Discussions

News: Applying JRL: A developer fixes the JDK

  1. Applying JRL: A developer fixes the JDK (19 messages)

    Andy Tripp decided to put the new Java Research License to the test, and documented fixing a few issues with the JDK step by step, through acquiring a copy of the source, compilation, coding, and sending patches back to Sun for inclusion in the next version of Java.

    His three bugs of choice were some compilation errors from the source, correction of some icons on Windows, and adding a Reader for Properties.load(). In each case he fixed the bug, and then sent the patches to Sun.

    In the issue of the compiler warnings, a Sun engineer responded within four days that the patch had been accepted and should be in the integration build in two weeks (which would place it in the integration stream around the second week of April, 2005).

    For the Windows icon issue, an engineer responded that the fix wasn't complete, that another engineer was fine-tuning the process for loading the icon from Windows.

    On the Properties.load() change, he had not received an update by the time of his writing the article, which might indicate that it was still under review.
    In all three cases, Sun did a good job of communicating: I got an automatically generated form letter immediately after submission of the fix, and a followup email from a real person once they had analyzed the fix. And once a fix had been incorporated, I got a third email telling me so (and letting me know I get a free T-shirt).
    And, of course, the good news is that my submitted fixes seemed to find their way to the right person within Sun. That may not sound like a big deal, but it's probably quite an accomplishment for a fairly large company like Sun.
    His conclusion final conclusion was also positive:
    Overall I'd give Sun a 'B+', or '3 stars out of 4', or 'very good' rating.

    Threaded Messages (19)

  2. Good News[ Go to top ]

    Good News for java.
    It means that it does not have to be open source in order to be open.
    Different models than open source can work.

    L. Orbach
    Multiple Book Price Comparison
    Javasight - Java News & Books
  3. Good News[ Go to top ]

    Different models than open source can work

    You unbeliever! :)
  4. Good News[ Go to top ]

    Work for whom? "Collaborative source" for instance seems to work more for the software company than for the collaborators...
  5. Good News[ Go to top ]

    "Collaborative source" for instance seems to work more for the software company than for the collaborators...

    Define work.
    If by 'work' you mean 'make money', then yes, this model 'works' better for the company than the collaborators (software sales).

    If by 'work' you mean better software, then I don't see many advantages for FOSS over this more "closed" model. In fact, the quality of Sun's guardianship is not likely to be surpassed by an independent, volunteer group. The requirements for submitting a patch to Sun is time-consuming, but rigorous. I don't see how this is a bad thing.
  6. Good News[ Go to top ]

    Work for whom? "Collaborative source" for instance seems to work more for the software company than for the collaborators...
    Won't you (me, everyone) benefit from Andi's patches?
  7. Good News[ Go to top ]

    Work for whom? "Collaborative source" for instance seems to work more for the software company than for the collaborators...
    Won't you (me, everyone) benefit from Andi's patches?

    +1 !!!
  8. Good News[ Go to top ]

    Work for whom? "Collaborative source" for instance seems to work more for the software company than for the collaborators...
    Won't you (me, everyone) benefit from Andi's patches?

    Sorry, let me clarify. I was not referring to Java itself in this comment. I have no issues with the JRL or how Java source code is protected, licensed, developed, etc. because I can still download and use the JDK from Sun for free. I agree absolutely with you, we will all benefit from the hard work, patches and contributions of people like Andy. My comment was referring to software projects with "pay or contribute" source code licenses that require you to write code for the company just to use their software. It turns out that few developers completely offset the licensing costs in this way.
  9. Good News[ Go to top ]

    Work for whom? "Collaborative source" for instance seems to work more for the software company than for the collaborators...

    Well how is it diffrent from model used by say jboss?

    And how do you imagine to have an open source project which gives commmit rights to random people? In good os projects there is always somebody who controls the code, makes planning and decides if patches are coherent with the line of the project.

    michal
  10. Good News[ Go to top ]

    Well how is it diffrent from model used by say jboss?

    There are two key differences:

    1) Ability to fork and distribute

    If JBoss doesn't accept your patch you can still
    do it yourself and *distribute* the result.

    2) Production usage

    You cannot use your patch in a commercial environment
    *until* Sun releases the official version with your patch
    included.
    In good os projects there is always somebody who controls the code, makes planning and decides if patches are coherent with the line of the project.

    Correct.

    xxxxxxxxxxxxxxxxxxxxxxxx
    Adrian Brock
    Chief Scientist
    JBoss Inc.
    xxxxxxxxxxxxxxxxxxxxxxxx
  11. Good News[ Go to top ]

    Correct.xxxxxxxxxxxxxxxxxxxxxxxx Adrian BrockChief ScientistJBoss Inc.xxxxxxxxxxxxxxxxxxxxxxxx

    Just for the record ...
    What does a Chief Scientist do ?
  12. two key differences[ Go to top ]

    Adrian Brock>>There are two key differences.

    Production usage:
    For any Quality consious company this is essentially a good thing. If I make a fix to JDK and want use the fix in production i definitely want to leverage the QA process that SUN got in place.

    Ability to fork and distribute:
    What percentage of teams want to own a JDK distribution? Probably 0.00001%? I say even that number is high.
  13. Good News[ Go to top ]

    There are two key differences:1) Ability to fork and distributeIf JBoss doesn't accept your patch you can still do it yourself and *distribute* the result.
    </blackquote>

    Isn't' it something that you discourage and actively protect yourself against it using LGPL licenses? Haven't JBOSS accused Geronimo of forking pats of JBOSS and threaten them with lawyers?

    Don't get me wrong - I respect your rights to protect your business - but in "commercial world" forking is not something which you can always do with open source software
    and not every type of fork is allowed.

    <blackquote>
    2) Production usage You cannot use your patch in a commercial environment*until* Sun releases the official version with your patchincluded.
    This is no issue with JDK. For projects written in commercial environment nobody uses bleeding edge and unstable jdks. There is still
    a lot of companies considering switching now to jdk ... 1.4. And really proliferation of forked version of jdks is the worst thing which could happen to java.
    And I don't think that sun is the only entity which decides which patches will be included. API additions are at least partially controlled by JCP.

    Michal
  14. Good News[ Go to top ]

    This is no issue with JDK. For projects written in commercial environment nobody uses bleeding edge and unstable jdks. There is still a lot of companies considering switching now to jdk ... 1.4. And really proliferation of forked version of jdks is the worst thing which could happen to java.And I don't think that sun is the only entity which decides which patches will be included. API additions are at least partially controlled by JCP.Michal

    Forked versions would be the death of Java as they would inevitably lead to API changes and core language changes.
    After all, if you allow people to fork some things but not others what's to prevent someone from forking the core language and the classfile format?

    Byebye WORA, hello version hell.

    It's exactly why the GPL zealots want Java to be released into their clutches, so they can destroy its greatest appeal which is the platform independence.

    As it is anyone is already allowed to write a JVM and even call it such IF it passes the compatibility test.
    You can even make that thing open source if you want to, but forked versions will still have to pass the compatibility test.
  15. Good News[ Go to top ]

    Actually, you can use your patch in a commercial environment - you just can't distribute it under the JRL. There is a license available for those who wish to distribute a patch to the JVM, as well.
  16. Good News[ Go to top ]

    Well how is it diffrent from model used by say jboss?

    There are two key differences:

    1) Ability to fork and distribute - If JBoss doesn't accept your patch you can stilldo it yourself and *distribute* the result.

    2) Production usage - You cannot use your patch in a commercial environment*until* Sun releases the official version with your patchincluded.

    Correct and well put. Note the difference between the two points that Adrian made and the original hypothesis:
    "Collaborative source" for instance seems to work more for the software company than for the collaborators...

    I don't think that Sun's approach is perfect, but I also don't think that there is a perfect. It's a balance between the needs of stability (which includes WORA) and the needs of those using Java. In that respect, this approach is a definite improvement, and it's not going to be free for Sun either, meaning they're not just dishing off all the work onto the community. Quality doesn't come cheap.

    Peace,

    Cameron Purdy
    Tangosol, Inc.
    Coherence: Cluster your POJOs!
  17. Spoiled by Open-Source[ Go to top ]

    Gosh, guys, where in the whole wide software business world can you actually submit patches and be heard?
    Ever tried to send a patch to Microsoft? ;-)
    Downloaded the source code of .NET lately?

    I'm still amazed by Sun's move and only good can come from it.
    It may not be your wet open-source dream, but it's very close.
  18. I think this is good news for Java. I´m asking myself if it would be good to release JDK as free software.
    I think the risk of forking the code in several platforms is quite high if this happens. It seems to me that Sun is doing a good job unifying the people´s expectations.

    But, from another point of view, maybe a truly open community could give more flexibility to the java platform.
    Anyway, I stringly believe this are good news for Java.

    Joserra
    New bussiness models with free software (oh!, in Spanish)
  19. But, from another point of view, maybe a truly open community could give more flexibility to the java platform.

    I lost my confidence in open communities after 2 best Linux distros (RedHat and Suse) are not for FREE anymore ...
  20. While I love and use open source, anyone who doesn't understand that without commercial interest, involvement and support, we wouldn't have the open source community we have; exactly where do you think this money comes from? If it wasn't for that support we'd all have to be farmers writing software at night. All these so-called leaders are paid very well for their time, thank you very much.

    Woodstock is over and we're not gonna hold hands singing how we want to buy the world a Coke. So buy a vowel or phone a friend, do whatever you have to; you people gotta calm down.