Discussions

News: Oracle-Sun deal, what it means for Java/JVM?

  1. Could it be the end of free Java/JVM and Java become propetary or the closed world again?, As Stephen Colebourne talks in his article that Java could become again with a new Java trap. It is posible never will be Java7 spec to continue the block of Harmony?. Article at http://www.jroller.com/scolebourne/entry/no_java_se_7_the1

    Threaded Messages (32)

  2. This is a different aquisition by Oracle so can't derive too much from earlier aquisitions. 1. Sun is a Technology R&D Vendor, Software Product Vendor, Hardware Product Vendor, Network Innovation Vendor, etc. So Oracle didn't do such multi-dimensional aquisitions, and this time will deal with Sun brands very carefully without loosing any base. 2. How is other Java community members to cooperate with Oracle? Yes it is very difficult esp. with IBM, however Sun was/in in the same shoes of Oracle 3. End of Java/JVM Obviously Oracle is big supporter of Sun not the killer of Java and since most of its revenues based on Java technologies based. So we can expect a good care and cure from Oracle. Fortunately Sun didn't slip into Microsoft kind of companies, they will certainly kill Java. 4. 3 Major players: Microsoft, IBM and now Oracle in Stream A? Yes its true, who knows probably Google may go into Oracle or vice versa in future depends upon their standings in the market 5. Too early to predict any outcome now
  3. Many like me feel very sad about this news when Java (and the whole community ) is put under such pressure of how Java will be taken care by Oracle? With the same pace and passion of Sun? Lets wait and see.
  4. How about JDBC?[ Go to top ]

    I've always thought JDBC was an incredibly clumsy interface. Basic use is incredibly cumbersome, primarily for that 1% of users so people who do things like scroll through cursors and update them. 99% of us just want to execute a simple operation for update/insert/delete and maybe return the entire contents of a select query to an easy-to-navigate object (like the way ColdFusion handles queries). Why does JDBC have to be so difficult? Would a lot less people bother with JPA/Hibernate if JDBC wasn't so poorly designed? You'd think being acquired by a database company would dramatically help Java. Then again, Oracle's DB is anything but friendly and easy to use. Does anyone see a new future for JDBC and JPA?
  5. Re: How about JDBC?[ Go to top ]

    JDBC is past. Forget about it. More interesting question: will Oracle add to Java language something like LINQ?
  6. Re: How about JDBC?[ Go to top ]

    Forget JDBC
    JDBC could be simpler, but Oracle has a bad track record of making things simpler...
    Something like LINQ
    Sure, there is. For example JaQu at http://www.h2database.com/html/jaqu.html (I wrote that :-) There are other projects. JPA 2.0 will probably have a metamodel API: http://blogs.sun.com/ldemichiel/entry/java_persistence_2_0_proposed
  7. Re: How about JDBC?[ Go to top ]

    > Forget JDBC
    JDBC could be simpler, but Oracle has a bad track record of making things simpler...

    > Something like LINQ
    Sure, there is. For example JaQu at http://www.h2database.com/html/jaqu.html (I wrote that :-) There are other projects. JPA 2.0 will probably have a metamodel API: http://blogs.sun.com/ldemichiel/entry/java_persistence_2_0_proposed
    I checked out the examples of your project, and I'm not dogging on it at all, but without language modifications to Java you can't get LINQ. And I'm not even talking about comprehension syntax, but you really need closures..and gravy is extension methods. And just to put my $0.02 into this, Oracle is probably much worse than IBM for advancement of the Java language and definitely worse for the Java desktop.
  8. Re: How about JDBC?[ Go to top ]

    > Forget JDBC
    JDBC could be simpler, but Oracle has a bad track record of making things simpler...

    > Something like LINQ
    Sure, there is. For example JaQu at http://www.h2database.com/html/jaqu.html (I wrote that :-) There are other projects.
    -100 for LINQ-like extensions to java language.
    JPA 2.0 will probably have a metamodel API: http://blogs.sun.com/ldemichiel/entry/java_persistence_2_0_proposed
    JPA confirms to be an inferior spec about persistence. Metamodel is already available since a couple of months in JDO. But, shhhh, don't tell others. Guido
  9. Re: How about JDBC?[ Go to top ]

    -100 for LINQ-like extensions to java language.
    JaQu isn't an extension to the Java language. It also doesn't require an extension. JaQu works with Java 1.5.
    JPA confirms to be an inferior spec
    JPA is based on Hibernate, which is very popular. JDO is less popular. Google says 1 million pages for JDO, 4 million for JPA. This doesn't make the spec better any, but JPA is more popular.
    Metamodel is available in JDO.
    Do you have a link with some examples?
  10. Re: JDO Metamodel[ Go to top ]

    Do you have a link with some examples?
    Hi Thomas, try http://www.datanucleus.org/products/accessplatform_1_1/jdo/metadata_api.html as a start point. The structure approximately follows the JDO XML. --Andy (DataNucleus)
  11. Re: How about JDBC?[ Go to top ]

    > -100 for LINQ-like extensions to java language.
    JaQu isn't an extension to the Java language. It also doesn't require an extension. JaQu works with Java 1.5.

    >
    OK, but I was referring to the possibility to have a LINQ-like extension.
    JPA confirms to be an inferior spec
    JPA is based on Hibernate, which is very popular. JDO is less popular. Google says 1 million pages for JDO, 4 million for JPA. This doesn't make the spec better any, but JPA is more popular.

    > Metamodel is available in JDO.
    Do you have a link with some examples?
    JPA more popular ? Looking at your figure, yes. But, as you say, it does not make it a superior spec. About popularity, maybe part of the answer is this http://www.theserverside.com/news/thread.tss?thread_id=54328#307700 JDO was not in J2EE and I think the this was/is a problem for J2EE vendors, because persistence specification would be out control (i.e., read here, possibly not barbed wired to J2EE). And rarely data are not persisted :-) Guido Guido
  12. Re: How about JDBC?[ Go to top ]

    I've always thought JDBC was an incredibly clumsy interface. Basic use is incredibly cumbersome, primarily for that 1% of users so people who do things like scroll through cursors and update them.

    99% of us just want to execute a simple operation for update/insert/delete and maybe return the entire contents of a select query to an easy-to-navigate object (like the way ColdFusion handles queries). Why does JDBC have to be so difficult? Would a lot less people bother with JPA/Hibernate if JDBC wasn't so poorly designed?

    You'd think being acquired by a database company would dramatically help Java. Then again, Oracle's DB is anything but friendly and easy to use.

    Does anyone see a new future for JDBC and JPA?
    This is the reason why Spring JdbcTemplate is still popular. :)
  13. Re: How about JDBC?[ Go to top ]

    I've always thought JDBC was an incredibly clumsy interface. Basic use is incredibly cumbersome, primarily for that 1% of users so people who do things like scroll through cursors and update them.

    99% of us just want to execute a simple operation for update/insert/delete and maybe return the entire contents of a select query to an easy-to-navigate object (like the way ColdFusion handles queries). Why does JDBC have to be so difficult? Would a lot less people bother with JPA/Hibernate if JDBC wasn't so poorly designed?

    You'd think being acquired by a database company would dramatically help Java. Then again, Oracle's DB is anything but friendly and easy to use.

    Does anyone see a new future for JDBC and JPA?
    JDBC difficult? Wow, maybe you should get a new career. Hibernate sucks.
  14. About choice in VM land[ Go to top ]

    VM have become a central piece of code for numerous languages and businesses (just remember Beyond Java but not the Virtual Machine). So, it's unfortunate that, while there are plenty of open source librairies, and while VM choice may shorten, AFAIK, there is no business-friendly open source, (quite) efficient, and well-recognized VM, in order to let the market enough open, and to foster innovation. We might suffer from this lack. Dominique http://www.jroller.com/dmdevito
  15. Personally, I could care less about Java 7 or Apache Harmony - I always felt moving Java to open source was a bad, bad idea, and I hope Oracle resumes control in the same way Sun did for so many years. That worked well. I don't see any reason for Oracle to restrict Java in any other way. The fact that it is free is the reason it is popular, I think ... and Oracle would force people to turn away from Java if they changed that.
  16. Personally, I could care less about Java 7 or Apache Harmony - I always felt moving Java to open source was a bad, bad idea, and I hope Oracle resumes control in the same way Sun did for so many years. That worked well.
    There's a balance between democracy and control. Java being open source is a good thing, provided it's open source in a useful way, i.e. if people actually have a platform for contributing fixes and changes, otherwise it's open source only figuratively. On the other hand, control is also a good thing, provided it's used responsibly and for architecting the whole platform out of all the different pieces. I'd like to see the role of an 'architect' in the java platform. There's too much anarchy in the platform, even in the JCP. Some of the simplest things, for example DI, are implemented in at least 4-5 different ways in various JSRs. If java had an architect he could have come up with a single and coherent solution. Ara.
  17. wrong for several reasons[ Go to top ]

    note: posted this to the original blog first but ran into an overzealous and broken spam filter. There's a few points to be made here: 1) OpenJDK is GPLv2, so without the improvements regarding IPR licensing in v3. Sun was a big proponent of v3, so going for v2 must have been a deliberate move. Their IPR is likely the reason. 2) IPR is actually distributed across the industry. Oracle/sun not putting OpenJDK 7 through JCP also means that the whole thing could backfire: e.g. IBM sueing. 3) Other OSS projects seem to function fine with GPLv2, just one implementation, and no explicit IPR deals. E.g. linux. It's all about risk management here. Could Oracle sue. Yes. Would they? No, it would mean exposing their customers to counter suits: very bad for business. 4) The whole point of Apache Harmony is that closed source & IPR friendliness (i.e. having it and not being forced to expose it). It's primary supporters are notorious for having big IPR portfolios. So, don't be naive about this. Harmony would not exist if people could live with the GPLv2 which is arguably a lot more liberal on these issues. 5) The test suite is a nuisance. But you are allowed to use it with OpenJDK derivatives. I.e. someone could fork OpenJDK and still use the test suite. Secondly, the test suite is not needed from a pure functional point of view. You could just write your own tests. It's only needed if you want the Java brand slapped on your product. Sun/Oracle is the rightful owner of that. 6) While OSS Java is very hip, much of the industry still uses non OSS Java 1.4 and 1.5. Java has gotten quite far without being open source. 7) Going from a full blown OSS GPL software base with an insanely huge amount of users (whole Fortune 500, world+dog, everybody else) to a closed source software base would be an unprecedented first. I think you give Oracle too much credit by even assuming they could pull it off. So altogether, I don't think Java and Java 7 are in trouble. I do think that if Oracle shows any hint of doing what you think they might do, the OSS community will step in and hijack Java OSS assets and continue under its own flag. That would be a bad deal for Sun/Oracle. Both Apache Harmony and any OpenJDK clone (e.g. iced tea) are perfectly ready for production usage. IBM or anyone else could just choose to throw their weight behind it, if they wanted. Mismanagement on behalf of Oracle will in my view lead to this situation very soon.
  18. Java as an ecosystem will die.[ Go to top ]

    It will be a death far more quickier than the one it would have had if IBM had bought Sun instead. Under slomo IBM, Java would have rusted and sat inertly to oblivion, but Oracle will try by any means to capitalize it, even if that means total distruption of the open source part of it. We will see Java run better under Oracle Solaris (it is already true btw), less bugs fixed on Linux, JCP driven to make Oracle offer the best one and similar techniques. IBM will fight this, producing far more fragmentation than the swt/swing dualism. And no, I don't think that .net will ever be a solution (6 incompatible and highly proprietary releases sinced 2002). Just my .2 eurocents. R
  19. swing/swt dualism[ Go to top ]

    IBM will fight this, producing far more fragmentation than the swt/swing dualism.
    I don't see SWT fragmenting Java. SWT was invented for one particular purpose, to speed up the UI part for IDEs. They did it right and the result it that we have Eclipse which is 100% Java except for the UI. I don't see SWT being used anywhere else, so all is good.
  20. Re: swing/swt dualism[ Go to top ]

    IBM will fight this, producing far more fragmentation than the swt/swing dualism.

    I don't see SWT fragmenting Java. SWT was invented for one particular purpose, to speed up the UI part for IDEs. They did it right and the result it that we have Eclipse which is 100% Java except for the UI. I don't see SWT being used anywhere else, so all is good.
    It was a de facto split from the canonical way to do GUIs. It could have bee softer simply by noy throwing away the entire Swing Set and simply by producing a Swing native implementation. And calling a project Eclipse, isn't it a statement of intents? R
  21. Why would Java die?[ Go to top ]

    .. less bugs fixed on Linux, JCP driven to make Oracle offer the best one and similar techniques.

    IBM will fight this, producing far more fragmentation than the swt/swing dualism.
    That assumes that it's not in Oracle's and IBM's best interests to be working together, which we already are on many fronts. Both IBM and Oracle have been actively supporting Java openness and standards, and even projects such as Apache Harmony. There are many more things in common between IBM, Oracle and other enterprise software companies than there are things to drive us apart. While Oracle actively competes with IBM on many fronts, IBM is today the _largest_ implementer of Oracle solutions, for example. Peace, Cameron Purdy Oracle Coherence: Data Grid for Java, .NET and C++
  22. Re: Why will Java die[ Go to top ]

    That assumes that it's not in Oracle's and IBM's best interests to be working together, which we already are on many fronts.
    Larry didn't want to shell out money just to buy a GPL technology, just like he didn't when he took RedHat Linux without paying a dime. But then he almost made it in his pants when he realized that IBM could take away Java. Today's (and tomorrow's) Oracle is a strongly Java-dependent company.
    Both IBM and Oracle have been actively supporting Java openness and standards, and even projects such as Apache Harmony.
    IBM did support open source in general and Harmony in particular. I seem not to recall big contributions to FOSS from the part of Oracle, sorry.
    There are many more things in common between IBM, Oracle and other enterprise software companies than there are things to drive us apart.
    I agree that big companies are alike: they have similar big numbers of employees to be fed. Sun on the other hand was somehow different: it had ideas and good implementations. If only didn't they buy that crappy MySQL. If IBM or Oracle, in this economic downturn and after it, shall find an economical advantage in sacrificing the Java ecosystem, they will at the speed of light.
    While Oracle actively competes with IBM on many fronts, IBM is today the _largest_ implementer of Oracle solutions,
    Yeah, this will be more and more true, since now they will be charted by Oracle, their biggest enemy, on the Java development. Do You think they will accept it? The more I think about this the more I realize that Java is screwed for good. R
  23. Why would Java die?[ Go to top ]

    Today's (and tomorrow's) Oracle is a strongly Java-dependent company.
    That is correct.
    I seem not to recall big contributions to FOSS from the part of Oracle, sorry.
    First of all, this isn't a race to the bottom. If giving away your IP via open source were the way to win, Sun would be crushing everyone. Second, it isn't some sort of religious contest. It's not like trying to show off how much money you give to your church. If businesses donate to your FOSS god, then they better have a really good business reason to do so .. otherwise they deserve to go out of business. Please recall that they are called "businesses" and not "charities" for a reason. Also, regarding your inability to recall, (and you seem to have forgotten your last name, Mr. "N/A" ;-) .. on the topic of Oracle and open source, consider: * MySQL sits on top of InnoDB .. that's an Oracle company. * The JPA reference implementation, a.k.a. EclipseLink, is the open-sourced TopLink product .. also from Oracle. * The most widely used database ever is BerkeleyDB .. also an Oracle company. Oracle contributes significantly to quite a few standards bodies and open source projects. In addition to Apache and Eclipse, consider the JCP, OSGi, etc. Oracle doesn't contribute just for the fun of it. The contributions are often for strategic purposes, i.e. things that directly or indirectly will help Oracle's business. Regardless of the topic of this thread, regardless of how it plays out, helping Java helps Oracle's business. It also helps a lot of other companies. It's a good strategic investment. Peace, Cameron Purdy Oracle Coherence: Data Grid for Java, .NET and C++
  24. Re: Why would Java die[ Go to top ]

    MySQL sits on top of InnoDB .. that's an Oracle company.
    InnoDB is a piece of *software*, currently owned by Oracle, after buying some other company that wrote it.
    EclipseLink, is the open-sourced TopLink product .. also from Oracle.
    Oracle owned software, bought from some other company that wrote it
    BerkeleyDB .. also an Oracle company.
    Some software bought from some other company. Warning : pattern emerging
    Oracle Coherence: ...
    Damn you done it again. Some software bought from some other company. Don't they write anything of their own ? ... Peas!
  25. Oracle and Open Source[ Go to top ]

    Just to add a couple more things that Oracle has done with Java and open source: Apache MyFaces Trinidad - probably the largest set of JSF components out there - Developed by Oracle Various Eclipse IDE projects (JSF, JPA/Dali, DB) - Developed by Oracle For more things that Oracle did with Open Source both in the Java and non-Java field have a look here: http://oss.oracle.com/
  26. Magnanimous Larry![ Go to top ]

    Just to add a couple more things that Oracle has done with Java and open source:
    Apache MyFaces Trinidad - probably the largest set of JSF components out there - Developed by Oracle
    Various Eclipse IDE projects (JSF, JPA/Dali, DB) -
    Oh! You're right! Trinidad and Dali!!! How could I miss them!
    Developed by Oracle

    For more things that Oracle did with Open Source both in the Java and non-Java field have a look here:
    http://oss.oracle.com/
    I looked at the list and I see an emerging pattern: All the projects are of the form: we take from open source, and we try to twist and suck from them (see Xen, or outragious unbreakable linux). The vast majority of the community projects that sees Oracle involved are: - drop to open source support some technology that the company sees as a cost - make some good open source project (like php) work better with Oracle DB Then Oracle vastly takes from Apache Foundation, for one the httpd server that it shameless rebrands, but it is not present in the list of companies that make donations: http://www.apache.org/foundation/thanks.html Where You can find GOOG and (Yuck!) MSFT !!! R PS: Don't take this as an Oracle bashing. I'm only spilling out the facts. MegaPeace!
  27. Re: Why would Java live?[ Go to top ]

    First of all, this isn't a race to the bottom. If giving away your IP via open source were the way to win, Sun would be crushing everyone.
    Correct, Cameron. Mind You that I'm no FOSS bigot. I only think that companies should trust the openness of their development tools if they have to invest into them. .Net is not a good choice not because it's not free but because its non-free status makes it a mean for microsoft to force you to upgrade often that every other year, thus increasing dependency and TCO. This is what i fear will happen to java.


    Second, it isn't some sort of religious contest. It's not like trying to show off how much money you give to your church.
    :-D I'd like to see some big managers in church! Following the analogy, they would steal from the plate!


    If businesses donate to your FOSS god, then they better have a really good business reason to do so .. otherwise they deserve to go out of business.
    FOSS is not a god but if I invest in tools I have to be assured that those tools won't become a black pit sucking money.
    Please recall that they are called "businesses" and not "charities" for a reason. [...] In addition to Apache and Eclipse, consider the JCP, OSGi, etc.

    Oracle doesn't contribute just for the fun of it. The contributions are often for strategic purposes, i.e. things that directly or indirectly will help Oracle's business.
    I agree to this point. That (plus the modus operandi of Big Companies) IS the scaring part. At least for us outside Oracle. Peace R
  28. Re: Why would Java die?[ Go to top ]

    I don't think Oracle will be a problem. Especially since we've put a lot of effort into Iced Tea. -- Bill Burke JBoss, a division of Red Hat Inc. http://bill.burkecentral.com
  29. Re: Why would Java die?[ Go to top ]

    I don't think Oracle will be a problem. Especially since we've put a lot of effort into Iced Tea.

    --
    Bill Burke
    JBoss, a division of Red Hat Inc.
    http://bill.burkecentral.com
    IcedTea? Ahem... You are joking Bill, right? I can't figure a big Telco or Bank deploying its vital systems onto that. R
  30. IBM = slowmo for Java?[ Go to top ]

    It will be a death far more quickier than the one it would have had if IBM had bought Sun instead.

    Under slomo IBM, Java would have rusted and sat inertly to oblivion, but Oracle will try by any means to capitalize it, even if that means total distruption of the open source part of it.

    We will see Java run better under Oracle Solaris (it is already true btw), less bugs fixed on Linux, JCP driven to make Oracle offer the best one and similar techniques.

    IBM will fight this, producing far more fragmentation than the swt/swing dualism.

    And no, I don't think that .net will ever be a solution (6 incompatible and highly proprietary releases sinced 2002).

    Just my .2 eurocents.

    R
    Surely you must be joking? IBM had a larger number of developers working on core Java enhancements than even Sun itself. Forget that their motives were not altruistic - the contributions of IBM to the Java and FOSS worlds are clear and undeniable. Xalan, SOAP, Eclipse, JavaOS - none of these would exist without IBM. Meanwhile the only technologies 'contributed' by Oracle were invented by companies that they bought. Even Toplink was bought from WebGain, which bought it from company of the same name. But I don't kid myself, if not for IBM's struggle with M$ to retain control of the server environment it would never have supported Java. But as long as Microsoft remains a threat in that arena they will continue to champion it, so I think they would have been a better choice than Oracle to take over the reins. With Oracle in charge of Java I predict M$ has a golden opportunity to regain alot of mind share in .NET technologies. Java's time is running out ...
  31. JPA and Metamodel[ Go to top ]

    Take look at http://www.leeonsoft.com for a new approach to persistence and its metamodel. It eliminates the complexity of JPA 2.0 metamodel and boosts productivity. You can download it from http://code.google.com/p/dreamsource-orm/downloads/list. Jim Xie
  32. mysql will die[ Go to top ]

    the consequence of this buy out is that 1) Mysql will die eventually . Oracle cannot have two databases competing against each other. 2) Solaris may survive depending on whether Oracle sees Linux as its future OS. Today Oracle has a strong market position in the Unix OS arena. However the question is that Oracle is not an OS vendor and doesn't have experience in this field. I suspect Aix may steal a good percentage of its market share. A lot of clients may go to AIX because they don't like Larry's "face". just like some oracle clients may go to DB2 now. Clients may want to diversify their suppliers or vendors to avoid lock in. 3) Java will survive. Oracle has got too much invested there. However, I suspect there will be 2 Java, one "free" and one "not free". Oracle does not have a reputation of giving full blown applications free without any constraint or restrictions. 4) Netbeans may die from the deal. Oracle most likely will keep its own IDE. Too bad I like Netbeans more than any other IDEs. 5) for Sun's appserver business, anything can happen there.
  33. Oracle buying Sun was purely commercial... just the same reason they acquired BEA. Overalapping products will obvisouly be sidelined and some may disappear. Regarding Java and JVM, everything will stay... Java has inflitrated too much into the commercial dataprocessing today. JVM (Virtual Machine is a total misnomer... It's just Java Runtime Environment) will continue to stay... Why should Oracle move away from either of them, when so many Oracle products/applications are dependent on them? JDBC/Hibernate/Spring etc Oracle won't bother so long as all of them are directly or indirectly using some of the Oracle products (Database, Weblogic etc).... Regarding mySQL, it will be around as a free database for quite some more time... Regarding 'Open Source' produsts we have too many of them. Every defect in one existing product is giving rise to a new wonderful product...I will be happy if Oracle with their muscle power put an end to some of them at least. It is nice for people to have something which is not so open but available free for testing. Oracle is already providing a lighter version of database free for test purposes. Once they do the same with other produts, many of the open source systems will become irrelevant!