Slashdot Mirror

← Back to Stories (view on slashdot.org)

Blackdown Releases a 1.4.1 JDK

Posted by michael on from the choices-choices dept.

gholmer writes "The Blackdown project has finally released a production version of Java 1.4.1 for both ix86 and Sparc on Linux. This much-awaited release gives Linux users another choice for Java besides Sun's and IBM's."

20 of 64 comments (clear)

  1. gcj? by crow · · Score: 4, Interesting

    What about using the Java front-end for the Gnu Compiler Collection?

    (I'm not a Java developer, but I was under the impression that it, also, was another choice besides Sun's and IBM's.)

    1. Re:gcj? by smileyy · · Score: 4, Informative

      If you're talking about the native compilation aspect of gcj, there's a goodly number of things that won't work when compiled with gcj. Most notable is Swing.

      If you're talking about the bytecode compilation aspect of gcj, then you still need a JVM to run those bytecodes in.

      --
      pooptruck
    2. Re:gcj? by KDan · · Score: 2, Interesting

      So hold on, maybe someone knowledgeable can answer this question for me... if I write, say, a servlet, that I integrate with, say, Jetty, then I can compile that with gjc and have it running as a NATIVE application under linux?

      If the answer is yes, is there a way to do this under windows too??

      Daniel

      --
      Carpe Diem
    3. Re:gcj? by Spy+Hunter · · Score: 3, Interesting
      you still need a JVM to run those bytecodes in.

      GIJ, the GNU interpreter for Java, is also included in the GCJ package. GCJ isn't a finished product yet, but when it is, it will be a complete JDK and JRE (compiler, VM, and libraries) with the added ability to compile to native code if you want to, and all open source under the GPL.

      --
      main(c,r){for(r=32;r;) printf(++c>31?c=!r--,"\n":c<r?" ":~c&r?" `":" #");}
    4. Re:gcj? by damiam · · Score: 2, Informative

      Yes. You might have to distribute it with the cygwin DLL, but the user wouldn't need a full cygwin installation.

      --
      It's hard to be religious when certain people are never incinerated by bolts of lightning.
  2. Comparison to Sun's Java? by stefanlasiewski · · Score: 2, Interesting

    I'm curious, why would I want to use Blackdown or IBM's Java over Sun's Java? Am I missing some wonderous features or something?

    Any tables out there comparing the various Java Flavors?

    Seems to me that Sun's Java is the most mature of the Java's, and that Sun engineers have the most experience with Java and therefore will probably create the best Java implementation.

    Am I wrong?

    --
    "Can of worms? The can is open... the worms are everywhere."
    1. Re:Comparison to Sun's Java? by tradervik · · Score: 5, Interesting

      IBM's 1.3.1 JDK for Linux had noticably better performance than the Sun JDK in certain areas. There is an extensive performance report posted on www.javalobby.org. I think you have to register (free) to access the report.

    2. Re:Comparison to Sun's Java? by Anonymous Coward · · Score: 2, Informative

      PowerPC linux does not have binaries from Sun. IBM and Blackdown have binaries for PowerPC Linux. IBM Java is much better than Blackdown Java preformance wise.

  3. good stuff from blackdown by Anonymous Coward · · Score: 3, Interesting

    Jvaa WebStart rocks! apt-get, emerge, windows installer, bsd ports all pale in comparison to WS. You just go the web page of the app click a link and the program downloads and installs itself. The downloaded app doesn't have full access to your system resources (printer, network, loacl disks etc.) until you give it permission. So you can download random safe self-contained applications without worring about malicious authors. Each time you run the program it will (if the networks available) check the server for new versions, and automatically upgrade. Its so quick & painless. Where i work we have a intranet app thats rolled out to 600 people and since we started using websart what was a real admin head-ache is now something that just happens. The app is maintained by us and upgrades could only be once a month, because we had to guarantee everyone used the same version - not as easy as you think in practice, but webstart is my favourite thing ever!!!

    finally with the blackdown release the webstart icons will be integrated into the gnome desktop so the java app will launch just like a native app (is done this in windows for ages) our linux users will be so happy!!!!

    i'm so pleased about this i just pissed my pants

  4. Finally! A JDK compiled with GCC 3.2! by jake_the_blue_spruce · · Score: 3, Interesting

    Now we can finally build mozilla with 3.2 and drop all this crazy crap we've been doing to work around it. I can't believe Sun hasn't put out a 3.2 compiled version yet (plans for 1.4.2 to be though). I don't know about IBM. Anyone?

    --
    "There's so much left to know/ and I'm on the road to find out." -Cat Stevens
    1. Re:Finally! A JDK compiled with GCC 3.2! by Anonymous Coward · · Score: 2, Informative

      So compile Sun's JDK yourself. Yes, you CAN do that. Anyone who says otherwise is an idiot who hasn't actually looked into it.

  5. Re:Heh by Anonymous Coward · · Score: 2, Insightful

    everything in java is about choice

    from mobile-phones too mult-million dollar servers

    on linux-bsd-windoze-solaris-osx-symbian-....

    code compiled to JVM in hundreds of programming languages other than java

    compile to native using gcj

    use naitive widgets using SWT (think Eclipse)

    full (no embrace extend) support for open standards

    inclusive community based evolution through JCP

    allows open source implementations

    support from oracle, ibm, appache, nokia, motorola,

    native java operating systems such as savaJe

    java accelerated hardware components in modern cell phones

    so where do you want to go today?

  6. they already have a superior option by HaiLHaiL · · Score: 2, Informative

    http://www.bea.com/products/weblogic/jrockit/index .shtml

    awesome server side JVM

    --


    reech bee-yond ur clip-0n
  7. But how's it diffrent ? by noselasd · · Score: 2, Interesting

    I sure would like to se some detailed changlog on the blackdown work. Its based on the sources from Sun, now what the heck is diffrent ?

  8. Re:The other way around by mythr · · Score: 4, Informative

    J++ was never the problem. The problem was the broken Java VM that was being distributed with Windows. It's the same thing that happened with IE, actually. Few people would actually go out of their way to install the real VM, so Java was broken on a large share of the computers out there.

  9. Re:The other way around by Anonymous Coward · · Score: 2, Interesting

    "J++ was never the problem."

    Tell it to Sun and the Courts. The antitrust ruling against MS was based on the theory (that I don't subscribe to) that developers would develop a Java application on Windows using MS's extentions and be suddenly suprised it wasn't cross-platform. It wasn't about standard Java breaking on Windows.

    If there are incompatibilities at the JVM level, it's probably because Sun wouldn't allow MS to update the JVM due to their dispute about J++. Now, of course, MS doesn't want to include Java at all which also pisses off Sun.

  10. Computer Software Ideology by LinuxXPHybrid · · Score: 2, Insightful

    > I'm curious, why would I want to use Blackdown or IBM's Java over Sun's Java? Am I missing some wondrous features or something?

    I doubt it. Also Sun's technologies from Java to UltraSparc are well specified and designed so that their technologies are more open (standard) than open source, so we don't really have to go for Blackdown.

    But from Computer Software Ideology stand point of view, it is nice to see that Java spec is well defined and has good license terms so that other vendors can implement their own version (production level quality right?) and distribute it legally and freely. That has been proved by Blackdown today.

    Personally, I'd stick to Sun's Java (even though some claim that it's not best Java), but announcements like this makes Java even more attractive to me (than ... dot something).

  11. NUMBERS: by thufir · · Score: 5, Interesting

    A project I am working on involves using soap (ApachesSOAP) as a transport layer and performing serialization of data to xml using Castors xml abilities (so xerces as well). It also uses the JDO part of castor to persist the data, and also to keep logs and some more complex things as well (PostgreSQL for this test). (I am running system on a Linux 2.4.19 machine with a 2.20GHz Intel CPU. (No swapping occured)

    Here are some numbers for a test involving simply serializing one of the complete object trees of data using castor, in a loop executed 1000 times:

    Sun JDK 'java version "1.3.1": (avg/3) 5.8s
    Sun JDK 'java version "1.4.1_01"': (avg/3) 6.4s
    Blackdown 'java version "1.4.1": (avg/3) 5.3s

    Sending a message with the the above generated xml full cycle through the system (multiple threads of execution here, multiple database connections as well (pooled), passing data over soap, etc), looped 200 times took the following times: (again, avg.)

    (sun 1.4.1): 44.2s
    (sun 1.4.1): 44.6s
    (bd 1.4.1): 41.4s

    In both the coded test, and the real world situation, Blackdown's JDK outperformed Sun 1.4 and 1.3 jdks.

    1. Re:NUMBERS: by aled · · Score: 2, Interesting

      Did you tried the one compiled 3.2? that may explain the difference. supposedly gcc 3.2 generates faster executables than previous versions and Sun's java (I guess) has been compiled with gcc 2.9x.

      --

      "I think this line is mostly filler"
  12. Re:The other way around by jazuki · · Score: 2, Interesting

    "It wasn't about standard Java breaking on Windows."

    You're off a little bit. It was about standard Java breaking on Windows, namely JDK 1.1. When Sun released JDK 1.1, the first non-toy version of Java, Microsoft decided that they would take bits and pieces of new JDK, but substitute their own incompatible pieces for the rest. The trademark fight, the first legal battle between Microsoft and Sun on the matter, was whether Microsoft could still call the result Java. The court said they couldn't. And so they called it J++.

    For example, Microsoft refused to use JNI, using their own RNI instead. Of course, they claimed RNI was easier to use, and perhaps in some ways it was. However, it meant that if you wanted to integrate native code and support the Microsoft VM, you either had to do dual development, or code to MSVM only.

    Interestingly, Apple has had their own native interface too, called JDirect, which was even easier than RNI. But they fully suppored JNI as well, and you could mix the two. (The interface has changed since then, but there are still dual-supported interfaces.)

    Why didn't Microsoft take this approach if they really thought they could do better? That's the crux of the matter. Sun and others would argue that Microsoft was trying to break cross-platform Java. I would argue they have a point.