Slashdot Mirror


Sun Application Server 9.0 PE Open Sourced

farble1670 writes "Sun Microsystems has released their Application Server 9.0 PE platform as open source, under the code name Glassfish. Version 9.0, when complete, will be J2EE 5 compliant. Code is released under Sun's CDDL (common development and distribution license), the same license used to cover the Open Solaris, the open-source Solaris operating system. This is most likely a response to the popular open-source application server JBoss, which has cut into profits for Sun as well as other major application server vendors such as BEA and IBM."

31 comments

  1. It was already free by Eric+Giguere · · Score: 4, Informative

    This is just an evolutionary step. The application server was already free for production use (deployment, not just development). But yes, it would be a response to JBoss and others. Java app servers are pretty much a commodity at this stage, pushing vendors to release them for free and then sell expanded versions with additional bells and whistles.

    Eric
    A blog about book publishing, making money, Google, etc.
    1. Re:It was already free by afd8856 · · Score: 1

      Of course, not just Java app server. Their python equivalent, Zope, has been open sourced for quite some time now.

      --
      I'll do the stupid thing first and then you shy people follow...
  2. Ruby? by jericho4.0 · · Score: 2, Interesting

    I wonder how people with a financial stake in Java app servers feel about Ruby on Rails?

    --
    "A language that doesn't affect the way you think about programming, is not worth knowing" - Alan Perlis
    1. Re:Ruby? by astrashe · · Score: 4, Insightful

      I'll probably get flamed, and rightfully so, because I'm really uninformed on this topic, and I'm going to shoot off my mouth anyway. I'm just a dilletante. But I don't think that rails is much of a competitor.

      Rails is more about rapid and agile development, and these big application servers are more about running big heavy duty enterprise apps.

      I have the feeling that most apps that run in these containers are speced out pretty completely, developed in concert with layers of management, etc. And then there are all of the java APIs, that do just about everything you'd want in an enterprise app.

      I sort of see rails as occupying a middle ground between PHP and a java app server. You get the structure of a java framework with the ability to know stuff out of PHP.

      But there's more of a learning curve than you have with PHP, and you don't get the really robust admin tools of a good app server and the fabled java APIs.

      So I think that they're really different tools for different kinds of jobs.

    2. Re:Ruby? by Eric+Giguere · · Score: 1

      different tools for different kinds of jobs

      Pretty reasonable statement for an "uninformed" view. You're right, get ready for the flames...
    3. Re:Ruby? by Anonymous Coward · · Score: 2, Interesting

      Back when I was a "web developer", we dabbled in J2EE (servlets & JSP, really) and it generally was just fine. However, after I left that project, the other developers tried to go whole-hog J2EE with enterprise beans and everything with WebLogic.

      Well, they talked professional and all and tried to be proud of their creation, but one thing they ran into was too many layers of abstraction. Their database was failing, and WebLogic's implementation ate all the important debugging information! They spent weeks trying to understand why their connection pools would go bad without warning and other problems.

      I don't blame J2EE at all, either. What I blame is that they took on too much stuff (resume stuffing, I think) too quick and without enough caution and patience. They didn't take the time to make the abstractions work for them rather than against them.

      J2EE is also one area where I really really wished the contract would have provided two weeks of real training. There is just too much documentation, and a nice primer would have made all the difference. But, of course, programmers are gods and don't need training. Just read the book and look at these examples in JavaPro. I quit that company after a few years.

    4. Re:Ruby? by Ogerman · · Score: 4, Interesting

      I sort of see rails as occupying a middle ground between PHP and a java app server. You get the structure of a java framework with the ability to know stuff out of PHP.

      That's a surprisingly accurate assessment for supposedly being "uninformed" about the topic. I have PHP and Java background and have recently researched Ruby a bit. At this point, Ruby would now be my choice over PHP for all but the lightest-weight jobs. If Ruby improves in the area of templating and component-based UI frameworks, it could knock PHP down quite low on the list and start to compete in the low-end Java arena.

      Java app servers, on the other hand, still run in an arena that the scripting languages can't begin to touch and by architectural nature never will. What is a more interesting comparison is modern lightweight Java tools (like Spring/Hibernate) vs. Ruby. These Java tools have a ways to go regarding loosing additional XML baggage, but they are still a significant improvement over current heavyweight J2EE app servers. What is most interesting is the convergence. Java tools are rising to the challenge of becoming simpler. Even the EJB 3.0 and J2EE 5 specs are heavily geared towards this goal. The question becomes: If Java tools get closer to the simplicity of scripting languages without losing the power of their architecture, where is the incentive to use scripting languages? Conversely, the scripting languages have simplicity nearly mastered but they have no headroom architecturally.

    5. Re:Ruby? by archeopterix · · Score: 1
      Similar experiences here. My last J2EE project lost a few precious weeks trying to get Websphere implementation of entity beans to work on MSSQL. Even though each transaction inserted some fresh records and operated on them (distinct sets of records), we still managed to get database deadlocks (strange kind of index range locks). We tweaked all there was to tweak (locking schemes, isolation levels, both on DB level and app server level). Interestingly enough, we couldn't recreate the deadlocks outside Websphere. We also tried substituting Oracle for MSSQL and the deadlocks vanished. Too bad that Oracle was not an available option in the production environment.

      Finally we dumped the entity beans altogether and went back to hand-written SQL. This got rid of the deadlocks and even busted the throughput almost twice, not to mention the relief from EQL which at that time sucked a big one (I don't know whether it improved, and I am not eager to try.)

      This was more than a year ago and I have moved away from J2EE. I am curious what's the status of the entity beans layer nowadays.

    6. Re:Ruby? by iwadasn · · Score: 1


      This is exactly correct. Appservers are often used for backend data processing. Usually something like, run a JMS server, let people publish data on it. Take the data, do something with it and a bunch of other data in a database. Make the results available through EJB, or possibly a web page. These tend to not be used for just webpages based on databases (simple shopping carts), but more for realtime trading, risk analysis, data processing, etc...

      Ruby on Rails and a J2EE appserver are completely different. However, it is possible to use an appserver to do the sort of things you'd do with Rails, the reverse is not true.

    7. Re:Ruby? by Anonymous Coward · · Score: 0

      very simple, rails does not scale.

  3. Glassfish? by cjsnell · · Score: 0, Troll


    The codename "glassfish" doesn't inspire a lot of confidence in this product. May I suggest "stonefish", "rockfish", or perhaps "swordfish"?

  4. Benefits from IBM too by mparaz · · Score: 3, Informative

    Glassfish uses Derby, which was open sourced by IBM from the Cloudscape embedded database.

  5. Bad Link by hritcu · · Score: 4, Informative

    The "J2EE 5" link in TFA links to OpenSolaris, not to http://java.sun.com/j2ee/5.0/index.jsp or http://www.jcp.org/en/jsr/detail?id=244 .

    --
    If you don't fail at least 90 percent of the time, you're not aiming high enough. (Alan Kay)
  6. Reaction to Geronimo as well by cant_get_a_good_nick · · Score: 1, Informative

    JBoss is not the only open source Appserver. Geronimo is moving quickly (recently passing the has passed the J2EE 1.4.1, on it's way to full J2EE compliance). IBM is providing support, a lot of it by donating Gluecode and other code to the apache project, and even BEA (even though they still sell WebLogic) contributing code to the Apache project, allowing their tools to target geronimo.

  7. This is great news for the open source community by hritcu · · Score: 0

    ... but apparently since it was NOT MADE HERE, GlassFish just can't be in any way better than the open source application servers, isn't it? The same way Open Solaris is not in any way better than Linux? So, of course, it is justified to bash Sun, even when they make a very good move for us.

    --
    If you don't fail at least 90 percent of the time, you're not aiming high enough. (Alan Kay)
  8. Geronimo by Ratbert42 · · Score: 0, Redundant

    Apache Geronimo is probably going to gain a lot of ground, especially with IBM backing it.

  9. Microsoft tightens Anti-Piracy policy! by jakin45 · · Score: 0, Offtopic

    As of today, July 26 2005, visitors to the Microsoft Download Center and Windows Update will be required to participate in Windows Genuine Advantage to access all content. That means a user of Windows 2000/XP must provide a software product ID to prove your software is legally licenced before full access is allowed. The world-wide anti-piracy initiative of WGA started today helps to further combat software piracy.

  10. Re:This is great news for the open source communit by GuyWithLag · · Score: 1

    Three words: Read The Licence

  11. Re:This is great news for the open source communit by hritcu · · Score: 1

    The CDDL is an OSI-approved open source license. What is your problem with it? It may not be compatible with the GPL but why should it be?
    And wait ... very few licenses are GPL-compatible. Are you also blaming the Apache License, the BSD license, the Mozilla Public License and many other open-source licenses, for beeing GPL incomplatible? Are they all evil or something?

    --
    If you don't fail at least 90 percent of the time, you're not aiming high enough. (Alan Kay)
  12. ...and once again... by Glomek · · Score: 1
    ...Sun releases a product which there are already credible Open Source versions of rather than what the community has been begging them for, an Open Source version of Java itself.

    Open Solaris doesn't matter at this point since we have Linux. Glassfish doesn't matter at this point since we have JBoss. When will they stop doing things that don't matter and do something that will matter, like Open Sourcing Java?!?

    1. Re:...and once again... by farble1670 · · Score: 1
      first, i don't think these two things are related. it's not like the sun execs sat down and said "okay we can either open source java or open source or app server, what should we do?"

      second, competition is good. even if you think app servers are a commodity. why don't you apply your "there already is one of those, we don't need another" logic to the entire economy? hey hondas are good cars, right? what is the point of making any other type of car?

      if you think that one application server is as good as another, that they are a commodity, that just illustrates tht you haven't worked with more than one of them. they are all j2ee containers, sure, but configuration is different, administration is different, HA options are different, scalability + performance are different.

    2. Re:...and once again... by Glomek · · Score: 1
      To be clear, I don't think it's bad that Sun has Open Sourced the things that they have. Yes, more options are good. I'm sure that both Open Solaris and Glassfish are cool. The thing is, neither of them will do as much good as Open Source Java would.

      The difference between zero credible Open Source implementations and one is much greater than the difference between 1 and 2 or 2 and 3. Another free operating system is good, yes, but it doesn't have the same impact as the first one did. Another j2ee container is good, yes, but it doesn't have the same impact as the first one did.

      We still don't have the first credible Open Source implementation of Java. That would make a much bigger difference than the nth (for n > 1) Open Source implementation of pretty much anything else.

    3. Re:...and once again... by Anonymous Coward · · Score: 0

      Open Solaris doesn't matter at this point since we have Linux.

      Using your logic Linux is pointless since we already had BSD.

  13. What's an "application server"? Really. by dotzie · · Score: 1

    No, I mean, really.

    What's an application server?

  14. Re:This is great news for the open source communit by Anonymous Coward · · Score: 0

    "Are they all evil or something?"

    No, there are very good reasons why large groups of developers have _chosen_ to wall themselves off from the FSF. The FSF is way too political and way too over-bearing for many people. In fact, I'd bet that most GPL fans are young and are still running on unchecked idealism. By the time they are 30, especially if they have real responsibilities like a business or family, they'll see why the FSF is good but not all that it's cracked up to be.

  15. At this point it's (sadly) irrelevant... by rkhalloran · · Score: 1

    After the whole Kiva -> Netscape -> iPlanet -> Sun fiasco, the company that *owns* Java (like it or not) has been completely unable to market a viable J2EE server product, and had their lunch split up and eaten between IBM & BEA. Then to add insult to injury, those nassssty OSS folks come along and roll their own (JBoss, Geronimo, JOnAS), and *those* have more street cred than Sun's product.

    They may as well throw the source out there, nobody really cares at this point...

  16. Re:This is great news for the open source communit by Anonymous Coward · · Score: 0

    Three words: Read The Licence

    Three words: Stop spreading FUD

  17. Mod parent down (a little) for advert NOT in sig by tod_miller · · Score: 1

    Sigs should be in sigs.

    I hate to do it, but come on Eric: Your signature isn't even a signature: - I have signature viewing turned off, yet your 'make easy money with google.com' website advertisment is still showing.

    Keep your signature IN your signature, I don't believe you went to the trouble of copy/pasting a signature into your messages each and every time just to by pass peoples OWN choices in viewing signatures, that slashdot gives them. Luckily I just RIPped that link out of slashdot, so I won't see it again, but please respect the netiquette of slashdot (what little there is!).

    Mods: Don't be fooled by the location of the advert - it is technically part of the comment, thus OT (it would have been fine, and totally ok with me if this advert was in his signature).

    --
    #hostfile 0.0.0.0 primidi.com 0.0.0.0 www.primidi.com 0.0.0.0 radio.weblogs.com
  18. Re:Mod parent down (a little) for advert NOT in si by Eric+Giguere · · Score: 1

    It's not a conspiracy, Tod, just laziness on my part. I vary my signature on every post, that's all. Sometimes I point to my latest book, sometimes to my HTTP header tool, sometimes to my cereal box commentary, sometimes to nothing... It's just easier to cycle through them that way than changing the preferences all the time. Sheer laziness, which is kind of funny, because laziness is why signatures were invented in the first place... I guess I'll have to reform my ways...

  19. Re:Mod parent down (a little) for advert NOT in si by tod_miller · · Score: 1

    Hrm, well you got me there. Laziness is the undeniable excuse for everything. Sorry I over reacted a bit I guess. I realise I don't actually care :-)

    Your comment was actually pretty insightful too, well, stet.

    --
    #hostfile 0.0.0.0 primidi.com 0.0.0.0 www.primidi.com 0.0.0.0 radio.weblogs.com