Oracle's Bold Plans for Java Bode Well

Discussions

News: Oracle's Bold Plans for Java Bode Well

  1. Oracle's Bold Plans for Java Bode Well (50 messages)

    Fatal Exception's Neil McAllister suggests that any doubts regarding Oracle's stewardship of Java have been dispelled now that Ellison and company have made it clear that they are very interested in making Java an even stronger alternative to .Net. 'We have the money to invest in Java, because Java is a very profitable business for us already,' said Ellison, whose plan for integrating Sun technology is ambitious, serving an even more ambitious goal: to create a soup-to-nuts tech juggernaut akin to IBM in the 1960s. Java will remain a key component of this push, with a new Java runtime, greater modularity, better support for non-Java languages, improved performance, and multicore-optimized garbage collection in the works, McAllister writes. Also revealed are plans to unify the Java SE and Java ME programming models and APIs and to enable JVM to run natively on hypervisors, allowing developers to run multiple Java instances on a single virtualized server. http://infoworld.com/d/developer-world/oracles-big-bear-hug-java-bodes-really-well-021

    Threaded Messages (50)

  2. Re: Oracle's Bold Plans for Java Bode Well[ Go to top ]

    How close is Mr. Neil McAllister to Mr. Larry ? This debate wont end till we really see the next few years go by... no point speculating, reading between the lines of every comment from Larry or any other Tom, Dick or Neil.
  3. speculating[ Go to top ]

    "no point speculating, reading between the lines of every comment from Larry or any other Tom, Dick or Neil. " You must be new here.
  4. Shawn,
    No point speculating, reading between the lines of every comment from Larry or any other Tom, Dick or Neil.
    I agree wholeheartedly. While it is definitely a good idea to keep a close eye on things to make sure things are really going in the right direction (I think they are), it's also pointless/too early to speculate/interpret way too much right now (unless of course one enjoys it - which I personally don't). Cheers, Reza
  5. Excited for Oracle and Java[ Go to top ]

    Any plans to compete more with .NET excite me. For years, the .NET guys (lower paid, less thinking in my opinion) have been given all the great tools to make things better for their development. Java developers sometimes get the tools but many times due to cost and the fact that free stuff is available are forced to learn it all. And, many times this learning is done at night (for the love of Java). If Oracle brings more horsepower to Java that may be a good thing. It still may not. The jury is still out on this. But, the JCP is almost dormant it seems, things are very slow, not a lot of interest, etc. Maybe a change was needed. I hope so. I love Java and I still believe it to be the best enterprise solution. David L. Whitehurst
  6. Re: Excited for Oracle and Java[ Go to top ]

    David,
    The JCP is almost dormant it seems, things are very slow
    From where I stand, this appears to be more perception from a certain set of developers than is reality. I would suggest carefully taking a look at the changes in Java EE 6 alone. It's not too difficult to see that they are far more significant both in terms of impact and number than any other single comparable framework or platform. If that isn't software evolution at work, I am not sure what is. It is certainly the case that those of us working with the JCP are always working to find ways to improve it, whether in the standard or in specific implementations like GlassFish, Seam, JBoss, Hibernate, Resin, CanDI, etc (and typically the best features end up in the standard). Cheers, Reza
  7. Re: Oracle's Bold Plans for Java Bode Well[ Go to top ]

    Fatal Exception's Neil McAllister suggests that any doubts regarding Oracle's stewardship of Java have been dispelled now that Ellison and company have made it clear that they are very interested in making Java an even stronger alternative to .Net.
    OMG, save us from Oracle executives that think Java is alternative to .NET. I just thought .NET is an alternative to Java!
  8. Sun ceo quits?[ Go to top ]

    Fatal Exception's Neil McAllister suggests that any doubts regarding Oracle's stewardship of Java have been dispelled now that Ellison and company have made it clear that they are very interested in making Java an even stronger alternative to .Net.


    OMG, save us from Oracle executives that think Java is alternative to .NET.
    I just thought .NET is an alternative to Java!
    not that it matters too much - but as expected i heard SUN CEO quit. Tweeter quitter they call him.
  9. Re: Sun ceo quits?[ Go to top ]

    Shawn,
    SUN CEO quit.
    That's what I heard as well. As far as I am concerned, this is a good thing. Management and not engineering (well, at least for the most part) is where Sun's real problems had been. In comparison, it seems Larry Ellison is better liked at Oracle, although it's hard to see why personally given some of the over-the-top statements here and there... Cheers, Reza
  10. Java is not alternative to .NET, it's true. But only because .NET is now much better than Java.
  11. Java is not alternative to .NET, it's true.

    But only because .NET is now much better than Java.
    I've been using both Java and .NET for over 5 years. My experience is different. .NET isn't "much better" than java. It's comparable. If you're talking thick client, .NET has many advantages. If you're talking middleware, Java has many advantages. It all depends on what you're doing. Most of the .NET projects I've been on resulted in "re-inventing" the wheel because Microsoft doesn't provide the solution I need. for me competition has been good. it's nudged both sides and both offerings better. peter
  12. Re: Oracle's Bold Plans for Java Bode Well[ Go to top ]

    Java is not alternative to .NET, it's true.

    But only because .NET is now much better than Java.


    I've been using both Java and .NET for over 5 years. My experience is different. .NET isn't "much better" than java. It's comparable. If you're talking thick client, .NET has many advantages. If you're talking middleware, Java has many advantages. It all depends on what you're doing. Most of the .NET projects I've been on resulted in "re-inventing" the wheel because Microsoft doesn't provide the solution I need.

    for me competition has been good. it's nudged both sides and both offerings better.

    peter
    Yeah. Pretty funny seeing how I am right now finishing up converting a Winform app to Eclipse RCP because it wasn't better. I've been doing .NET longer than 5 years and I will admit that there are somethings that are better and for somethings it is better (let me think for a bit on that :) ). But the Cons outweigh the Pros. Back to the main topic, I hope Oracle can shore up the loose ends. And take things forward, like JavaFX for instance.
  13. I've been using both Java and .NET for over 5 years. My experience is different. .NET isn't "much better" than java. It's comparable. If you're talking thick client, .NET has many advantages.
    Yes, SUN is incapable of creating a good GUI framework. SWT is nice, but it's already obsolete.
    If you're talking middleware, Java has many advantages.
    3 years ago I would agree with you. But not now. Let's see: 1) LINQ + Entity Framework/NHibernate beats Java in data access, hands down. 2) ASP.NET is _now_ way better than JSF can ever hope to be. 3) For lightweight stacks there's Cassini + Castle Windsor/Spring.NET - roughly equivalent to Tomcat+Spring. 4) Enterprise Services are a bit inferior to EJB+JCA, I'm ready to concede this point. What else?
    It all depends on what you're doing. Most of the .NET projects I've been on resulted in "re-inventing" the wheel because Microsoft doesn't provide the solution I need.
    Thanks, I choked on my drink. How many web frameworks were invented for Java because SUN can't standardize its way out of a paper bag? Personally, I stopped caring to count them looong ago. And we STILL get new ones, witness the 'Get to know Apache Click' link on the TSS frontpage. And, most importantly, a lot of Java developers don't seem to care. They think it's OK if Java becomes a new COBOL.
  14. Alex,
    3 years ago I would agree with you. But not now.
    I could not agree more. .NET has made some pretty impressive gains, not to mention Microsoft's historic strength on the UI/IDE end (including Silverlight). The Java world would be foolish to ignore these facts. My understanding is that the JSF 2 and JPA 2 folks took a close look at ASP.NET and LINQ and feel they are competitive at this point. On the component model end, we did not see much to add from the .NET world into CDI/EJB 3.1 (if you see something missing that is obviously beneficial, I'd certainly love to hear it). Cheers, Reza
  15. My understanding is that the JSF 2 and JPA 2 folks took a close look at ASP.NET and LINQ and feel they are competitive at this point.
    Nope, JSF/JPA folks are just as deluded as ever. Just temember EJB2... LINQ is a 'game changing' technology in data access, I'm now pretty sure about it. And I don't see any real attempts to make a comparative technology from Sun.
  16. My understanding is that the JSF 2 and JPA 2 folks took a close look at ASP.NET and LINQ and feel they are competitive at this point.


    Nope, JSF/JPA folks are just as deluded as ever. Just temember EJB2...

    LINQ is a 'game changing' technology in data access, I'm now pretty sure about it. And I don't see any real attempts to make a comparative technology from Sun.
    To me, LINQ seems like a throwback to the ancient past with shiny new label. I guess we'll never get away from coupling code to data structures. Around and around we go.
  17. >To me, LINQ seems like a throwback to the ancient past with shiny new label. I guess we'll never get away from coupling code to data structures. Around and around we go.
    It is called "data processing" for a reason :-). And lately, in the Java World, we had a wave of LISP like languages on the JVM where (loosely speaking) code *is* data..... :-)
  18. >To me, LINQ seems like a throwback to the ancient past with shiny new label. I guess we'll never get away from coupling code to data structures. Around and around we go.

    It is called "data processing" for a reason :-). And lately, in the Java World, we had a wave of LISP like languages on the JVM where (loosely speaking) code *is* data..... :-)
    It's just exhausting to see how excited people are to embed queries in their code. I lived through a few years of powerbuilder maintenance and this is not something I would ever recommend to anyone. It's not that LINQ isn't clever. It's that 'integrating queries' into your code is a really bad idea. LINQ is like a really well designed crack pipe. It will seem great at first but I predict the code will become a maintenance nightmare. That's really the MO of MS, though: "forget eating your vegetables, have some more dessert." It's unfortunate that few people in this industry understand that most of the cost of software development is in the maintenance cycle. Most everyone worries about how fast they can build the initial release and not what happens after that.
  19. >To me, LINQ seems like a throwback to the ancient past with shiny new label. I guess we'll never get away from coupling code to data structures. Around and around we go.

    It is called "data processing" for a reason :-). And lately, in the Java World, we had a wave of LISP like languages on the JVM where (loosely speaking) code *is* data..... :-)


    It's just exhausting to see how excited people are to embed queries in their code. I lived through a few years of powerbuilder maintenance and this is not something I would ever recommend to anyone. It's not that LINQ isn't clever. It's that 'integrating queries' into your code is a really bad idea. LINQ is like a really well designed crack pipe. It will seem great at first but I predict the code will become a maintenance nightmare. That's really the MO of MS, though: "forget eating your vegetables, have some more dessert."

    It's unfortunate that few people in this industry understand that most of the cost of software development is in the maintenance cycle. Most everyone worries about how fast they can build the initial release and not what happens after that.
    That seems very self rightious when in fact everyone learns maintenance is 80% of the cost at Uni but after 10-20 years most experienced people dont believe its true any more for most projects and often the additional quality costs FAR more than a simple RAD and rewrite what doesnt work well approach. This has led to a lot of OO methodologies specifying write as quick as you can with no though on reuse/maintenance until the code is used repeatedly. The original stats came mainly from the large projects in the 60-80s , these days small services and many client modules are more often replaced than enhanced ften by a new language. This is especially true for Java script /Ruby / php/ Web pages / VB forms apps etc . Try quoting a Java app against a dev house writing one of those.
  20. re power builder and Sql code , this sound like Logic in the form ( ie Data access) only someone very inexperienced would do that these days and is not good , even most good RAD shops put all their queries ( including LINQ for .NET ) in a shared and easy to maitain data access assembly. Have you ever tried to maintain 1000 Stored Proc with roll out procedures , coordination with DBAs , counting the work and cost the DBAs do ( they are paid more than devs) ? Especialy when you cant do select * ( due to your app not covering the whole DB/table) , even adding a lousy field costs a fortune. Using a single LINQ lib with your queries is cheap and easy to maintain and the latest MS stuff actually encourages View-ViewModel-Model which has a completly independent view , is easy to write and maintain but the learning curve is a bit steep.
  21. It's not that LINQ isn't clever. It's that 'integrating queries' into your code is a really bad idea.
    I can't quite agree. Why would it be a bad idea? Take, for example a technology like JPA, where essentially queries are embedded in your code as well (via annotations and via class design and via mapping declarations). Maintainable code for me is code that people can understand with from itself without consulting too many external files, documents, zero value adding layers, thought twists up and down the call stack.... etc. etc. Or too rephrase: It is a good idea to make things configurable that are configured and not hardwire the rest (hard stuff does not break when thrown against the wall), whereas what I see is that a lot of stuff is configured that never changes whereas a lot of stuff (such as database names...) that should be configurable is hardwired.
  22. Because if your database table structure changes you'll have to hunt through your entire code base to make the changes to fix it.
  23. Because if your database table structure changes you'll have to hunt through your entire code base to make the changes to fix it.
    In all likeliness, once your database table structure changes you will need to do that anyway, since the abstraction usually breaks here, not only on a technical level, but from a business perspective. That said, when you use a mapping technology like, say, JPA, when your database structure change you will essentially need to rewrite all your database access code....
  24. Because if your database table structure changes you'll have to hunt through your entire code base to make the changes to fix it.


    In all likeliness, once your database table structure changes you will need to do that anyway, since the abstraction usually breaks here, not only on a technical level, but from a business perspective.

    That said, when you use a mapping technology like, say, JPA, when your database structure change you will essentially need to rewrite all your database access code....
    First of all, I don't consider JPA to be a very good solution to this. But that aside, it's not the case that you would need to change your code because the schema changes. I have repeatedly built subsystems that process data from a number of different sources (database and otherwise) with markedly different structure. The structure of the data and the meaning of the data are too very different things. If you change the meaning of the data (logical structure), you will need to change the code. But if the meaning is consistent and all that changes is the way it is persisted, there should be no need to change the code. When you build the query into your code, you've conflated the meaning of the data with how it is represented in a database. Note that tools such as LINQ don't create an issue. It just doesn't solve the real problem. The real problem is that people want to use it and are using it to couple code to queries.
  25. Just temember EJB2...
    Oh please, this is just asking to get slapped in the face with ... MS-DOS, Visual Basic, asp ... come one, MS has always played catch up and always will. Remembered, they just figured out an internet search engine would be nifty ...
  26. 2) ASP.NET is _now_ way better than JSF can ever hope to be.
    I certainly won't disagree with your upfront on this, as ASP.NET indeed always seemed a little ahead of JSF in the past. Having not used ASP.NET for quite some time though, I'm genuinely curious on what points exactly ASP.NET is currently way better than JSF 2.0. Would you care to elaborate?
  27. I'm genuinely curious on what points exactly ASP.NET is currently way better than JSF 2.0.

    Would you care to elaborate?
    1) ASP.NET is tightly coupled with VisualStudio. Smart tasks and other features are very helpful. Though I still prefer to write UI manually. 2) Databinding. 3) ASP.NET MVC And generally ASP.NET just feels easier than JSF, but that's subjective. Of course, Wicket or Tapestry in my opinion are better than ASP.NET, but I'm only comparing ASP.NET and JSP.
  28. Alex,
    Wicket or Tapestry in my opinion are better than ASP.NET, but I'm only comparing ASP.NET and JSF.
    I'm afraid you've lost me here. I definitely don't see Wicket or Tapestry as being "better" than JSF 2+Facelets in a tangible way, just different in approach. You might want to take a look at a similar discussion on the most recent item on Java EE 6/GlassFish v3. Cheers, Reza
  29. I think the folk Alex, Does not have a clue what he is talking about. His points that he said about ASP.Net is superior to JSF 2 are laughable and irrelevant. Im fan of Wicket but Wicket is better than JSF 2? cmon, It seems he again does know have a complete clue about JSF2 and facelets and What is goind on in the Java World lately so why bother to write stupid and irrelevant comments. Look at JPA2 and Hibernate, Ibatis, Spring, CDI/weld, Hadoop, Apache Click its an awesome and super easy component web framework, Struts2 and Stripes for MVC(ASP.NET mvc does not get close to this frameworks) and thousand and thousands of projects behind the Java ecosystem. Take a close look at apache Pivot for RIA business apps,We dont need silverlight or if you want eye candy take a look at JavaFX. IDE's we have 3 awesome Intellij, Eclipse and Netbeans. and much more. What .Net offers to a Java programmer? LINQ? I can get Linq functionality with libraries, by the way Linq was prouposed first as a library add-on but the designer opted to include it in the core. What else closures with C#? Java will have it be patience. WPF? Swing and JavaFX are still kicking pretty well. The server side Java have more ecosystem and projects as I said. What really else .Net offers new to a Java developer because Im still dont know what is it.
  30. What .Net offers to a Java programmer? LINQ? I can get Linq functionality with libraries, by the way Linq was prouposed first as a library add-on but the designer opted to include it in the core.
    LINQ _is_ a library. However, it required core language changes in C# (expression tree support, extension methods, anonymous types and some other minor features). For example, NHibernate uses these extensions to create its own typesafe query support which is somewhat different from LINQ. The strongest point of LINQ are typesafe _composable_ queries. I.e. I can write: //Reusable expression var students = from db.p in People where p.age > 18 select p; //We're making a further restriction on data. val goodStudents = from p in students where p.gpa > 3 select new {student=p, gpa=p.gpa}; //Anonymous types! Array.ForEach(goodStudents.ToArray(), c=>Console.WriteLine(c.student, c.gpa)); //Typesafe! and compiler will check that all the types are correct. For example, if class "People" doesn't have a field "age", I'll get an error. And it all can work with collections in RAM, databases, XML and it even can be automatically parallelized - http://en.wikipedia.org/wiki/Parallel_Extensions And you can't get LINQ in Java, sorry. The closest possible thing is QueryDSL which uses codegeneration to emulate features of LINQ (typesafe queries). The above sample in QueryDSL looks like: EBoolean students = new HibernateQuery(sess) .from(QPeople.people) .where(p.age.gt(18)); //Alas, we have to use non-typesafe tuples List<object[]> people=students .where(QPeople.people.gpa.gt(3)) .list(QPeople.people, QPeople.people.gpa); for(Object [] row : people) System.out.println(((People)row[0]), (Double) row[1])); </object[]> Maybe I can write an article about it...
  31. Alex, I am aware of the "embedded" model that LINQ has if this is what you are referring to as a concrete advantage. Personally, I think the Criteria API in JPA 2 is just fine and is pretty much functionally equivalent without binding things too tightly to the JVM. That being said, I am quite impressed by .NET overall and do think it is a very capable development platform, especially compared to where things were a few years ago (as you pointed out). Cheers, Reza
  32. Alex,

    I am aware of the "embedded" model that LINQ has if this is what you are referring to as a concrete advantage. Personally, I think the Criteria API in JPA 2 is just fine and is pretty much functionally equivalent without binding things too tightly to the JVM.
    Criteria API is not typesafe and it uses string identifiers so it's not refactorable. I can automatically refactor classes in .NET and all my LINQ expressions will be refactored automatically. Also, Criteria API of Hibernate is inadequate, it still doesn't support some very required functionality.
  33. Reza is referring to the type safe criteria api in Java EE 1.6. http://www.ibm.com/developerworks/java/library/j-typesafejpa/
  34. Serge,
    Reza is referring to the type safe criteria api in Java EE 6.
    That is indeed what I was referring to (specifically to the JPA 2 meta-model as to the type-safety concerns). One practical reason JPA would likely never go that route is that it is likely to have a chilling effect on competition at the data access layer to tightly couple a specific data access paradigm to the JDK. That's arguably OK for the .NET world because it is so dominated by a single company... Cheers, Reza P.S.: An article comparing the current state of .NET and Java does seem like a very good idea, especially if the research for it was well done and it was written in a balanced fashion. What would be really fantastic would be recommendations from both the perspectives of improving Java and .NET.
  35. Here is a comparison of Querydsl and JPA 2 Criteria API syntax : http://blog.mysema.com/2010/04/querydsl-as-alternative-to-jpa-2.html
  36. I'm afraid you've lost me here. I definitely don't see Wicket or Tapestry as being "better" than JSF 2+Facelets in a tangible way, just different in approach.
    For Wicket: Faster learning curve, richer library, better support for custom component creation, better validation handling and error display. The main problem with JSF is that it is a base technology. Any real world productive JSF project will always be a IceFaces/Richfaces/SomeOtherFaces project....


  37. For Wicket: Faster learning curve, richer library, better support for custom component creation, better validation handling and error display.

    The main problem with JSF is that it is a base technology. Any real world productive JSF project will always be a IceFaces/Richfaces/SomeOtherFaces project....
    And the famous Oracle ADF Face/MyFace/Tinidad... If you look at it, it is the two last remaining JEE server vendors behind JSF, i.e, Oracle and JBoss. Has IBM (the other remaining JEE server vendor) got behind JEE 1.6?
  38. George,
    It is the two last remaining Java EE server vendors
    This is hardly the case if you look carefully. Besides the IBM/Oracle/JBoss offerings, there are excellent Java EE application servers like Geronimo, Resin, JOnAS, JEUS, etc, all of whom are implementing Java EE 6. Here is a good overview: http://agoncal.wordpress.com/2009/12/12/o-application-servers-where-art-thou/. As IBM stated on this thread, they are on path to publishing their road-map for WebSphere and WebSphere CE, likely including OSGi in addition to Java EE 6. Hope it helps, Reza
  39. To be completely fair, I should add SAP NetWeaver here. SAP's implementation is very good and I certainly like the SAP implementation team personally. Cheers, Reza
  40. 1) ASP.NET is tightly coupled with VisualStudio. Smart tasks and other features are very helpful. Though I still prefer to write UI manually.

    2) Databinding.

    3) ASP.NET MVC

    And generally ASP.NET just feels easier than JSF, but that's subjective.

    Of course, Wicket or Tapestry in my opinion are better than ASP.NET, but I'm only comparing ASP.NET and JSP.
    1) I somewhat agree with that. Normally tight coupling is bad, but on the other hand I'm not really fond of e.g. Eclipse, where nothing is a first concern and all attention (read support) has to be shared with 10 other frameworks. Maybe Netbeans fares better, but at least in Eclipse JSF support is sub-par. 2) That's not really elaborating. That's merely pointing into a direction. In what way do ASP.NET controls bind better to data than JSF components do? 3) I have no idea what you mean with that. "ASP.NET MVC" just by naming a term, how does that educate me on the advantages of ASP.NET? I do agree with the others that this discussion, though interesting, is slightly off-topic here. The only relevance is how serious Oracle is in battling Microsoft's ASP.NET. It's no big secret that Larry Ellison isn't exactly a big fan of Microsoft, so I really am curious what Oracle's plans are here.
  41. ... ASP.NET just feels easier than JSF, but that's subjective.

    Of course, Wicket or Tapestry in my opinion are better than ASP.NET, but I'm only comparing ASP.NET and JSP.
    Well, I guess most Java advocates would agree that JSP is a bit primitive, but almost nobody build webapps using just JSP. There are lot's of alternatives, like the ones you mention. Personally I prefer Vaadin but that's just me.
  42. ... ASP.NET just feels easier than JSF, but that's subjective.

    Of course, Wicket or Tapestry in my opinion are better than ASP.NET, but I'm only comparing ASP.NET and JSP.


    Well, I guess most Java advocates would agree that JSP is a bit primitive, but almost nobody build webapps using just JSP. There are lot's of alternatives, like the ones you mention. Personally I prefer Vaadin but that's just me.
    It was a typo. He meant 'JSF', not 'JSP'! You can see that easily from the context.
  43. ASP.NET vs JSF[ Go to top ]

    1) ASP.NET is tightly coupled with VisualStudio. Smart tasks and other features are very helpful. Though I still prefer to write UI manually.

    2) Databinding.

    3) ASP.NET MVC

    And generally ASP.NET just feels easier than JSF, but that's subjective.
    So you are looking for an IDE that makes the UI creation easier and that provides declarative binding of UI components to business services and also has a better MVC architecture. Start with this video: http://www.youtube.com/watch?v=qeAk0TQCMZ4 to understand what Oracle is doing already to make Java easier than .NET in these specific areas.
  44. Alex,
    Would you care to elaborate?
    I would similarly want to know what you think is missing from JPA 2 that could be added. I have not done .NET development for a while, but do keep an eye on things and no serious gaps are apparent to me? Cheers, Reza
  45. OMG, save us from Oracle executives that think Java is alternative to .NET. I just thought .NET is an alternative to Java!
    Given that 'alternative' is a symmetrical term, this seems like a rather strange statement.
  46. Focus on the thread[ Go to top ]

    Hello, I'm always disappointed to see how people instead talking about the main subject of the thread they elaborate some sophisticated answers about some others. Here the main subject is Oracle and the future of java platform: - Oracle position on open source/licensing (JDK) - what is the governance of JCP - What strategy about application servers: Glassfish, Weblogic Oracle application server - UI and Rich client is JavaFX still the solution - javaDB and MySQL (this one is not related to Java platform but open source community) My guess is Oracle is still in thinking process but the market time pressure is here and they have to take a CLEAR decisions soon to dissipate the actual ambiguity. Please do use/create another thread to talk about Java vs .NET. Tarik
  47. Re: Focus on the thread[ Go to top ]

    Tarik,
    My guess is Oracle is still in thinking process but the market time pressure is here and they have to take a CLEAR decisions soon to dissipate the actual ambiguity.
    Actually, I think Oracle has been pretty clear about these issues. That being said, I agree the proof is in the pudding (i.e. actions in the near term other than just words).
    Please do use/create another thread to talk about Java vs .NET.
    You are right - while this is a valuable discussion it is a tangent at best... Cheers, Reza
  48. About Hypervisors[ Go to top ]

    It is a really interesting subject imop, to have a virtual machine in hypervisor. I'm an enterprise java developer, and had alway missed the opportunity to have my applications virtualized. Currently I'm using VMWare plus some sort of debian based linux (I love that packet manager...) but here is a really big (imho) scalability problem: if my app takes - let's guess - 64MB of ram, than theorically I could virtualize that 32 times in a 2GB phisical machine. But with current technology, if I want to install java I need some OS, and decent ram for linux is AT LEAST 256MB so I can virtualize up to 8 virtual machines. I found a trick to spare HD space: use a LiveCD distribution let me recicle the OS space, so I would mount the ISO with the OS and a little virtual disk for the app+data. But the RAM is the real bottleneck. I found that JRockit has similar funcionalities, but was beyond my budget (that is 0), and I know there was an incubating project (but nothing more than an experiment) in Sun... Now I see the subject here in TSS, but could not find any references. Where did Oracle mentioned that? I could not find that anywhere....
  49. JRockit VE[ Go to top ]

    It is a really interesting subject imop, to have a virtual machine in hypervisor.
    ...
    I found that JRockit has similar funcionalities, but was beyond my budget (that is 0), and I know there was an incubating project (but nothing more than an experiment) in Sun...

    Now I see the subject here in TSS, but could not find any references. Where did Oracle mentioned that? I could not find that anywhere....
    The technology you are talking about is called JRockit VE, has been in development for a few years, was presented at last year's Oracle OpenWorld in San Francisco and is being prepared for release. You may be able to find the presentations online (I don't have a link, sorry). The Sun project you are talking about is called GuestVM and is a research project. Regards, Henrik Ståhl Director PM, JRockit Products Speaking for the combined Sun and JRockit JVM teams
  50. Hello everyone, Look, I base my concerns about Oracle from 10 years of experience working as an consultant on large healthcare projects, had to work with Oracle products because of corporate mandates. So, from the Oracle application server (before and after Orion), Oracle installer, Oracle Enterprise Manager, Oracle Database, Portal, SSO, LDAP, HTB, just to name a few, it was absolutely brutal, the quality, performance, installation issues and support just was very difficult to work with and added significant time to projects. Remember the Informix DSA database was superior in quality and performance, actually worked on many Oracle to Informix conversions. The last in-house Oracle developed attempted product, Oracle J2EE Application Server was an incredible failure, yet, customers where actually paying for it. Then Oracle had to run out and buy the “open source” J2EE Orion server just to get by. Anyway, GlassFish, OpenESB, and MySQL are all disruptive to Oracle’s image of being a Fortune 500 industry grade software provider. I understand, corporate IT managers need to pay IBM, Microsoft and Oracle because the relationship is more important than innovative software. No problem with that, but SUN was breaking ground on a new tier of sophisticated uses of software to drive businesses. SUN was implementing open and quality infrastructure software, because IBM, Oracle, BEA and others were not delivering on the Java potential Instead, we’re seeing bloated, proprietary and obscured API’s from these companies, the usual vendor-lock in agenda. Wondering, how long does it take to install and run Fusion, or is it already embedded in a server chip – Oracle is a hardware company now… Have not seen single evidence on how Oracle is going to make bloated, closed source, proprietary, standard facade and vendor-lock software obsolete!
  51. Hello everyone,
    Look, I base my concerns about Oracle from 10 years of experience working as an consultant on large healthcare projects, had to work with Oracle products because of corporate mandates. So, from the Oracle application server (before and after Orion), Oracle installer, Oracle Enterprise Manager, Oracle Database, Portal, SSO, LDAP, HTB, just to name a few, it was absolutely brutal, the quality, performance, installation issues and support just was very difficult to work with and added significant time to projects.

    Remember the Informix DSA database was superior in quality and performance, actually worked on many Oracle to Informix conversions.
    The last in-house Oracle developed attempted product, Oracle J2EE Application Server was an incredible failure, yet, customers where actually paying for it. Then Oracle had to run out and buy the “open source” J2EE Orion server just to get by.

    Anyway, GlassFish, OpenESB, and MySQL are all disruptive to Oracle’s image of being a Fortune 500 industry grade software provider. I understand, corporate IT managers need to pay IBM, Microsoft and Oracle because the relationship is more important than innovative software. No problem with that, but SUN was breaking ground on a new tier of sophisticated uses of software to drive businesses. SUN was implementing open and quality infrastructure software, because IBM, Oracle, BEA and others were not delivering on the Java potential Instead, we’re seeing bloated, proprietary and obscured API’s from these companies, the usual vendor-lock in agenda. Wondering, how long does it take to install and run Fusion, or is it already embedded in a server chip – Oracle is a hardware company now…

    Have not seen single evidence on how Oracle is going to make bloated, closed source, proprietary, standard facade and vendor-lock software obsolete!
    Great insight. thanks. one comment thouigh - SUN wasnt really this innocent bystander. we have paid millions to SUN for their web server( java one web server), their hardware failed many a times in a big insurance giant i was working with and the cost of workign them was no different than cost of a stupid IBM consultant. So on technology side - i think SUN was better thasn most of the bloatware companies but not too different. on management side - SUN had too many managers and too much hierarchy. I remember at one point in 2003 they had close to 200,000 employess - selling what ? just hardware. how can a hardware company sustain 200K employees ? Too many managers was the main cause of SUN setting...it was too late later on to cut costs , they had lost too much money already.