ... "

    Any explanations to this result from jboss devs on this (in a calm, rational tone please)?

    Also, it seems not only Rickard Oberg is doubting jBoss's AOP performance. Tirsen from Nanning claims the same and suggests the use of AspectWerkz.
  76. Performance[ Go to top ]

    I want to make sure I'm reading this right: with the Interceptor on, it took 10 million function calls took a total of 2 million 2 hundred thousand millis (aka 220 seconds). With the Interceptor turned off, 10 million function calls took 20 milliseconds in total.

    Is that right?

        -Mike
  77. Performance (multiple choice)[ Go to top ]

    That is what's claimed. Seems quite
    ( ) unbelievable
    ( ) drastic
    ( ) flawed
    ( ) impressive
  78. Performance[ Go to top ]

    Markus, can you clarify if it is 22ms vs 20ms or 2200000ms vs 20ms? As freeroller is down, please post here. Thanks!
  79. JBoss AOP vs. POJO[ Go to top ]

    I'm not a big fan of artificial benchmarks, but it seems that JBoss AOP instrumented POJOs are about 1/5 the speed of a plain vanilla POJO (PPOJO?) :)

    http://www.jboss.org/thread.jsp?forum=151&thread=34270

    That's slower but on the order of magnitude I would be willing to sacrifice for the expressiveness of AOP.
  80. JBoss AOP vs. POJO[ Go to top ]

    Thanks for the link, good catch. But still in the link to the jboss thread you provide, the tests on 500000 calls state that AOP instrumented POJO's are about 5 times slower than plain POJO's (when writing to dev/null). That seems a lot overhead compared to AspectWerkz in offline mode (which I can also use with other app servers). I dunno about AspectJ, never used it.
  81. JBoss AOP vs. POJO[ Go to top ]

    Bob,
    do you have any timings for AspectWerkz?

    My timings were VERY unscientific, but it was interesting to see that the performance of JBoss AOP was comparable to AspectJ. That was rather unexpected.

    For me the most interesting timing is the difference in speed with the changed POJO that writes a similar amount to stdout, in that case the speed difference was about 1.5 times.
    (=> AOP-JBoss 1.4 times POJO, 1.66 POJO /dev/null,
    Aspect J: 1.5 times POJO, 1.52 POJO /dev/null)

    Rickard said something interesting: what happens when you have a chain of advices? I have the horrible feeling that he is right.. certainly JBoss AOP gives me a feeling of being "unfinished", but then again thats what you would expect of a "DR" release!
     
    PS:Has anybody tried to use AspectWerkz with JBoss?

    Erlend
  82. JBoss AOP vs. POJO[ Go to top ]

    Hi Erlend.

    > Bob,
    > do you have any timings for AspectWerkz?

    I have done some performance testing for AspectWerkz (the same ones as Rickard did and blogged about).

    I have posted them on my weblog: [http://blogs.codehaus.org/people/jboner/archives/000063.html] (scroll down to the bottom)

    - Jonas

    >
    > My timings were VERY unscientific, but it was interesting to see that the performance of JBoss AOP was comparable to AspectJ. That was rather unexpected.
    >
    > For me the most interesting timing is the difference in speed with the changed POJO that writes a similar amount to stdout, in that case the speed difference was about 1.5 times.
    > (=> AOP-JBoss 1.4 times POJO, 1.66 POJO /dev/null,
    > Aspect J: 1.5 times POJO, 1.52 POJO /dev/null)
    >
    > Rickard said something interesting: what happens when you have a chain of advices? I have the horrible feeling that he is right.. certainly JBoss AOP gives me a feeling of being "unfinished", but then again thats what you would expect of a "DR" release!
    >
    > PS:Has anybody tried to use AspectWerkz with JBoss?
    >
    > Erlend
  83. JBoss needs to do less better.[ Go to top ]

    The problem I begining to sense with JBoss, is a tendency to develop new features (AOP e.t.c.) rather than spend time improving existing ones.

    The reason I use a J2EE application server is their commodity nature, I look to JBoss as a Bea or websphere replacement if it can do what they do and do it faster, more reliably and with better or easier monitoring and maintaince.

    The weblogic console is far easier to work with than the jboss jmx-console, for instance I can browse to a JMS queue and click the refresh button and watch as I play messages through the system.

    Weblogic's clustering is relatively simply to set up and well tested, last time I tried setting up a JBoss cluster I found it far from a trival matter and certainly I didn't get the impression it was a well tested configuration.
  84. JBoss needs to do less better.[ Go to top ]

    last time I tried setting up a JBoss cluster I found

    > it far from a trival matter and certainly I didn't
    > get the impression it was a well tested configuration.

    That's strange. For me all I need to do is start JBoss in the 'all' configuration. The clustering just works. I don't have to do any setup at all!

    Maybe you did something wrong?

    /T
  85. JBoss: A modern day plague?[ Go to top ]

    Very funny and TRUE rant. It's just what many people think about the fat pig faces.
  86. It's just what many people think about the fat pig faces.


    Coming from a SwiftMQ guy. I know which JMS server I will be avoiding like a plague from now on. And I am not even going to go so low as to call you a pig.

    Have a nice day. And work on your anger issues. It will pay off later on in your life, trust me on that.

    /T

    PS. Do you think that reliable messaging requires sync on a physical storage yet?
  87. Coming from a SwiftMQ guy. I know which JMS server I will be avoiding like a plague from now on.

    Mattson, you are a freeware sucker [and a MF brown-nose too], aren't you? Look, as a freeware sucker you don't have to avoid our product. You don't have money, so you can't buy it anyway. We don't care about you. Understand? ;-)

    PS. Do you think that reliable messaging requires sync on a physical storage yet?

    It depends. But why do you care about disk syncs? You are a freeware sucker and use JBossMQ which doesn't have an option at all to sync with the disk [except you hack it into the source, of course]. So why do you care?

    -- Andreas
  88. Mattson, you are a freeware sucker [and a MF brown-nose too], aren't you? Look, as a freeware sucker you don't have to avoid our product. You don't have money, so you can't buy it anyway.


    I personally have very little, or should I say modest, middleware requirements. However, my client's do not and they come to me for advice,

    We don't care about you. Understand? ;-)

    I understand perfectly. Thank you for clearing that up.

    /T
  89. J2EE cert.[ Go to top ]

    Hello

    Ok, after reading I have something to ask. JBoss is not J2EE certified. I get that, but why not? I have heard different things from different people. Some people say it is because JBoss won't pay the fee. The JBoss people say that Sun won't let them take the certification test. Which is it?

    Mike
  90. J2EE cert[ Go to top ]

    You might want to add 'incomplete' to the list of potential reasons why JBoss has not been certified. I'm not the expert on this, but the word I got from people who should know is that IIOP support is just one of the outlying specs that JBoss does not support. Interoperability and complete support of *all* aspect of the J2EE API familiy are part of the purpose of certification.

    The assertion I have heard is that they couldn't pass the certification test, even if Sun gave it to them for free.

    Brad
  91. J2EE cert[ Go to top ]

    \Schneider\
     I'm not the expert on this, but the word I got from people who should know is that IIOP support is just one of the outlying specs that JBoss does not support. Interoperability and complete support of *all* aspect of the J2EE API familiy are part of the purpose of certification.
    \Schneider\

    From what I've seen this is correct.

    Personally, I don't understand all the fuss about certification. Since JBoss is open source, you can quite easily look at the source and see what it does (and doesn't) do. Unfortunately for my needs, JBoss' JMS implementation doesn't make a very credible effort, and definitely isn't usable for production use from my inspection of the code. When you don't even attempt to make your XA support recoverable at all, well there's no much hope for this iteration of their JMS support.

    Anyway, unlike Weblogic Websphere et al, you don't need to guess. You can see up close and personal what JBoss does and doesn't support. From my own peeking here and anon, I'd say there's a glimmer of truth in what Rickard says - the public statements about compatability, or certification, or whatever - don't really match the reality. In some areas it's pretty good, in other areas an attempt hasn't even been made. It certainly was a surprise to me to look at the XA JMS code and see the recover method implemented as:

        return (new Xid[0]);

    Incidentally, even if you're not up to the task of looking at the code, the forums at jboss.org have most of the information you need on what JBoss, the reality, is good at and what it's not.

    As for certification - from what I've heard there are enough holes in the certification that it doesn't interest me very much. It's more of a marketing thing then anything else. What interests me is how the code matches the published specifications.

         -Mike
  92. J2EE cert[ Go to top ]

    Oh great. The old "friend of a friend" or "some people I know" routines.

    My, aren't we a technically sophisticated crowd here ? JBoss's support for IIOP has been relegated to the pages of Urban Legend, on TSS no less.

    Yes, I can see it now, right under the entry for travellers having their kidneys harvested by organ smuggling criminal gangs.

    "Be careful. My sister-in-law works with a lady who said her son's neighbor who lives in Houston had a partner whose cousin had a problem with the IIOP service in JBoss. Tell everyone you know. And oh...... post it on TSS. Their readers don't know any better."

    Too bad.....
  93. J2EE cert[ Go to top ]

    \nick w\
    Oh great. The old "friend of a friend" or "some people I know" routines.

    My, aren't we a technically sophisticated crowd here ? JBoss's support for IIOP has been relegated to the pages of Urban Legend, on TSS no less.
    \nick w\

    From:
       http://jboss.org/index.html?module=html&op=userdisplay&id=developers/projects/jboss/IIOP

    the main JBoss IIOP Page:

    "Status

     JBoss/IIOP is currently in alpha state."

    There are also only 13 topics/24 messages in the JBoss IIOP forum.

    I suppose your response is that The JBoss group is now also engaging in FUD about its IIOP support :-) They obviously must have a hidden agenda in declaring their IIOP support alpha.

        -Mike
  94. J2EE cert ... NOT![ Go to top ]

    Spot on Mike! Looks like Nick Wind is now gasping for air trying to prove a few unprovable facts!! Nick Wind/Whippersnapper/warmonger/whatever, it is obvious you don't have a clue re: JBoss and yet, you profess to be a JBoss user and advocate. Continue drinking the kool-aid my friend and may the force be with you.
  95. Useless Posting[ Go to top ]

    There are better things in the world.
  96. Blaa blaa blaa[ Go to top ]

    I read this guys comments on various topics, and if this is the crap TSS is going to post, then I'll quit reading. I can get stuff like this from IRC flame wars.

    This guy obviously has no idea what he's talking about from a technical or marketing perspective. So what's the point?
  97. re: Blaa blaa blaa[ Go to top ]

    Well, it could be that TSS isn't quite fond of the JBoss project or Marc Fluery. And wouldn't mind parading insulting blog entries as genuine technical positions that they deem appropriate for their readers. And just so that it looks like they aren't posting nonsense, they include as an aside comments from Rickard so it looks like the original entry is credible in itself.

    However, I think that they've succeeded. We've all taken the bait. I am sure we'll soon receive a note from the TSS folks saying that this was merely an IQ test and that we all failed miserably.

    Sandeep.
  98. My Rant[ Go to top ]

    I am sick and tired of the personal attacks on JBoss, Marc Fleury, and JBossGroup. I'm sick and tired of the mud that's thrown at us on a constant basis against the nominal success we've achieved so far. I'm sick of the freerollers who chip at our hard work or try to make a name for themselves by bashing Marc or JBoss or JBossGroup in their ego-centrical BLogs.
    You don't see any of us writing blogs attacking these people. Hell, we don't even have blogs! No, you only see us come out when somebody is attacking the project we love. We are aggressive because we care passionately about what we're doing. Both in code and in business.

    Yes we are aggressive, yes we may hurt people's feelings, yes some people have fragile egos that can't take our passionate rebuttals. It's all for 2 common goals. To promote the JBoss OSS project and to build a business on open-source.
    All of us at JBG have taken huge risks staking our livelyhood on open-source. We are interested in building something bigger than ourselves. We are interested in building a real company.

    I've been around the JBoss world for more than 2 years. Seen it grow from Marc giving sporatic training courses while living in his in-laws garage, to a growing company of 25-30 individuals.
    Yes everyone at JBG is guilty of saying stupid things that we'd like to take back. Yes mistakes were made. But I think we're learning from them. We believe in what we're doing and will defend ourselves.

    Bill
  99. My Rant, back atcha Bill[ Go to top ]

    You guys don't have blogs? No wonder everybody hates you. :) You don't ever SAY anything!
  100. My Rant[ Go to top ]

    I am sick and tired of the personal attacks on JBoss, Marc Fleury, and

    > JBossGroup. I'm sick and tired of the mud that's thrown at us on a constant
    > basis against the nominal success we've achieved so far.

    Have you ever stopped to ponder why there are "personal attacks" such as the above? Do you think there's a reason for it, or are they just unfounded, out of the blue, sporadic rants?

    > You don't see any of us writing blogs attacking these people. Hell, we don't
    > even have blogs!

    Come on Bill, don't be shy. Sure you do. Here, let me remind you:
    http://www.jboss.org/index.html?module=journal&op=search&bywhat=uid&forwhat=4624

    > No, you only see us come out when somebody is attacking the project we love.

    All I see is ego protecting. I have seen very little in terms of protecting the project, which I used to love as well.

    > We are aggressive because we care passionately about what we're doing.

    I think you're aggressive because you don't understand why people have a problem with JBoss. If you did you could respond differently.

    > All of us at JBG have taken huge risks staking our livelyhood on open-source.

    Please Bill, don't go there... it's a pretty lame "I'm an OpenSource martyr" ego-trip thing.

    > We are interested in building something bigger than ourselves. We are
    > interested in building a real company.

    You are interested in building something that looks and feels like a real company, I'll give ya that. Judging from the comments from people who left (and from some who are still in JBG) there's a difference between what you say and what you do.

    > I've been around the JBoss world for more than 2 years. Seen it grow from
    > Marc giving sporatic training courses while living in his in-laws garage, to a
    > growing company of 25-30 individuals. Yes everyone at JBG is guilty of
    > saying stupid things that we'd like to take back. Yes mistakes were made.
    > But I think we're learning from them. We believe in what we're doing and will
    > defend ourselves.

    But that's just it. You've made mistakes for a long time, the mistakes have been pointed out (by myself and many many others), and you keep doing them over and over and over again. You're not learning, because you're not "gettin' it". If you actually understood what the mistakes were and how to avoid them I very much doubt that you'd be in the situation you are right now.

    *sigh*

    /Rickard
  101. Re: All the rants[ Go to top ]

    <Rickard>I think you're aggressive because you don't understand why people have a problem with JBoss. If you did you could respond differently. </Rickard>

    <sincere>
    Rickard, exactly what *are* these problems? I'd like to know, and I am sure there are a few others who don't get it either. From what I've seen so far, I think that a lot of people have problems with Marc and his attitude. And perhaps his endeavor to make a consulting practice around JBoss. But why is JBoss hated so much? What would *you* list as the top 5 reasons?
    </sincere>

    Sandeep
  102. Re: All the rants[ Go to top ]

    And perhaps his endeavor to make a consulting practice around JBoss. But why is JBoss hated so much?


    Not that I think there's anything wrong with that sort of business approach in general.
  103. Re: All the rants[ Go to top ]

    I too would like to know the problems. All I know from my limited experience is that my apps work on Jboss and are easy to port to other app servers. Admittedly we haven't used the jms implementation much and i've heard they are reloading it to make it a world class jms implementation.
    Ive heard the IIOP argument but i don't use it and don't care if works or not. I guess if you wanted a good IIOP implementation you couuld go with Iona, etc.

    Perhaps you can enlighten me.
  104. Re: All the rants[ Go to top ]

    <sincere>

    > Rickard, exactly what *are* these problems? I'd like to know, and I am sure there are a few others who don't get it either. From what I've seen so far, I think that a lot of people have problems with Marc and his attitude. And perhaps his endeavor to make a consulting practice around JBoss. But why is JBoss hated so much? What would *you* list as the top 5 reasons?
    > </sincere>

    I'm glad you asked, so I'll give you my top five:
    1) Whenever a problem is raised, there is a knee-jerk reaction: "but it's free", ergo there is no problem as it can potentially be fixed by the individual who raised the issue.
    This helps to hide problems, and continue to live in the fantasy that JBoss is the latest and greatest app server. It's brilliant, in it's circular logic, but is also sure to piss people (like me) off, who don't live in the JBoss fantasy world.
    2) The claims that "JBoss did X first" (e.g. dynamically generated client stubs) whereas I have yet to find any single thing that JBoss did first. And many of those claims are for things that I added to JBoss in 2.0.
    I was surprised by Marc's claims back then, but shut up for "the greater good". I shouldn't have.
    But, the point here is that I think Marc actually believes what he says. I don't think Marc ever lies, properly. It's his perception of reality which is distorted, which causes him to say things like this.
    3) The claim that JBoss is "management and administration friendly" just because it runs on JMX, even though there's no decent console (even now) and the configuration file jungle is hideous (as outlined in Hani's post).
    4) The disconnect between JBoss the idea, and JBoss the reality. I think this is one of the more subtle points. Marc has always tried to expose his view of JBoss the idea, but instead of referring to them as ideas they are described as reality (e.g. "JBoss is an J2EE server" as opposed to "we want JBoss to implement J2EE").
    (note: it's actually quite funny that Marc refers to JBoss as liberating people from the corporate Matrix, as what he's really doing is actually the opposite: placing people in a state of mind where JBoss can walk on walls and perform marvellous feats, and which only JBoss, as The One, can do.
    Even though it's not real, those who are entranced by Marc's words will believe it is actually the case). If I understand the CDN guys correctly, this is a major reason why they forked off.
    5) The attitude that people should be grateful that Marc&Bill and gang have sacrificed themselves on the OpenSource stake in order to bring free J2EE to the masses. This ties in quite heavily to Marc's general ego-problem, which is an attitude towards "mortals" that is severely elitistic.
    This also shows in the lack of documentation, meaning, if you can't read code to figure things out, you don't really count (or you do count, but am willing to pay money for consultancy time and support to make up for your inability to read code).

    To sum up, whether or not you like JBoss or not depends largely on if you are able to discern between JBoss the idea and JBoss the reality.
    For those who can't see the disconnect any complaints are going to look like unfounded general rants and attacks on The Project which can be disregarded as FUD. Reminds me a lot of certain socio-political philosophies, but Hani said I'd lose the argument if I go there, so I won't.

    I'm not sure if I managed to get all the major problems into that list, but I think it's a good start. And the above only marginally touches on the technical problems, which is a whole other discussion in itself.

    Hope this helps.

    /Rickard
  105. Mississippi cat fight[ Go to top ]

    Man, not one serious technical argument in the list. This thread was started based on Rickard's reputation and valuable opinion on JBoss. It now appears to have been nothing more than a little McCarthy-style black book of technical issues to front a personal grudge against MF.
    I give the win to JBoss on this one since the one person who would know the faults, whose view is respected, has made some great contributions to OSS, and who seems to remain objective and technical (most of the time), hasn't made a serious argument.

    JBoss is good stuff and the community needs it. Most users don't care about MF. They do care about software that works. Why don't you help CDN to fulfill your vengence. That way everybody wins.
  106. Mississippi cat fight[ Go to top ]

    Man, not one serious technical argument in the list.


    Sandeep asked why people hate JBoss, and that's what I replied to, and I doubt that there are technical arguments for it. But I might be wrong. Do you see any technical arguments why people would "hate" JBoss(/Group)? Please enlighten me.

    > This thread was started based on Rickard's reputation and valuable opinion on
    > JBoss.

    The referenced blog entry from me refers only to technical arguments, which still hold. If you think any of the points I made were wrong, either comment on them here or in my blog.

    > It now appears to have been nothing more than a little McCarthy-style
    > black book of technical issues to front a personal grudge against MF. I give
    > the win to JBoss on this one since the one person who would know the faults,
    > whose view is respected, has made some great contributions to OSS, and who
    > seems to remain objective and technical (most of the time), hasn't made a
    > serious argument.

    This is exactly the kind of response I was referring to. Any of the issues I listed are disregarded as FUD and pure ranting, without any objective reality, which helps maintain the illusion of JBoss.

    > JBoss is good stuff and the community needs it.

    This is nr 5 on my list.

    > Most users don't care about MF.

    This is nr 5 on my list (the users who can read code don't care, because they're "in").

    > They do care about software that works.

    This is nr 4 on my list. The idea is that it works, and works well. Read Hani's original post about why this isn't reality.

    > Why don't you help CDN to fulfill your vengence. That way everybody wins.

    You are probably not aware of it, but you are actually helping me here, since your response is a great example of what I was referring to in my previous post.

    /Rickard
  107. Man, aren't you extrapolating a personal problem towards MF, and reflecting it into the whole JBoss stuff? I´m not a shrinker, but by reading all this thread, sounds like the whole problem is MF himself, not the product, since I've read a lot of posts saying Jboss works fine for them.
    So the product should be ok. It's not the holy grail, but it works. The problem lies in the people (as always)... The problem is MF, not Jboss software, nor JBoss group as a whole, so it seems. Of all 5 reasons you gave, 3 cited his name directly.

    I wouldn't want to be in MF's shoes, no way.

    Take it easy.
  108. Mississippi cat fight[ Go to top ]

    Rickard, I hear you. It sounds like you are against the sales hype and are frustrated that JBoss is not in line with your philosophy. Most who are not as close to it would like to see bugs and personality issues worked out. If JBoss does suck, time will tell. Right now they appear to be OK. As Tomcat gets stronger it seems that the tide is turning against the EJB container anyway.

    Pt.1 I would think JBoss uses bug tracking? I would also think JBG's clients would get first say in what needs to be fixed.

    Pt.2 This is just a harmless sales gimmick. Apple ripped off Zerox Parc and Microsoft ripped off Apple. Given JBG's marketing budget compared to BEA and IBM - not bad. And how many times has a vendor said it does X only to find it doesn't. I'm thinking of Websphere. I sounds like JBG doesn't do that.

    Pt.3 JMX is friendly. Not having a WebLogic-esque console is a soar point for sure but does it stop people from using the core container. Maybe it's not a priority.

    Pt.4 Hmm. Can't help you on this one. I think business people buy this kind of crap but technical people kick the tires and if it works they keep it. We would like more more OSS projects to be run by Linus Torvalds (no BS, no ego) but alas Finland has but so many people to spare.

    Pt.5 I'm with you on the documentation. If they gave it away for 4.0, I bet they would see a surge in protoyping with a percentage going on to be operational. It's a short term revenue loss with long term payback in consulting. Yes, the price is cheap for documentation, but it may be too much for the IT geek in the corner to risk $60 of his beer money in an enterprise bake-off with IBM and BEA.

    On the other point about being grateful. En lieu of money, MF better be grateful to all JBoss developers, even those who have left to form other companies. I am grateful the the cool stuff JBoss and you have produced, it makes my life easier - daily. So MF and others may use it as a defense. It doesn't appear to be working based on these posts.

    cheers
  109. Never was, never will be! Get over it and stop whining!
  110. Mississippi cat fight[ Go to top ]

    You are probably not aware of it, but you are actually helping me here, since

    > your response is a great example of what I was referring to in my previous post.

    And I sincerely want to thank everyone else who posted their rants after this post. It's exactly that kind of stuff I was referring to. Ignoring the issues. Attacking the messenger. Making wild unsubstantiated accusations. Etc. etc. Have fun in the JBoss-Matrix ;-)

    /Rickard
  111. Mississippi cat fight[ Go to top ]

    And I sincerely want to thank everyone else who posted their

    > rants after this post. It's exactly that kind of stuff I was
    > referring to. Ignoring the issues. Attacking the messenger.
    > Making wild unsubstantiated accusations. Etc. etc.

    So I guess this means we won't be hearing your technical arguments why JBoss is not a J2EE application server? Only your worthless whining due to your personal issues with Marc Fleury.

    Figures. Big mouth and a blog. That's all you have.

    /T
  112. Mississippi cat fight[ Go to top ]

    That's simply delusional
  113. I have a doubt![ Go to top ]

    Have you read fates rant at all? _That_ is how this all started. I suggest you read it and come back here and tell us what is not true about it.
    Kinds.
  114. Rickard Oberg--JBoss Stalker[ Go to top ]

    Rickard,
    Frankly I and I'm sure other people are sick of hearing your rants about JBoss and Marc Fleury. Granted, Marc Fleury might be an *******, but at least he's not a self-righteous hypocrite who can't get over the fact that a certain project didn't croak and die just because he left it over two years ago.
    From what I remember you were every bit as rude and obnoxious as any of the current JBoss developers in the days when you were on the project.
    Unlike them, you're into image management so you'd do things like cut and paste so only the parts of threads where you were right would be apparent.
    I guess someone like yourself wouldn't know what it is to risk anything for open source, since apparently you were paid a salary while you contributed to JBoss and XDoclet.

    Marc Fleury lives in a fantasy land? Maybe. However a brief visit to your blog leaves the impression of a manically jealous paranoid schizo. Remember the reason that Marc and Bill and not you were at JavaOne getting some attention and you weren't is that the CIA is after you man.
  115. Whiner[ Go to top ]

    All this chest-thumping and I have yet to see a substantive argument/reason *not* to use JBoss. A few vague posts stating it's broken, doesn't work, or doesn't implement the full spec - but when pushed, no one has really substantiated or demonstrated anything, other than maybe to point out a bug and a shortcoming or two. BFD.
    Just one long rant. Look at any product in detail (J2EE or otherwise) and you're bound to come up with your own bitches about it. Ever use the original WL console? Let's just say it's come a long way.

    Hani:
    People like you will never be convinced, or swayed by pragmatism or even reality. You keep whining, we'll keep implementing. Seemingly against all odds (if one believes your rants), our applications (several of them) are humming along nicely on our JBoss platform, with more planned. I am pretty sure we're not alone. Draw your own conclusions.

    Without it, we would likely not have been able to implement J2EE here. What was departmental has become enterprise and we now have purchased training, docs, and support from JBG. All for (quite literally) a fraction of cost WS/WL.

    Bottom line: JBoss works for us, and works well. To generalize that to say that it would work for everyone, would be as asinine and egomaniacal as many of the other claims and assumptions made elsewhere in this thread.

    Mike
  116. Re: All the rants[ Go to top ]

    1) Whenever someone pays, they get an answer. And lots of people get answers for free.

    2) If JBoss did something first, and you implemented it, does the fact that you no longer contribute to JBoss invalidate the claim? However, there are many things that JBoss has (and continues to do) first, but in the spirit of your own rants, I will not list them.

    3) I'm stupid, granted, but I can't make sense of WebLogic or WebSphere. Maybe if I paid for training, but I didn't have to to get JBoss working. Personally, I think Orion takes the prize in ease of use, but I use JBoss because it is free, and I can get free support online too.

    4) I don't see anything wrong with Marc talking about what he wants JBoss to become. In the Matrix, Neo could "see" the code, which enabled him to manipulate it (walk on walls.) The CDN guys left because of the money. They got a big fat venture capital investment from folks who saw JBoss as a growth market that could support more than one consultancy.

    5) Who cares about Marc & Bill? I bet you still have more money than them; and so what if next year they're billionaires and you're not.
  117. Re: All the rants[ Go to top ]

    1) Whenever someone pays, they get an answer. And lots of people get answers for free.

    >
        Will you work for free mate?
    > 2) If JBoss did something first, and you implemented it, does the fact that you no longer contribute to JBoss invalidate the claim? However, there are many things that JBoss has (and continues to do) first, but in the spirit of your own rants, I will not list them.
    >
       > 3) I'm stupid, granted, but I can't make sense of WebLogic or WebSphere. Maybe if I paid for training, but I didn't have to to get JBoss working. Personally, I think Orion takes the prize in ease of use, but I use JBoss because it is free, and I can get free support online too.
    >
         I stick to my guns. Getting something free does not make it is the better one. If you can't makes sense of WebSphere or WebLogic, do you honestly think ,its IBM or BEA's fault?

    > 4) I don't see anything wrong with Marc talking about what he wants JBoss to become. In the Matrix, Neo could "see" the code, which enabled him to manipulate it (walk on walls.) The CDN guys left because of the money. They got a big fat venture capital investment from folks who saw JBoss as a growth market that could support more than one consultancy.
    >
       I dont know this stuff. But personally, i would never get bought into JBoss.
    > 5) Who cares about Marc & Bill? I bet you still have more money than them; and so what if next year they're billionaires and you're not.

                Again, it does not matter. I dont care about if Marc and Bill becomes billionaires. I have nothing to do with them. Its all about JBoss the so called J2EE server. Nothing against, Marc and Bill. All the best to them.
  118. My Rant[ Go to top ]

    No offense,
       but maybe you guys can meet somewhere and take it "outside" and get it over with. I'm actually quite surprised that this whole thread was even posted, or that I'm even wasting my time writing this :-), but someone needs to take a timeout.
    Maybe you guys were lovers in past life with unresolved conflicts, maybe someone should take the high road, maybe Dion Almaer should read the "No obvious flame bait" bullet on the site he's responsible for, maybe both sides deserve/need an apology. Who really cares? Please go find your nearest pub, drink a brewskie, and get over it :-)
  119. This is only a test[ Go to top ]

    This thread was actually an intelligence test. You have all failed miserably. You cretins.
  120. Well said[ Go to top ]

    <Byung>
    No offense,
       but maybe you guys can meet somewhere and take it "outside" and get it over with. I'm actually quite surprised that this whole thread was even posted, or that I'm even wasting my time writing this :-), but someone needs to take a timeout.
    Maybe you guys were lovers in past life with unresolved conflicts, maybe someone should take the high road, maybe
    Dion Almaer should read the "No obvious flame bait" bullet on the site he's responsible for, maybe both sides deserve/need an apology. Who really cares? Please go find your nearest pub, drink a brewskie, and get over it :-)
    </Byung>

    The above post sums up my feelings as well. Rickard your posts appearing more and more like those of a disgruntled employee. I really don't care if your posts are accurate or not, your personal involvement and baggage undermine your arguements.
    One one point your are pointing to some form of empirical data on the other you muddle your posts with biased personal mini rants. Move on man.
  121. Rickard's Rant[ Go to top ]

    Admit it Rickard - this is a personal vendetta of yours against the JBoss team. We all make mistakes sometimes, what is so unforgettable that you have to 'bash' them professionally?

    Yes MF can be rude and immature and the group has lot to learn businesswise. But the fact is that the JBoss is a decent J2EE server. You heard me right, J2EE. I started to learn J2EE stuff some 2 1/2 years ago by reading documentation from SUN and by using their J2EE server.
    You can imagine the relief when after six month struggle I switched to JBoss. Still reading the SUN docs and now also armed with Mastering EJB2 from Ed Roman I continued to study and experiment and - after a while also do some serious development work on JBoss.

    The only JBoss specific code I have is the jboss.xml (extension to ejb-jar.xml) and couple of methods using the MBean interfaces in session bean (here's the 'offending' code)

    [code]
    public void flushXXXCache() throws Exception
    {
      initializeRMIAdaptor();
      ObjectName name = new ObjectName(XXX_M_BEAN);
      jboss.invoke(name, "flushCache", null, null);
    }

    public int getXXXCacheSize() throws Exception
    {
      initializeRMIAdaptor();
      ObjectName name = new ObjectName(XXX_M_BEAN);
      Integer result = (Integer) jboss.invoke(name, "getCacheSize", null, null);
      return result.intValue();
    }

    private void initializeRMIAdaptor() throws NamingException
    {
      if(null == jboss)
      {
        jboss = (RMIAdaptor)EJBHomeFactory.getContext().lookup("jmx/rmi/RMIAdaptor");
      }
    }
    [/code]

    The project has been (and is still) a long and interesting one and there have been problems. But the problems are/were not related to JBoss. The way you design and architect your app is much much more important than the server you use.

    If someday the JBoss doesn't meet our requirements anymore, we'll switch. (Resin-EE, Orion, etc..)

    Here's my list of problems/shortcomings on JBoss
    - Authentication mechanism is almost impossible to customize if you need extra info besides the regular username & password. (I know that this is not the case in e.g. Weblogic)
    - Classloader mechanism doesn't 'allow' you to deploy multiple versions of the same component. This means that you cannot have version X.x and X.y running simultaneosly within single JBoss instance.
    - The binary distro of 3.0.7 was accidentally built on JDK 1.3 which caused the JDBC 3 level operations to be unavailable. Fixed it by building it myself. (Double click on the jboss-3.0.7-src\build\build.bat and collect the output from jboss-3.0.7-src\build\output - hard? :-))
    - Even the paid documentation doesn't describe the changes one needs to apply to deployment descriptors when there has been a change in them (I usually compare the new examples against my previous ones and extract the info from there - tooks about one hour)
  122. Calling your bluff[ Go to top ]

    Okay, Rickard, lets hear it.

    What is the cause for the personal attacks? What are the mistakes that cause so many people to apparently hate JBoss and anyone who uses it? We all know why you left, it was for the money. And now Marc & Co. have some too. Maybe there were personal reasons as well. I don't read your web log regularly so I may have missed where you enumerated them.

    I haven't seen the JBoss 4 developer release code, but I'll take your word that there are probably too many callbacks in the AOP monitoring. But is that any reason for personal attacks, or to make unfounded insinuations about JBoss's J2EE compatibility or even its usability?

    So go ahead, list some of the mistakes made by JBoss, and please give us something more than link to a personal IRC chat Marc Fluery had years ago.
  123. My Rant[ Go to top ]

    Why is TSS spreading this stuff?

    I agree with Bill Burke. Things are too personal and it detracts from the jboss product and the purpose of this site. just let them (Jboss and Jboss group) be. And Dion, come on dude, you knew what was going to happen with this post. you're fanning the flames.
    i read hani's original blog, and its just plain rude. BEA's CEO is ugly too - do you need to advertise it on the serverside front page? What does Hani look like? i'd guess that no one cares. you are making this site look bad. i don't come here to read personal flames - leave them on the personal blogs where they belong.

    at the end of the day we are passionate about technology, but we still are people. People may be agressive, and people make claims we may diagree with, but you'll get more mileage if you disagree respectfully and post with that in mind.

    have a great week!! :)
  124. My Rant[ Go to top ]

    Bill Burkha, if you stopped for a moment and thought for yourself, you wouldn't be in this predicament. The problem is that folks like you are too deeply brainwashed and are unable to see the truth.
    What is this I hear about "nominal success" in your posting? Listening to MF it sure doesn't seem that way with all your red and blue pills and 7.2 billion downloads. I thought you were going to conquer the world, one download at a time? No? Oh ... wait, you folks are NOT even J2EE-certified, for chrissakes.

    Jboss folks have probably the biggest egos I've ever seen. Remember, even if it is free, free crap is still crap!
  125. My Rant[ Go to top ]

    Again I ask you which appserver do you use Mrs. Bloom. Apparently you hate jboss and ibm. Who do you bless with your approval? Or is it all negative name calling premature rants.
  126. Matt Finish Smith?[ Go to top ]

    Mr/Mrs MattFinish:

    I never said I didn't like WebSphere. It is just that I can't stand the phony JBoss trolls like yourself.
  127. Matt Finish Smith?[ Go to top ]

    Mr/Mrs MattFinish:

    >
    > I never said I didn't like WebSphere. It is just that I can't stand the phony JBoss trolls like yourself.

    more negative comments from cary
  128. Matt Finish Smith[ Go to top ]

    Hey MattFinish ... looks your are truly matte finish! I was referring to the shady accounting deals at IBM, that wasn't a comment about WebSphere!

    Isn't this a classic JBoss strategy: obfuscate the issues and portray yourself as the knight in shining armor!!! Great strategy, no wonder it is not working despite your 7.2 bn downloads.
  129. re:My Rant[ Go to top ]

    cary, have you used jboss? if so what didn't you like about it
  130. ok, it isnt. but someone tell me what does NOT work in jboss? i uses much of the J2EE spec in my apps, and everything works. and its easy to use. what else do i want?

    For those people ranting, please just tell me *what* does not work. and post example code showing something that doesnt work. (relevant stuff..) i havent encountered anything. when i deploy in Jboss, and i made a mistake in my ejb's, it sometimes even tells me what portion of the j2ee spec i violated. so i'm happy.

    Rickard, didn't you code the Jboss core a while back? i wonder, do you still use jboss? if not, whats your appserver (if you *had* to choose) of choice and why? would you recommend jboss to the typical app developer (you're atypical :)) out there?
  131. App clients in .ears and the classloader, among others.

    As far as the "IS certified..." are you retarded? Even The JBoss Group doesn't make that claim, nor should they. For that matter, many J2EE licensees don't make the claim yet, since they're LICENSED but haven't passed the CTK yet. JBoss, of course, hasn't got the CTK to pass, so certification is an impossibility so far.

    Give them time and, well, money. Eventually I'm sure they'll have it if they choose to go for it. They don't have it yet, period.
  132. Jboss is NOT certified[ Go to top ]

    Wonder why? The "greatest coolest" app server isn't even J2EE certified!!! Whoa!
    It is amazing how the JBoss trolls who are so indoctrinated by MF seem to think that it is certified, when it isn't!
  133. Jboss is NOT certified[ Go to top ]

    Of course it's not 'sun certified'. at least read the post. we all do our own mental certifications with everything we do. that's I was alluding to with the outrageous title.
    Rickard says Jboss 'maybe' complies with all the j2EE spec. he's right - maybe is all anyone can say. But based on my j2ee work, it complies. so that 'maybe' is good enough for me. it may not be good enough for you - that's cool too because it depends on your needs and experiences.

    if it is at all possible for you to disagree with any real substance, please tell me what doesn't work in jboss. all i see on this whole post is 'client app in an .ear' - cool, thats 1 thing - i'll look into that. tell me more - i sincerly want to know.

    every opinion is the end result of some sort of indoctination. MF can say whatever his heart desires - my indoctrination happens when i run ant and deploy.
  134. Posted By: Cary Bloom on June 17, 2003 @ 04:50 PM

    >Wonder why? The "greatest coolest" app server isn't even J2EE certified!!! >Whoa!
    >It is amazing how the JBoss trolls who are so indoctrinated by MF seem to >think that it is certified, when it isn't!


    Who gives a flying **** wether JBoss is J2EE certified or not?

    JBoss is simple and elegant J2EE app server that works, PERIOD.

    I and a million JBoss users would care less if JBoss in't J2EE certified yet. Sun has to be blamed 99% for not allowing JBoss group access to the J2EE cert test suite (which they wouldn't unless they get their freaking $$$$). JBoss group will not pay a dime for the cert test suite, so Sun can take their freakin test suite and shove it up their own ass.

    Afterall other big $$$$ J2EE app server vendors who claim to be J2EE certified
    are not 100% certified ..so whats the point. The certification process has lot of holes in it and Sun folks can let you slip through those holes if you just shove wads of cash in their pockets.

    So who is Sun trying to fool?
  135. First thing,
         It would be better, if we can mind of our language in this forum.
    This is a technical forum (exception few threads), not something else.

    <quote>
     I and a million JBoss users would care less if JBoss in't J2EE certified yet.
    </quote>
         This is biggest, and funniest lie i have ever heard of. Dude, downloads does not mean that many number of users. I downloaded some version of JBOSS, could not digest it, so i thrashed it. Count me out from your numbers.

    <quote>
       So who is Sun trying to fool?
    </quote>

    None. Sun is not trying to fool anybody. Sun has a business todo, thats what they are doing. Besides that, there is no meaning in trying fool a FOOL.

    my 2 pence.
  136. JBoss is not J2EE certified[ Go to top ]

    Srinivas:

    Why the pathetic language? Are you upset because your H1-B visa has expired and you are scared the Feds will soon find out that you've been here for the past year without a job and thus in violation of your H1B visa rules? You can't even spell straight for chrissakes (leave alone the language).

    Getting back to J2EE certification, either an app server is J2EE certified or it isn't. You can't be half-pregnant, can you? For a lot of companies, J2EE certification means a lot. It suggests that there is some minimum guarantee that an application running on J2EE-certified appserver #1 will run on a similarly certified appserver #2 with minimal changes.

    JBoss has been playing the underdog card for far too long now to be taken seriously. It may have worked in the beginning, but we can all see through MF's gameplan and nobody is fooled. There is nothing preventing Jboss from running the certification tests (Sun has even waived the fees). If JBoss is this manna from heaven (as we're led to believe) and has all this hotshot developers working on it (again, as we're led to believe), what's preventing them from being J2EE compliant for the past *TWO* years? It doesn't take a genius to figure the answer: because JBoss is fundamentally broken w.r.t J2EE compliance and it would take a major rewrite to fix it.

    So, 'nuff said. Let JBoss fix the J2EE compliancy and we'll talk. Until then, it is simply a case of "big hat, no cattle".
  137. Cary;

    You are so low class calling out spelling.

    And besides, there is no such word as "chrissakes".

    At any rate, while I am sure you are eminently qualified to comment on J2EE certification, it is obvious that you are completely in the dark on JBoss's ability to get it.

    Kudos to you and whatever certified product you're using that you must love so much. We are all waiting with bated breath to know what it is.

    All bluster and no balls.
  138. Nick, since you profess to be a JBoss expert, can you provide the definitive answer to "is JBoss J2EE-certified"? We all know what the answer is, but you seem to be delusional in thinking otherwise. Sorry to rain on your (delusional) parade.

    You have NO business to know what app server I use. It is NOT JBoss, as is evident. Funny you should ask, hiding behind a false name! Nick Whoa?
  139. Cary;

    I realize you do not keep up with the news, but here is a news flash: JBoss is not J2EE certified. Apparently you are the last person to know. The funny part is how you of all people keep the debate alive.

    So here is some more news:

    News Item A. Many people do not care if it is certified or not. They like the level compatibility it has.

    Since the difference between "certified" and "compatible" clearly eludes you, I thought it might be a nice opportunity to outline the difference:

    compatible: Capable of existing or performing in harmonious, agreeable, or congenial combination with another or others.

    certified: to be issued a license or certificate

    News Item B. Sun does not really consider JBoss to be incompatible with J2EE (incompatible means the opposite of compatible). In fact, they have made it clear that it could be certified quite easily. (See http://www.infoworld.com/article/03/06/12/HNjbossdevs_1.html). The argument is whether or not the certification should be paid for.

    Seriously though, no one cares what app server you use, you big bag of wind.
  140. JBoss is not J2EE certified[ Go to top ]

    There is nothing preventing Jboss from running the certification tests (Sun has even waived the fees).


    Do you work for Sun? Have you seen the contract Sun gave to us?

    Bill
  141. JBoss is not J2EE certified[ Go to top ]

    Do you work for Sun? Have you seen the contract Sun gave to us?



    http://openenterprisetrends.com/cgi-bin/page_display.cgi?193

    "
    Saletta: They say [JBoss products] are not compatible because Sun wouldn't give them the test suite. Under the previous licensing arrangement, they [JBoss] couldn't get the test because it was tied to source license. That was an issue; however, that is no longer the case.JBoss was the first OS group we offered the compatibility test license, and it was not accepted
    "
  142. Cary;

    Thanks for your nice Saletta quote dated 4/1/2003. Now please refer back to this one which is dated 6/12/2003. http://www.infoworld.com/article/03/06/12/HNjbossdevs_1.html

    You are getting a little more creative with your names, but most of them are still derrivative. Try coming up with some more on your own.
  143. Happy with JBoss[ Go to top ]

    I recently designed and implemented a general use Pub/Sub notification system for a LARGE company, developed and deployed on JBoss 3.04 and Oracle 8i. It seems to me that JBoss is J2EE-compatible and performant enough to be useful. And the price is right (even big public companies try to save money now and then).

    The App. itself uses Stateless Session, Entity (CMP/CMR/CMT) and message-driven EJBs within JBoss. All seemed to work pretty much as advertised.

    Of course, there's small problems here and there. I did end up buying some of the JBoss Docs. Ooh ooh, $20 out of my pocket. I was happy to both obtain better information about the product, while offering some meager financial assistance to the team. Is that such a big deal?

    Here's a quick rundown of some of the problems I did encounter:
    1) JBoss CMP seems to have a column name length limitation, at least in 3.04.
    2) Using Oracle sequence numbers for CMP PKs was a little tricky.
    3) Yes, I'll admit to having a little trouble getting the stupid DataSource configured

    Other than those things, I was quite happy with JBoss. I had WAY more trouble with XDoclet (but that was well worth it too, in the end).

    Deployment on JBoss is especially sweet -- just drop in the EAR and go! For anyone coming from Weblogic (please don't flame, I'm stuck with WL 6.5 right now), that's a nice change of pace (EJBC -- UGH!).

    I say 'Keep up the good work' to the JBoss team!

    -Matthew S. Ring
  144. Happy with JBoss[ Go to top ]

    Matthew, your messaging works ok? I heard many times that JMS implmentation in JBoss is not production quality. Had no experience myself with it though...
  145. RE: Happy with JBoss[ Go to top ]

    Stepan,

    That's a fair question. I haven't heard how the JBoss messaging is holding up in production, as I changed jobs recently. I guess I would worry a bit under heavy load. But I bet it would only take a day to plug in a different JMS implementation, if needed.

    Luckily, AFAIK, the MDB wasn't really a mission critical part of the system, it simply broadcasts the event info to the subscribers (mostly via email). Since the events themselves are persistent entities, they are not lost in the event of a messaging failure.

    -Matt
  146. JBoss: A modern day plague?[ Go to top ]

    JBoss, if you dont like it, you can change it.

    Standards are good, but doesnt mean they are 100% correct.

    If you could, what would you change in EJB, JSP, servlets ?? I'm sure you have a lot of things in mind. Isn't the same with JBOSS ? Stop complaining, start helping.

    <Eternal life to open source/>
  147. I just want to say that it's wonderful to finally see so many developers out there finally come to their senses and discredit Marc's deceptive marketing strategy.
    I especially like Cary Bloom's comments. And about Rickard Oberg, people always said he was the world's greatest programmer: I don't know if it's true, but he is making very sound technical points here and if I didn't give him the benefit of the doubt before I do give it to him now.
  148. From my blog at:

    http://www.freeroller.net/page/hugopinto

    Feel free to comment.

    I guess the JBoss-themed discussions are now the state of the art for rant-blogging, so I'll add a couple of lines of my own.

    I use/dig/promote JBoss since it was named EJBoss (and was actually not more that Marc's mock-up code for EJB 1.0) and had it's only list on eGroups (now yahoo groups). I believe it was the end of 1999.
    I remember the whole story of Rickard's famous interceptor-based design for 2.0, the complete re-write, and the cool architecture thereafter. I ever wrote a couple of lines back then, that are still on the server's codebase.
    I remember when things started to cool down between Marc and Rickard, and when their close collaboration split off - and hey, up to this point it's pretty normal open-source, people come and go, just as I did, just as Rickard did.
    And up to this point, there was little business around the server (Marc had a professional services company back then, - Telkel or somethink sounding like it, I believe - that flopped), and JBoss Group The Company was still a mirage.

    Boundaries start to blur when JBoss Group is formed. Still today, many people cannot make the distintion between JBoss (the server) and JBoss Group (the company), and I am still not sure if Marc intended it to be this way or not.
    Marc created the JGB after the Telkel flop, and it was a company that would provide services over the codebase. Turns out that this time the momentum was correct, and it took off fine.
    Actually, the average seat price for trainings is so high that I'm surprised it did, but OK, good for JBG. Anyway, the this was that there was JBoss, the server, free, and there was a company named JBoss Group, that employed occasionaly the core JBoss developers, splitting revenue, to deliver certified services over the codebase.
    The only thing this company should have that made it special over other companies delivering services over JBoss was that it was owned and managed by the JBoss creator Marc Fleury (wich also is the brand's trademark owner).

    Then came the certification issue: Marc always upheld that it was his intention to certify JBoss as soon as Sun would let him. And with the latest revision of the JCP, Sun actually DOES allow to get open-source products certified.
    The problem is - Sun says - that JBoss is a for-profit organization, and does not qualify for free certification.
    Marc replies "gee, this Sun guys really don't like us! They're afraid will kill the J2EE market with a free certified server", and blames Sun over it. BUT WAIT A MINUTE! JBoss is not for profit! Not JBoss THE SERVER.
    Marc's company is. But he lets it pass (amongst other Sun-JBoss discussions, but never mind those, this public argument is enought for here). And this is the point where I got sure that there was "Marc's private agenda" written all over it.

    And with JBoss 4 (with it's ultra-cool AOP features), this became clear. Marc wants to stay in the grey area, where the server is not really certified (he is happy with "compatible", altought I don't know, nowadays, what it is compatible with... EJB 2.0? IIOP? Servlets? Wich versions?) and he can blame Sun for it, at the same time that they improve the non-J2EE features instead of pushing the not-so-cool compatibility flag.
    It's so easy. JBoss is not certified because Sun won't let them certify it because they say the server is for-pay, while they said it's free. Only problem is that Sun is (awaringly) not refering to the server but to the company.
    And Marc is (also awaringly) letting this pass away, because the blured boundary between JBoss Group and JBoss The Server is so convenient for him. In the end, he says they would certify JBoss happily paying for the certification if somebody supported the cost, but they still don't have the money. And the circle is closed.

    Couple of months ago, I asked on TSS these same questions, out in the open. Marc e-mailed me in private (I won't disclose the mail) and replied in a very pissed off way (as usual in him) - but did not address the questions I had posted.

    JBOSS is not 100% J2EE compatible. There is, I believe, a series of minor issues, such as CORBA compatibility, that are still lacking in implementation (and that hardly anyone uses). However, I believe that JBoss is one of the most compatible servers out there, probably above Websphere, rivalling with Weblogic and the rest of the gang. I use it, and 3.2 looks really solid. For J2EE applications, JBoss is in fact my first choice.

    Core Developers Network enters the arena. And some very important issues rise. What makes JBoss Group more qualified than CDN to certify JBoss, besides the ownership of the brand by Marc?
    With the entry of the CDN in the marketplace, JBoss Group is what is always should have been: just another JBoss-services company - which, by the way, should have as many right of using the JBoss brand as CDN - after all, Marc sould use the JBoss brand ONLY WITHIN THE CODEBASE AND THE PROJECT, not on his own private company.

    Now everyone has a choice on where to get services and support. More JBoss companies will rise. And this is when things start to get interesting. This is when open-source projects start to mature.

    Thanks for your time.

    Hugo
  149. Since the code is free for anyone to download...
    ...Why doesn't Sun run the tests in their 'labs' and disclose the results? Or any other company (BEA, IBM, Oracle, Orion) what's so secret about the test? If the Oracle is so spec compliant wouldn't it be benefical for them to release the results Oracle vs. JBoss.

    They are perfectly free to do so and all these companies would love to see JBoss fail miserably.

    Why not?
    Choice 1: Even if JBoss would fail the tests would show it to be more compliant than the 'big ones'
    Choice 2: The success of JBoss is critical for all the big ones because a good quality OS server brings more people to the market. The .NET side has much cheaper servers and more marketing power than all the J2EE groups combined.
    Think about apache. Without it the whole world would now be running on IIS. 90% of apache users chose it because of it's price. Those 90% would be now running IIS. Every JBoss download and installation takes a slice from Microsoft's cake.
    The big ones really need to lower their prices to make the products even remotely competitive to Microsofts offerings.
  150. The Certification issue black hole[ Go to top ]

    ...Why doesn't Sun run the tests in their 'labs' and disclose

    > the results? Or any other company (BEA, IBM, Oracle, Orion)
    > what's so secret about the test? If the Oracle is so spec
    > compliant wouldn't it be benefical for them to release the
    > results Oracle vs. JBoss.

    What I've heard is that each and every one of the certified servers needed to have significant work done on the test suite to pass -- mostly because the test suite itself is not up to par and contains bugs, the spec is ambigious, etc.

    So there is no single certification test suite that is run on all certified app servers and where you could compare the results. Therefore none of the J2EE certified server vendors actually publish any kind of results of the test. They get their "certified" stamp from SUN and are happy with it.

    That's the reason why Oracle for instance will not publish a test result comparison with JBoss (if it is even allowed under the licensing rules from Sun), nor is Sun willing to put the test suite to public to reveal the shambles it is in.

     
    > They are perfectly free to do so and all these companies would
    > love to see JBoss fail miserably.

    It would require some engineering effort (which means spending money) plus I suspect the vendors are not allowed to publish any results (so BEA cannot publically point out IBM's bugs for instance).

    /T
  151. Oracle, BEA,and IBM is not going run a comparison test with JBOSS. b'cos its not their work. They have other things to do.
  152. life's too short[ Go to top ]

    I've just wasted half an hour reading this thread. I'm gutted...
  153. Nipsudoofus:

    Obviously you don't have a clue as to how the certification works. Why would IBM or any other company spend their time running the certification tests for Jboss? Don't they have other things to do?

    Let's face it: JBoss DOES NOT PASS the certification test. If it did, we would've heard from the Jboss trolls by now. The fact that it is been 2yrs+ and still no certification should tell you that it is all a house of cards.
  154. Re: JBoss[ Go to top ]

    I can understand if someone doesn't feel comfortable with the attitude of certain people at the JBoss Group, dislikes the coloring of their web site, and perhaps thinks that this and that part could have been implemented better or doesn't conform correctly to specification JX-30089.1c. Big deal. Glad to see that there aren't any real issues. Thanks, and keep up the good work!
  155. Re: JBoss[ Go to top ]

    Man u took my previous name! :-)
    Beware, you'll start to get comments about that. Alpha and Omega ;-P
  156. So what's the deal here ?[ Go to top ]

    who is this guy ?

    what exactly is it that would make me say: "Hmm, maybe he has a point?"

    Can somebody explain me why one would attack JBoss this way ? I mean, these guys did a lot for the open-source community, making a great application server available for everyone to download, as well as keeping up with the latest technologies... as long as it can be downloaded, checked out from CVS, access to the forums etc... it's great stuff.
    If it wasn't for JBoss the open-source J2EE landscape would not be as great as it is now, and you know it. Admit it, JBoss is quality stuff... sure there are issues, but you really think other app.servers don't have issues ?

    You cannot attack an open-source project! It makes no sense, think about it... If you don't like it, don't use it... turn away and go do something else, but don't write/say BS, there's enough of that already

    I have succesfully used JBoss in over 4 medium-sized projects, ranging from standard PetStore-like projects to a very complex PKI platform, not once have I experienced issues that could not be easily and elegantly resolved...
    what can I say, yes I love JBoss, because I have absolutely no reason to dislike it. Maybe it's just me but if someone else has another opinion I only see three possibilities:

    1. they have no clue of what they're doing, and get frustrated by the issues
    2. they have a personal 'thing' with JBoss (or J2EE in general)
    3. they are God-like geniusses that understand some horrible problem no else is able to grasp (although the type of 'rant' kinda rules this one out)

    I fall in neither of these categories, I am just a guy passionate (not fanatic) with Java development (coming from C/C++ and ASM makes it even stronger ;-)

    I have found nothing interesting to read in this 'rant', (I use quotes, real rants make me laugh... this was a rather boring one, sorry); maybe the guy lacks social skills and this is his way of venting (yeah, we've seen it all before)

    I think everyone agrees that Java is here to stay, being backed by a huge community of developers, open-source projects (jakarta, jboss, sourceforge,...) it's getting better everyday...

    just my 2 euro cents.

    greetz
    Wouter

    ===============================
       Java Enterprise Teamwork
     http://jeteam.sourceforge.net
    ===============================
  157. So what's the deal here ?[ Go to top ]

    "You cannot attack an open-source project!"

    Hmm.. Exactly why is that? Of course you always have a choice regarding using it or not using it.
    But this whole discussion about not having a right to critisize something because "its free" is plain stupidity.
    If you cannot point out short-comings, bugs and flaws, how are you going to progress?
    The "we wont take any criticism"-attitude is nothing but ego, its a reaction of denial that you as a human being actually can make mistakes!

    My tip for the JBoss people and just about everyone else: if criticism is unfounded, stupid or otherwise ridiculous, dont comment on it! People are intelligent enough to see thru it.
    And if (God forbid, how is this going to effect my ego?!) there should be some foundations to the criticism, deal with it, solve it, instead of trying to put it under the rug, plain deny it or excuse it with "hey, its free, so you dont have to use it".

    Isnt progress what we are all really after? I would like to see a bit more CONSTRUCTIVE criticism, and a bit more CONSTRUCTIVE receival of said criticism.

    If you think you cant make mistakes, well, you are obviously a lazy bastard not doing enough in that case!

    And for the record: I have used JBoss, I like the product personally, and for "plain" (Session and BMP) EJB development and use it is lovely, never had any problems (Not entirely true, had some problems with Hibernate and JBoss). However, I would be interested in knowing how its JMS implementation holds up?
    I havent looked into this in more detail (read: I know theres something in there called JBossMQ, and thats about it)...
  158. So what's the deal here ?[ Go to top ]

    "You cannot attack an open-source project!"

    >
    > Hmm.. Exactly why is that? Of course you always have a choice regarding using it or not using it.
    > But this whole discussion about not having a right to critisize something because "its free" is plain stupidity.

    you're right,

    as I was replying to the 'rant' I probably should have written "You cannot attack an open-source project in this way!"

    of course criticism is a good thing, especially when it is used in a constructive way, but I don't think the 'rant' here qualifies as critisizm ...

    as a matter of fact, I rarely read something which makes me say: "Hey, he's right! How about this <any-issue-here> ?"

    you have to admit, most of the 'critisizm' is unfounded and colored by personal feelings... in the case of open-source I think it's a very sad thing... and frankly it's getting a little annoying (but maybe it's unavoidable)

    I just asked myself the question: "So what did I learn from the 100+ posts in this thread ?".
    Not too much.

    anyway, before being sucked into the nerd vortex (as Cameron Purdy once eloquently put it, if I recall correctly), I know we all agree on this, so I'll just get back to work ;-)

    cheers

    Wouter.
  159. So good for these people's reputation. Of Jboss'. Or even Java. Like a bunch of kids fighting over a toy. Can we grow up now? Thanks.
  160. So what's the deal here ?[ Go to top ]

    Then we agree. I was more answering the line of discussion here, because besides the unfounded ranting of some people, I also found som valid points that should be brought to light (sometimes in the midst of otherwise unfounded rants). Such as some of the more specific technical points.

    The bad blood between some people (Whatever the reasons) and personal attacks from and against the JBoss people I could care less about (as long as I am not in the middle of it, which I am most definitely not).

    I care about facts.
    Although its always fun to discuss quasi-philosophical issues in a civilized manner.. :)
  161. So what's the deal here ?[ Go to top ]

    Maybe this discussion isn't all that stupid. Maybe there's an issue here that needs to be addressed.

    Personally, I couldn't care one bit about Marc or any of his people at JBG. Their attitude is their issue, and their loss. That stance is probably true (not to mention appropriate) for most people, except a few cheezed off users and developers who have the right to be discontent. The rest of those who are annoyed are probably doing it because its fashionable.

    That said, the incremental amounts of vitriol flung against JBoss over the past several months is just not understandable. I asked Rickard for an explanation, and frankly I am not satisfied.
    Remember, we are talking about a project that is one of only two credible open source projects in the J2EE app server realm today. Also remember that it beat several other app servers with more market share to become a finalist in the recent JavaWorld awards (for whatever that's worth, and yeah, yeah so there were nomination issues).

    Rickard just blogged again about JBoss AOP (dreambean.com). OK, maybe this version of it sucks. Allrighty. On to the next issue.

    I take it the performance of the EJB2 implementation hasn't drastically changed since 3.2.1 (Which I know performs decently enough. I build products using it.) I don't think that too many IT projects around the world are going to come to a standstill if these AOP-related performance issues are not resolved soon.

    It's difficult to understand the *true* basis for all the hate flying around. When I hear all the shotguns going off, I ask myself - has the JBoss group done more to harm the J2EE world than help it?
    I don't think so. Neither do most of the people taking potshots at the project, I am guessing. As I said earlier, I think it's must seem downright fashionable to do this today. Wow.

    So you folks out there with the guns, please put them back in your pickup trucks and mosey along, will ya? There's REAL software that needs to be built today.

    Sandeep.
  162. So what's the deal here ?[ Go to top ]

    Spot on Sandeep.

    I personally think it comes down to jealousy. Marc and his gang are living life that the ranters only dream of. They don't have the skill set so they are jealous.

    Rickard is doing a good job of playing referee. I consider Rickard to be one of the best technical analyst out there today. He has a the background since he has implemented an aop framework and product based on it. A few months back he was complaining about jboss aop not being true aop because it did not handle introductions.
    Well guess what, it does now and he is not complaining about whether or not jboss aop is aop or not. Now he is complaining about performance, well guess what i have confidence that by the time jboss 4 puts out a production release these issues will be addressed.
    I for one am glad Rickard is paying such close attention to the details. He did it with the petstore and now he is doing it with jboss aop. In a way Rickard is making jboss aop better by being so vocal.
    He sees something he does not like, tells everyone about it, brings it out into the open, it gets discussed, it improves. Just goes to show that developing software in the open produces better quality software.

    The thing I find troubling is the amount of emotions that are involved. I am probably guilty of that myself. Religion and technology do not mix. I personally have a lot invested in jboss.
    Spent time learning it. Attended training. Put apps in production using it. And personally feel it is one well designed piece of software.
    So of course I would like to see it succeed. Those that want it to fail have probably not invested much time in it. I invested time in it out of necessity (read free), but now it is not out of necessity that I use it. I use it because when it comes down to it it is a good piece of software.
    And I said thanks by buying the doco.

    Thank you jboss group and core developers net for your hard work in which I am jsut one beneficiary.

    tx

    Matt
  163. So what's the deal here ?[ Go to top ]

    Its probably a mix.. bad attitude at times from some of the JBoss folks (Hey, they are only humans!). The only thing is that when you are in a public position such as these people are people will jump on you in an instant if you make a wrong step, its sadly the price of being a somewhat public figure.
    So in that aspect: it wont be the first or last time this kind of thing comes up, they are after all only human. Even though you probably should think once or twice before getting emotional when you have the spotlight on yourself.
    A humble and low-key attitude never got anyone hated. But it has gotten a lot of people respected.

    As for jealousy? You can bet your shiny nurd *** that jealousy plays a big part!

    I am from Sweden originally, a country where people will backstab and badmouth you if you have any considerable success. And dont even think of parking that Mercedes in a public place, someone with a WV WILL use their keys to scratch the paint just because they are jelous.
    I´ve seen it all when it comes to those bad sides of human nature, and it will with 100% certainty play some part here as well.

    Why do people always put Marc F and Bill Burke down? Probably because they are there! They are active in the community. I bet not many people here have the slightest clue of whos responsible for WebShphere or Weblogic, hence they dont get to take the crap the JBoss people get.

    Ok, I wont brown-nose the JBoss people anymore now. They do have a lot to learn when it comes to Public relations and general attitude and ego-control. They are doing a good job in my eyes, but never let ego blind you from learning from your own mistakes.
    As I said earlier: If you dont make mistakes, you are not doing enough.
  164. JBoss: A modern day plague?[ Go to top ]

    Bottom Line: It works. Cost: a little time learning the basics. A pretty good deal I would say. I never needed to purchase documentation either.

    Craig Martin
    Rochester NY.
  165. Look here for the press release:

    http://www.webmethods.com/press_release_detail/1,1075,2988-2003-2,00.html
  166. Professionalism[ Go to top ]

    Its a bit amusing seeing this guy comment on professionalism, when he obviously has no clue about it himself. If the JBoss Group is as unprofessional as he implies (which I personally believe they are not), then he should go to work for them as he would fit right in.

    What is amazing is that TSS would actually publish completely worthless information such as this. What is even more amazing is that so many people (like myself) are spending so much time replying to this completely worthless information.

    Scot.
  167. I think that many of you guys are blind Windows users.

    Try to see what's happening with patents today, look at SCO Unix diatribe, for exaple.
    You can't immagine how harmfull patents are.
    No matter if that software has a cool feature or runs better. You are only playng to the hangman game with your own future (look also at what's happened to some big companies in US, none of them is really shure).
    The fact is that JBoss is as mine as yours and with a software like JB I'm sure I don't run in any legal trouble, and it will grant to me always the same rights I have today.
    That's the point!!!!
  168. Great Advertising[ Go to top ]

    For those who truly hate JBoss your doing a great service to them with the best advertising possible.
    So poeple read this, what do you think they will want to do ? .. Try it!

    Now lets face it, is there any product on the market that is 100 percent perfect.

    Not one intelligent person would take this forum seriously because if something is truly that bad then poeple dont have to talk about it, poeple know.

    Jboss has a large user base and its not getting any smaller either, its free and it works.

    It's pretty simple if you like it use it, if you dont then dont, anything beyond that is just an aberation like this particular forum.
  169. JBoss: A modern day plague?[ Go to top ]

    I just find it iteresting that no one is addressing the issues raised.

    I installed jboss. The log upon start up was 1141 KB in size. It had 8448 lines, 12 WARN messages, 3 Exceptions, and one stack trace. This is from an unmodified jboss distrubution. Trying to find real error messages in all that crap when you actually deploy is like looking for a neadle in a haystack. This level of useless data makes it very hard to work with when (not if) you ever have a problem.

    The fact that my app can take over the logging of the appserver is bad. But whats worse is that the logs from the app server are now intermixed with my individual app's log.

    An app server that only allows me to deploy one version of a library at a time has a broken classloader. When I load a jar in my web-app it should not have any effect outside the web-app.

    JBoss has some nice ideas. But they need to take a step back and fix what they have befor they move on and add AOP to the mix.
  170. JBoss[ Go to top ]

    I just find it iteresting that no one is addressing the issues raised.


    Perhaps that is because knowledgeable and educated users don't consider this an issue. The logging levels are not set in stone you know. You **can** configure them. Go to jakarta.apache.org and figure out how log4j works.

    > I installed jboss. The log upon start up was 1141 KB in size.
    > It had 8448 lines, 12 WARN messages, 3 Exceptions, and one
    > stack trace. This is from an unmodified jboss distrubution.
    > Trying to find real error messages in all that crap when you
    > actually deploy is like looking for a neadle in a haystack.

    Learn how to filter out INFO messages if you don't want to see them. This is not rocket science.

    > This level of useless data makes it very hard to work with
    > when (not if) you ever have a problem.

    Trying to program an app server that is able to work around a stupid uneducated user is hard too.

     
    > The fact that my app can take over the logging of the appserver is bad.
    > But whats worse is that the logs from the app server are now intermixed
    > with my individual app's log.

    Learn how to configure log4j correctly.


    > An app server that only allows me to deploy one version of a library
    > at a time has a broken classloader.

    Agreed. This is not JBoss though. Learn about how to define separate classloader repositories.

    > When I load a jar in my web-app it should not have any effect outside
    > the web-app.

    Configure your classloader repository correctly.

    > JBoss has some nice ideas. But they need to take a step back and
    > fix what they have befor they move on and add AOP to the mix.

    Nah, I don't agree. There isn't a single app server out there that is able to help you if you don't read the documentation. Whining about the logging system when it is so obvious how to configure it is lame. Shows you haven't done your homework.

    Have a nice day and better luck next time!

    /T
  171. JBoss[ Go to top ]

    I just find it iteresting that no one is addressing the issues raised.

    >
    > Perhaps that is because knowledgeable and educated users don't consider this
    > an issue. The logging levels are not set in stone you know. You **can**
    > configure them. Go to jakarta.apache.org and figure out how log4j works.

    This is nr 1) on my list above.

    > > I installed jboss. The log upon start up was 1141 KB in size.
    > > It had 8448 lines, 12 WARN messages, 3 Exceptions, and one
    > > stack trace. This is from an unmodified jboss distrubution.
    > > Trying to find real error messages in all that crap when you
    > > actually deploy is like looking for a neadle in a haystack.
    >
    > Learn how to filter out INFO messages if you don't want to see them. This is
    > not rocket science.

    This is nr 1) on my list above.

    > > This level of useless data makes it very hard to work with
    > > when (not if) you ever have a problem.
    >
    > Trying to program an app server that is able to work around a stupid
    > uneducated user is hard too.

    Nr 5)

    > > The fact that my app can take over the logging of the appserver is bad.
    > > But whats worse is that the logs from the app server are now intermixed
    > > with my individual app's log.
    >
    > Learn how to configure log4j correctly.

    Nr 5) (and Nr 1)

    > > An app server that only allows me to deploy one version of a library
    > > at a time has a broken classloader.
    >
    > Agreed. This is not JBoss though. Learn about how to define separate
    > classloader repositories.

    Nr 1)

    > > When I load a jar in my web-app it should not have any effect outside
    > > the web-app.
    >
    > Configure your classloader repository correctly.

    Nr 1)

    > > JBoss has some nice ideas. But they need to take a step back and
    > > fix what they have befor they move on and add AOP to the mix.
    >
    > Nah, I don't agree. There isn't a single app server out there that is able to
    > help you if you don't read the documentation. Whining about the logging system
    > when it is so obvious how to configure it is lame. Shows you haven't done
    > your homework.

    Nr 5)

    Oh well.

    /Rickard
  172. JBoss[ Go to top ]

    Did you have a point or are you just now learning how to count numbers?

    /T
  173. jboss[ Go to top ]

    well, I guess Rickard was only demonstrating how nicely the items in his list of complaints (see above) about the jboss attitute matches the attitude of a certain group of jboss defenders.

    He does not say that defending JBoss is generally a bad idea, he only critisizes the WAY this is done, i.e., in the case of this last postimg, the angry tone and the inability to explain things without an implicit personal attack in it.

    Personally, I would like to have a good and strong JBoss App Server - but, seriously, i think some of their policies, e.g. the documentation policy, really bring more harm than good. No wonder that people begin to distrust them, and raise questions even about the legitimacy of JBoss project's claim being pure Open Source. Sure, some vital parts of it are OS, but if other, equally vital parts (e.g. documentation) are not freely available, how would you call that (i did not study marketing, so i do not know how this is called ;-) ?

    joe
  174. jboss[ Go to top ]

    He does not say that defending JBoss is generally a bad idea,

    > he only critisizes the WAY this is done, i.e., in the case of
    > this last postimg, the angry tone and the inability to explain
    > things without an implicit personal attack in it.

    I guess you're not familiar with the expression RTFM. I'd like to see the app server that works without the user bothering to read the documentation.

    > Personally, I would like to have a good and strong JBoss App
    > Server - but, seriously, i think some of their policies,
    > e.g. the documentation policy, really bring more harm than good.
    > No wonder that people begin to distrust them, and raise questions
    > even about the legitimacy of JBoss project's claim being pure
    > Open Source. Sure, some vital parts of it are OS, but if other,
    > equally vital parts (e.g. documentation) are not freely available,
    > how would you call that (i did not study marketing, so i do not
    > know how this is called ;-) ?

    But here again you are confusing JBoss Group and JBoss. No one is preventing you from writing a kick-ass documentation and giving it away free of charge. You're just expecting that the people who wrote the server should provide this very same service for you. And when they do, you complain about the quality of their work.

    Nothing in the definition of Open Source says anyone has the obligation to write a piece of documentation when bunch of whiny users come knocking on their door. Just like no one is forced to apply a patch into their code base if they see no need for it.

    What you are asking for is ludicrous.

    /T
  175. i agree to disagree[ Go to top ]

    He does not say that defending JBoss is generally a bad idea,

    > > he only critisizes the WAY this is done, i.e., in the case of
    > > this last postimg, the angry tone and the inability to explain
    > > things without an implicit personal attack in it.
    >
    > I guess you're not familiar with the expression RTFM. I'd like to see the app server that works without the user bothering to read the documentation.

    well, the term RTFM is normally used tongue in cheek, but you guys really BELIEVE in this term, right..

    a free app server needs a free docu, period. Yea, i think there is a free getting started docu, but this is a similar marketing teaser like 30 day evaluation periods.


    > But here again you are confusing JBoss Group and JBoss. No one is preventing you from writing a kick-ass documentation and giving it away free of charge. You're just expecting that the people who wrote the server should provide this very same service for you. And when they do, you complain about the quality of their work.
    >

    May i quote from jboss.org "Here at JBoss, the developers make money by selling their advanced documentation. We get the developers to directly write a professional document, professionally edited and given directly to you, directly from the source. No more cretins writing documentation on things they don't understand, no more poorly documented open source features."

    Okay, if i am a Jboss developer (my current skill set would not allow that anyway...) i would get problems if i don't play in their game.

    And if a jboss user writes documentation, he/she is only a "CRETIN WHO DOESN't UNDERSTAND", and i can already imagine how good and kind answers i would get if i need someone to ask a question for my documentation project...

    NOPE, the JBoss web site clearly says "GO AWAY" to potential documentation writers.


    > Nothing in the definition of Open Source says anyone has the obligation to write a piece of documentation when bunch of whiny users come knocking on their door. Just like no one is forced to apply a patch into their code base if they see no need for it.
    >
    > What you are asking for is ludicrous.
    >

    I don't think so. If they want to compare themselves to Apache, then they should try to deliver the same quality, which includes good and free documentation for a "bunch of whiny users".


    so, i stick with my opinion: An app server with non-free documentation is like trialware with some expiration date - you can play with it for free, but as soon as you want to do some professional work, you suddenly have to pay some money. (sure, the docs are cheap - but is it really cheap for students, for people in india, pakistan etc, or for people who just want to "give it a try" in their sparetime?)


    just my 2 cents,
    joe
  176. i agree to disagree[ Go to top ]

    a free app server needs a free docu, period. Yea, i think there is a

    > free getting started docu, but this is a similar marketing teaser
    > like 30 day evaluation periods.

    Believe it or not, there is a free documentation written by the "community".

    > NOPE, the JBoss web site clearly says "GO AWAY" to potential documentation
    > writers.

    JBoss web site is the JBoss Group website. They sell documentation.

    Stop making stuff up, do a web search on google, the web is full of JBoss documentation pieces. Nobody is telling all these people to "go away". If you don't like what is available for free, well what are you going to do about it? Write your own.

    > I don't think so. If they want to compare themselves to Apache, then
    > they should try to deliver the same quality, which includes good and
    > free documentation for a "bunch of whiny users".

    I doubt Marc Fleury wants his company be like Apache (i.e. make no money) but I could be wrong, you'd have to ask him ;-)

    > so, i stick with my opinion: An app server with non-free documentation

    As you already said, there's free documentation. Try to be more consistent. If you have an issue with it, that's fine. You don't like it? Ok. But claiming that there is none when you just said there is makes you look silly.

    > (sure, the docs are cheap - but is it really cheap for students, for people in > india, pakistan etc, or for people who just want to "give it a try" in their
    > sparetime?)

    Uhh, yes it is really cheap. It's cheaper than any book you can buy from a bookstore.

    /T
  177. sure[ Go to top ]

    All i said, is that i do not like the current policies (e.g. regarding documentation) and i don't think they are helpful for jboss in the long term.

    My mistake in this short discussion was obviously that i spoke about my idea how an OSS project should or shouldn't behave, but probably i did not make explicit enough that it is my "opinion" not necessarily the ultimate truth.

    I understand your counter-arguments (e.g. doing a google search for some pieces of JBoss docs, starting to write own docu - well maybe you are right and the CDN guys will provide a free docu!) but as long as Jboss (or JBG or whatever) stays in its current state, i will not be able to say "yes, this is a good open source project".

    joe
  178. i agree to disagree[ Go to top ]

    Joe,
        Few points.
    First,I dont use JBoss. I dont like JBoss. Before any JBoss supporter blow my head off, the reason i did not like it is b'cos it did not give me, what i was looking for at that time. And money is not a big concern, b'cos what my company was looking for was solid technology,not just saving dollars.

    But, i will not say, JBoss is a bad product. As i said in one of my previous posts, JBoss has its own market.

    About document, When you are comparing the cost of other products,spending little money on JBOSS is stil worthy.
    Why you are calling people in India, and Pakistan? Why you left so many other countries in the World??

    Let's make things clear. JBOSS is a company, with own business model. It is not a charity organisation. Neither they have loads of money to throw away everything they got.
    If people are interested, i would expect them to spend atleast something. Atleast.

    But It does not mean, JBOSS is anything better than other Servers. Giving something for free DOES NOT make JBOSS is better than other servers.

    If you ask me, will I use JBOSS in my next project, answer is probably NO.
  179. JBoss: A modern day plague?[ Go to top ]

    I just find it iteresting that no one is addressing the issues raised.

    >
    > I installed jboss. The log upon start up was 1141 KB in size. It had 8448
    > lines, 12 WARN messages, 3 Exceptions, and one stack trace. This is from an
    > unmodified jboss distrubution. Trying to find real error messages in all that
    > crap when you actually deploy is like looking for a neadle in a haystack. This
    > level of useless data makes it very hard to work with when (not if) you ever
    > have a problem.
    >
    > The fact that my app can take over the logging of the appserver is bad. But
    > whats worse is that the logs from the app server are now intermixed with my
    > individual app's log.

    Aagh! I promised to quit this thread but *locates an auxiliary power source* must.go.on.for.one.more.time

    Funny I checked my startup log and it's 48KB and 626 lines (not a single error). Yes, I've actually configured the server before production use! (After that 626 lines it stays silent if there aren't errors in application code.) I've also disabled some services like clustering, IIOP and Axis (JBoss version is 3.0.7)

    Here's my 'top five' list: ;-)

    Hint:
    1. Edit Jboss/server/xxx/conf/log4j.xml
    (where xxx is the name you specified for your configuration) if you didn't it's probably small, default or all.)
    2. Locate the section Preserve messages in a local file
    3. Locate appender name="FILE"
    4. Change param name="Threshold" value="INFO" to value="WARN"
    5. Logical, eh?

    This way you still get the full info to console (some 600 lines in my case) but the log file gets only the serious messages.

    If you bought the documentation there's also a pdf-file called JBossLog4j.pdf but it's not very good. (It's actually quite poor)

    > An app server that only allows me to deploy one version of a library at a time
    > has a broken classloader. When I load a jar in my web-app it should not have
    > any effect outside the web-app.

    Luckily we don't have to run multiple versions of our software simultaneosly. Otherwise this would be a problem.

    > JBoss has some nice ideas. But they need to take a step back and fix what they
    > have befor they move on and add AOP to the mix.

    Agree with you 75% Moving forward is an important thing but so is current quality. That's why I hope that they keep improving the 3.2 'series' as well as spending time with 4.

    No more replies, I swear ;-)
  180. JBoss: A modern day plague?[ Go to top ]

    From the
    horse's mouth

    -------------------------------------------
    It would be nice if you all posted your comments on our development forum rather than on your own blogs. Otherwise I have to search for it or have somebody email me saying somebody commented on it.

    Please comment on our development forums. I pay attention to them, I don't always remember to go look at somebody elses blog.

    Bill
    -------------------------------------------
    He sounds like a reasonable guy. Why is it that so many of you in this forum resort to personal insults. Someone called me 'nipsudoofus' and told me I haven't got a glue. Well maybe I don't. I'm just a programmer and I tried to come up with some explanations to this issue. There's no need to go on shouting: "THE FIRE ENGINE IS RED!"

    See ya all in next thread. This one's boring.

    P.S.
    TSS guys - please come up with more interesting articles! You could also moderate the threads and generally make it unacceptable to blatantly insult other people.
  181. QA[ Go to top ]

    All of Dion Almaer's posts should be QA moving forward because his posts constantly script divisiveness and "fanatical" spins about open source within our community. TMC seems to have a friendlier disposition for Microsoft than Java Open Source.


    Dion Almaer (Mr. Chief Architect),

    Don’t forget to test new functionality for TSS across different browsers (i.e. phoenix and opera). News related thread displays are not functioning properly on TSS. In case you are not aware there will no longer be a standalone version of Internet Explorer. It will be integrated within windows OS.
  182. JBoss Provides Solid Reflection.[ Go to top ]

    It's great that there's open source application server implementations such as JBoss to reflect and learn from. Unfortunately, this article does not address any solid technical reflections.

    I'm surprised TSS would bother highlighting such a useless post on their front page!?

    Dave./