Opinion: Java Community hits out at .NET Progress

Discussions

News: Opinion: Java Community hits out at .NET Progress

  1. Opinion: Java Community hits out at .NET Progress (92 messages)

    Kris Thompson comments on the SDTimes article ".NET Progress Worries Java". Kris hits out at the constant claim that Visual Studio.NET is the best thing since sliced bread, and that J2EE is ugly and complicated. He talks about his experience with VS.NET, and that in his opinion, it is great for small projects, but he misses a real MVC framework.

    The SDTimes analyst hits out at J2EE, and thinks that it needs to make some changes:

    "I believe it must start by scrapping JCP. Java needs two or three active vendors to define a handful of needed improvements and implement them well and quickly.

    Get rid of the hundreds of currently proposed JSRs. J2EE needs to be redefined—J3EE?—with fewer technologies and easier implementation.

    Finally, Java needs to generate true native binaries in addition to bytecodes. These steps will restore the feeling of a tight, effective enterprise platform—the right response to the unproven but steadily gaining competitor."

    View the report: .NET Progress Worries Java

    and read Kris Thompson's thoughts on the report

    Threaded Messages (92)

  2. <kris>OK, VS does kick ass and it does help me pump out a lot of shit code really fast. And the statement that J2EE is complex, well yes, for a good reason. J2EE adds complexity in the form of structure like MVC (what a great concept!) Building on top of such "complexities" allow your application to grow and be maintainable. </kris>

    There is nothing stopping you in developing a MVC styled application in ASP.Net using VS.Net. May be you need to work on your VS.Net on your new Dell Laptop a more.
    IMO, instead of ignoring competitor, a contribution towards how J2EE can be made better is needed. These kind of above comments take J2EE nowhere.
    I think we should stop being J2EE at heart and start being J2EE at brains :)
  3. Sure I could grow my own MVC in .NET or I could attempt to use Interface Building Blocks from MS or try some of their MVC practices that aren't MVC at all or maybe the best solutions is use Maverick.NET (which I like as a J2EE solutions but not sure if it can be simply ported to .NET) so ya, I have options.
  4. Okay.... I thought this subject was a dead subject... that is .NET is dead.

    However, time and time again I read these sdtimes telling me its alive and well.

    I mean this is the second article in 2 weeks from sdtimes talking about .NET besting java.

    Okay lets get the facts straight first:

    "Borland’s JBuilder and IBM’s Eclipse (WebSphere Studio) have changed little since Visual Studio .NET raised the high-water mark."

    What the hell is he talking about??? Since when did the introduction of VS.NET best Eclipse? He's got the story backward, since Eclipse actually came after VS.NET.

    "Eventually, the Mono project will have ported all of the Windows Forms libraries to Linux"

    Yeah right, when hell freezes over. The stuff is so windows specific that it'll never happen!

    "IBM is looking to AspectJ and aspect-oriented programming (AOP) as one way to find new ways to reduce programming complexity. It also is working with Java open-source vendor JBoss on AOP. But this step will not be sufficient.
    "

    So he's saying that AOP is needed because Java programming is too complex?? AOP attempts to solve OOP problems is general and that includes stuff like C#. This is ample evidence that the dude doesn't have a clue.

    "Finally, Java needs to generate true native binaries in addition to bytecodes. These steps will restore the feeling of a tight, effective enterprise platform"

    Huh? Haven't you seen the latest JDK 1.4.2 benchmarks??

    C'mon guys lets quit talking about paid advertisement from Microsoft.
  5. The article does overhype .NET, but I'd be hard pressed to call it 'dead' by any means.

    I think it's a dangerous practice to discount MS as a threat to Java. As long as MS is backing it, .NET won't go away.

    Though to be honest, I have no qualms with Microsoft. I just make my living off Java. I have seriously thought about learning what I can about the .NET Framework, especially since C# syntax isn't all that different from what I do now. I'm just not convinced I can make as good a living off of it yet.
  6. "C'mon guys lets quit talking about paid advertisement from Microsoft."

    I totally agree, I've tried VS.NET and it just isn't nearly as good as Jbuilder 9.

    Refactoring?

    It is like going back to Jbuilder 4.

    VS.NET makes crap coding easy. I don't think giving coders an easy path to make crap code is a good thing. Path of least resistance etc.

    Frameworks, patterns, coding standards and metrics is the only way to do anything less than trivial development. It also makes developers happy that they are doing something that feels ‘right’ not brown and smelly.

    Dot.NOT
  7. Time to pull our heads out of the sand[ Go to top ]

    As much as I see people complain about .NET, it hasn't changed the fact that it has become a serious contender.

    Businesses generally don't care about best practices or design patterns, they do see that an application can be pumped out faster with .NET than Java, and they begin to wonder why they use Java. Not only that, but Java application servers are expensive. Sure, you could use JBoss, but businesses are leery about something being completely free.

    .NET isn't faster to develop with than Java because it's such a better technology suite, but rather I think it's the Java communities obsession with complexity. I'm a big proponent of MVC design, but I'm also practical about it. As I've mentioned to others, the current project I'm doing I looked to do in ASP. Why? Because I've come to realize that things like Struts and JSPs are not good standards. Instead of making life easier, we make it harder. Why?

    The Java community has become fragmented. JSR's are the epitomy of bureaucratic committee management. The article had a lot of good points about what the Java community needs to do.

    We need more frameworks like Tapestry that free developers from design work and put it back into the hands of web designers. Not only that, but it also allows for rapid development once you become familiar with it. My only complaint is that it's not fully 'mature' yet. However, it makes life very simple (once you learn it) and it adheres to MVC.

    However, I think it's too simple, it's too easy to use. I have a feeling that it will be resisted because it doesn't need JSPs, it doesn't obfuscate itself enough. There are not enough layers of complexity.

    The Java community has developed a fascination with complexity that I think has become unhealthy. Just look at all the buzz around JSF, yet another framework to help me add more complexity.

    I see developers talk down about people who "just make it work" and don't take the time to implement design patterns or find a reason to use EJB's. I think the reason .NET is catching on is because it fits the business model of IT departments that care more about deploying an app than how it works.

    I have some real issues with .NET and C#, but those issues are meaningless to a business manager. They don't care about my issues. Ultimately, developers and IT professionals in general are supposed to be solutions providers.

    The attitude that has become prevelant, which is that companies should bow to the whims of IT, should have died off when the tanking economy proved that IT professionals are not as crucial as we always believed we were. We can't have our ideal little world anymore.
  8. Couldn't agree more. Eliminating complexity was how MS beat Sun in the PetStore-lane ... while the resulting app was 90% SQL code in stored procedures, it's still a valid point that you just don't have to implement all known best practices in one application!

    When it comes to web-frameworks, I also agree. It seems like the best-of-breed darwinism really has gone bad, open source programmers of bad frameworks just won't lie down and diedust :-) I don't like to program plain jsp-pages, so frameworks are welcome. By coincidence I tend to like the programming model in Microsoft .NET WebForms with server components and event-listeners. However, Java-frameworks with this approach are quite rare. SOFIA (JADE) and Millstone are the two I know of, and supposedly JSF has some of this functionality as well, I have understood. Tapestry I don't know. The MVC stuff in Struts, and all the additional taglibs are nice, but, well, I just don't want to know anything about all those requests, responses, javascriptlets and stuff that a nice framework can take care of for me.

    When it comes to native code compilers, just go for Websphere and JRockit, or gcc. Dictating that all Java compilers should produce native code is just plain stupid, and specialized products that do are available already, both open source and commercial.
  9. Quite balanced[ Go to top ]

    Jon,

    I like your reply, it appears balanced and objective, now I have a question; if you had the choice, talking objectively, which would you go with from a practical "lets get it to work quickly" approach? J2EE or .NET? And why?

    This is a question that I am keen to answer, but I do not have the knowledge to make a well reasoned judgement.

    Thanks,

    Stuart King
  10. Time to pull our heads out of the sand[ Go to top ]

    Businesses generally don't care about best practices or design patterns, they do see that an application can be pumped out faster with .NET than Java, and they begin to wonder...

    Agreed. Time to market is supreme. Rapid Application Development is what matters when choosing technology for a project that has less than half a million dollars for development. If the development budget is less than $200k, then the project goes furthest when regarded as throw away ^H^H^H^H er, as write only. This sentiment is implied by eXtreme Programing and perhaps also in agile methodology. And it's why Rational Unified Process languishes.
  11. Time to pull our heads out of the sand[ Go to top ]

    Jason said:
    >
    > .NET isn't faster to develop with than Java because it's such a better technology suite, but rather I think it's the Java communities obsession with complexity.

    Jon Said:
    >Couldn't agree more. Eliminating complexity was how MS beat Sun in the >PetStore-lane

    Brain said
    >words to same affect

    Vic says:
    +1 (ie, I agree with your points)

    KISS! PetStore 3 (w/ Struts) on iBatis makes .NET into a joke.
    But that is beside the point.
    .NET is better than J2EE as promoted by Sun (EJB + BluePrints + JSF, this stuff barely works in lab conditions). Do you think will stop positioning J2EE this way? No a chance. They think we are the enemny for saying KISS.

    http://sourceforge.net/mailarchive/message.php?msg_id=6350426

    .V
  12. Admittedly J2EE is unnecessary complex, however like what JPetstore and xpetstore have shown, you don't have to use it!

    The second thing that you have to realize is that the only think VS.NET helps you with is build the web part of the web application. If all you ever do is web applications against a database, then you might as well do the quickest and dirtiest way you can do it.

    However, nothing in Java prevents you from doing quick and dirty apps.

    Go take a look at Oracle JDeveloper 10g or Borland JBuilder X or BEA Weblogic Workshop 8. It's all drag and drop form creation against datasets. It's dirt simple.

    However, when you need to go beyond the web app you need serious equipment, that's when Eclipse and IntelliJ are invaluable.

    This bring up the question, can you do heavy lifting with Visual Studio.NET?

    I doubt it!
  13. Is J2EE complex?[ Go to top ]

    Just for clarification

    I keep hearing how J2EE is complex, but I have to wonder if it's complex or do we make it that way?

    EJB's are pretty bloated, but I consider them an aspect of J2EE. You can do plenty of J2EE development without them. In fact, J2EE seems pretty simple all by itself. It's all the crap people try to load into it that makes it seem so bad.
  14. .NET isn't faster to develop with than Java because it's such a better technology suite, but rather I think it's the Java communities obsession with complexity.


    I'm not all too sure that is true. If you take the time to learn to work with stuff like Xdoclet, Middlegen (ant of course) and you use JDO/Hibernate and the like you will save a lot of time by generating in stead of programming. The complexity can be automated away ever better. To learn this stuff takes some time and it's usually not all too straightforward as something in a Microsoft toolbox would probably be. But are these tools available for Microsoft at all?

    Consider a Hibernate, Struts web application. You can generate your Hibernate descriptors using Xdoclet, generate your database scheme using Middlegen. Or alternatively, generate your JavaBeans from Hibernate descriptors or directly from the database scheme. You can also generate your basic struts code through Middlegen.

    It takes a while to get it all to work, but when it does: who's faster now?

    Having said all that: if more time was devoted to documenting all these wonderful Java things in easy How-To's, a lot of the headache would go away and you don't have to say "if you get it to work, then...", but rather can say "it is faster, period".

    Having a tool like Eclipse has meant a lot in this area, because you don't stumble as much on the obvious problems (classpath issues and the like). Java is now more accessible to the average developer. And that's a spot that has always been occupied to Visual Studio.

    A lot of these articles don't take these new developments into account.

    Marc
  15. I love JDO and Hibernate because they do seem to be going in the other direction. It's not that EJB's are just so gosh darn hard to learn, but writing them and maintaining them is an exercise in tedium. True, XDoclet takes care of most of EJB's headaches. However, I don't think having to learn something else because of EJB's shortcomings is a good solution.

    Not that I think XDoclet has no value. I think it's plenty useful beyond EJB's. Oddly enough, it was a presentation by one of the authors of 'XDoclet in Action' that really sort of put the problems of EJB's in perspective for me. Prior to that I never really thought about it.
  16. resistance is futile[ Go to top ]

    "JSR's are the epitomy of bureaucratic committee management"

    Agree completely. There was not any study or benchmarks that made me choose .NET. It was the talk. It is out of question that I side with people that use such language, such ridiculous over complex theoretical crap (called "Javaspeak") in quasi-religious terms that encourage hyperbole, paranoia and hatred.

    And then they have not a clue what is going on. Frameworks!

    The new framework from MS is called "Windows Sharepoint Services" and is bundled free with the rock solid and fast WinServer2003. Combined with the new XML based Office 2003 and applications like InfoPath they eliminate the better part of custom contract programming. The real programmers are left comfortable writing low-level and non-trivial custom controls. This is a concept that has been tried many timed before in the computer industry - but now it works. Around the corner awaits Longhorn with Avalon, XAML and Indigo.

    Comparing .NET against J2EE without seeing the whole picture is worthless.

    Regards
    Rolf Tollerud
  17. resistance is futile[ Go to top ]

    Haven't you learnt your lesson yet.

    You're a real sucker for punishment.
  18. come on......again?[ Go to top ]

    Looks like "trolling" Australian forums is not enough and you have to come back from time to time.
  19. resistance is futile[ Go to top ]

    as opposed to companies that speak like this;

    "This landmark release, four years in the making, offered a unified development environment and programming model for constructing a range of software solutions. With the recent launch of Visual Studio .NET 2003, customers gained the benefits of enhanced tool and framework functionality, as well as increased performance, security and scalability for building enterprise-critical software.

    As the needs of businesses continue to evolve, Microsoft remains committed to ensuring the success of its customers by providing innovative developer tools that meet these ever-changing requirements."

    ((from ms "developer tools roadmap 2004" product page))

    Please explain what they mean with this bloated cheeseburger-royale marketing gizmoid speech.

    Here I will give you interpretations into plain english as far as my ability allows;

    "Unified development environment and programming model" ... does this mean it's impossible to program without using their tool (that's what it seems to say to me - language and tool one and the same i.e. 'unified').

    "constructing a range of software solutions" ... you can make stuff with it. some of this stuff might be a computer program. the rest are "solutions".

    "customers gained the benefits" ... we made it better and/or different to the last one.

    "enhanced tool and framework functionality" ... tools are enhanced. frameworks, well there's a word for all seasons that means all sorts of things to different people. but, like whatever, rest assured their functions are "enhanced". in what regard, I am unable to tell. but you can do more of the stuff with the "solutions" mentioned above. I think that's what they're saying.


    "increased performance, security and scalability for building enterprise-critical software" ... now here's an interesting one because for some reason it's one thing most J2EE critics concede that Java is better at. but while it deserves some much closer reading - and every product brochure written in the last five years has a similar phrase - I'm not going to bother with it here. Obvisouly more performance, more security, more scalability is critical, that is, very important, to the enterprise (companies), because everyone mentions it. And microsoft is not going to left out of this language bonanza!


    "As the needs of businesses continue to evolve" ... business conditions change over time.


    "Microsoft remains committed to ensuring the success of its customers" ... Microsoft is abandoning the traditional captialistic concern about making money for it's shareholder and instead implementing a humanitarian aid program of making money for YOU.


    "innovative developer tools" ... Supposedly, Microsoft products.

    "meet these ever-changing requirements" ... Microsoft products change too. (This is a new product announcement after all).
  20. J Technologies/terminologies[ Go to top ]

    Java, JVM, Swing, AWT, JFace, J2SDK, J2ME, MIDP, JRE, J2EE, JSR, JCP, JSP, Taglib, Servlet, EJB, CMP, BMP, CMR, CMT, MDB, Session Bean, Stateful Session Bean, Entity Bean, Container, Remote interface, Local interface, JNDI, Deployment descriptor, EJBQL, MVC, JCA, JDBC, JMS, JMX, JAAS, Struts, JDO, AOP, Patterns (Singleton, Abstract Factory, Value Object, DTO, DAO, Command, Business Delegate, Facade,....).
  21. J Technologies/terminologies[ Go to top ]

    I take it from the post you are expressing the many technologies and frameworks that a Java developer can build on you provide customer with

    Scalable, High Proformance,Reliable,Available,Extensible,Maintainable, manageable and secure apps.

    Or just saying it better to build your own pile of poo rather learn some proven patterns.

    Made me launch anyways.
  22. funny article[ Go to top ]

    There is a definate place for all the components,patterns and things.
    In java things will work out naturally and systems are more maintainable even complex ones or small ones.

    C++ ,C# you will have to work on things by ur self and come to a point where the architecture is inflexible not extensible when the system grows.

    APO is not used to solve java problems. You should go and learn object oriented languages and AOP .....

    If you are selling the idea that .Net is faster i will not buy it unless i am writing a Web page to display static HTML
     
    bottom line .net gets messy as the number of tiers increase.



    --whatever
  23. FUD[ Go to top ]

    This is one of the stupidest articles I have ever read. It all starts with the title of the article:

    ".NET Progress Worries Java"

    Since when did Java become a person or corporation? Even if he would have said it worries Sun that would be ok. What corporation doesn't worry about competive products (if you even want to call it that)?

    I also loved this tidbit:

    "What about the concern that .NET is not enterprise-tested? This remains true. But .NET does not need to conquer the enterprise in order to enjoy widespread success."

    Apparently a technology being proven doesn't matter to a .NET fanboi. As long as it looks pretty and you can code to it quickly, what else could possibly matter?

    SDTimes is trash. I can't believe this made it past the editors of the publication.
  24. J Technologies/terminologies[ Go to top ]

    I do not think it is complex!
    I do not think. It is complex!
  25. Java too complex[ Go to top ]

    I do not think it is complex!

    > I do not think. It is complex!

    Yes of course we make things much too complex, apache is a good example;
    example from Struts docu;

    <option name=
    The attribute name of the bean whose properties are consulted when rendering the current value of this input field. If not specified, the bean associated with the form tag we are nested within is utilized. [RT Expr]

    Couldn´t we just replace these 3 lines with the word; "bean"?
  26. J Technologies/terminologies[ Go to top ]

    Java, JVM, Swing, AWT, JFace, J2SDK, J2ME, MIDP, JRE, J2EE, JSR, JCP, JSP, Taglib, Servlet, EJB, CMP, BMP, CMR, CMT, MDB, Session Bean, Stateful Session Bean, Entity Bean, Container, Remote interface, Local interface, JNDI, Deployment descriptor, EJBQL, MVC, JCA, JDBC, JMS, JMX, JAAS, Struts, JDO, AOP, Patterns (Singleton, Abstract Factory, Value Object, DTO, DAO, Command, Business Delegate, Facade,....).

    But do you truly regard the Java system stack as teetering unsustainably? I think the ceiling is nowhere near in sight.
  27. .NET Technologies/terminologies[ Go to top ]

    Java, JVM, Swing, AWT, JFace, J2SDK, J2ME, MIDP, JRE, J2EE, JSR, JCP, JSP, Taglib, Servlet, EJB, CMP, BMP, CMR, CMT, MDB, Session Bean, Stateful Session Bean, Entity Bean, Container, Remote interface, Local interface, JNDI, Deployment descriptor, EJBQL, MVC, JCA, JDBC, JMS, JMX, JAAS, Struts, JDO, AOP, Patterns (Singleton, Abstract Factory, Value Object, DTO, DAO, Command, Business Delegate, Facade,....).



    C#, Visual Basic.NET, VBA, C++, C, J#, CLR, ASP.NET, Web Method, MIDL, ADO.NET, RDS, ADO MD, ADOX, MDAC, MSDE, MTS, COM, COM+, DCOM, ISAPI, Windows Service, Longhorn, XAML, MSMQ, IIS, ACTIVE-X, ATL, MFC, Crystal Reports, DLL, WIN32, Services, Web Control Library, CAB, OLE, Managed Code, Unmanaged Code, JUMP, MSDN, FoxPro, WhidBey, WFC, NGSCB, Win Forms, VBScript, JScript, GUID, CCLID, ClickOnce, WCL, Add-in (Some security standard), Patterns (Martin Fowler has a hole set of patterns applicable to both), etc...


    I can make the list go on and on as well. I can make Microsoft feel just as bloated as Java. Its all marketing. Don't be fooled. WebSphere Studio can create quick code also with a click or 2. So can other Enterpise IDE's. Visual Studio is nice for client side development, but it does not provide anything more than anything else for server side development. I would add more security API's but then again I have to patch my windows machine every other day because of a new security hole. I left out any Web Service technology. We can go back and forth with acronyms, this is not meant to start a tech list thread.
  28. Visual Studio .NET Mature?[ Go to top ]

    Funnily enough, VS .NET gets so much credit for being the power behind any .NET development edge, real or imagined. The thing that's funny about it is that VS really isn't THAT mature of an environment for end-to-end development projects.

    A lot of people forget that before the .NET edition, Visual Studio was really only geared toward writing ASP pages; and component development was done in a separate IDE for either C++ or Visual Basic. MS has been moving toward the common IDE for several years with a similar look and feel (VB was actually the one that introduced the current MS IDE paradigm, believe it or not), but they were not truly integrated until VS .NET Beta 1.

    For my money, as someone that has actually used it quite a bit in the past, it's kind of a bloated tool. It's got a lot of nice stuff for actual sophisticated, non-trivial development, but it's also got a lot of crap for quickie POC and RAD projects that I could live without. Its ability to do many very simple things have its opponents often completely disregarding the more sophisticated things it can do (I personally think MS should make more segregated releases of VS .NET targeting different scales of development communities, but somewhere I imagine Bill Gates could personally give a sh*t :)

    I thought for sure Borland was going to come along and be my White Knight with C# Builder, creating a toolset for C# with the same sophistication it has shown for other platforms. I was, needless to say, underwhelmed by Version 1.0.
  29. J Technologies/terminologies[ Go to top ]

    Patterns


    Patterns are an OO concept, not a Java technology.
  30. J Technologies/terminologies[ Go to top ]

    I couldn't agree with you more. I'm an application architect for a very large US bank. All of our teams (5 of them) have been forced to work with .NET and totally abandend J2EE because of cost and time to market.

    I totally hated it in the beginning, but now after a year...WOW it totally rocks!!!

    Try it before you knock it. I wouldn't go back unless I had to...
  31. <WithoutReadingAWord>

    Oh thank goodness, another .NET versus J2EE spat. It's been a few days.

    :-)

    Must...resist...tasteless jokes...about flamebait originating from San Diego...

    </WithoutReadingAWord>

    Now I'll go read for deeper insight into this matter.
  32. A very witty reply[ Go to top ]

    Mike,

    Which would you choose if you were being totally objective (If such a thing exists)?

    I am also confused by endless argument, my feeling is that either will and can work and either can produce rubbish, but beneath all of the diatribe is there actually a real advantage in either of them?

    Has anyone actually done any realistic, unbiased research into teh area of J2EE vs .NET on the cost/ease of development fronts? If so, do you know where I can find it?


    Stuart King
  33. My reply[ Go to top ]

    For me, there was no choice. I develop for a company that has invested in mainframes, rs/6000, linux on z/series, linux on intel, and windows. .NET isn't even a consideration here. I'm not biased at .NET except for that I dislike Microsoft as a company. Take that for what you will. I am biased against VB and VB developers, though. I use personal experience as my basis as I've only met one out of many of them that really understood programming concepts. (part of my job is porting our vb apps to java, and some are really blah)

    In sum, I've seen first hand some really really bad systems designed by programmers that wouldn't be programmers if it weren't for VB.
  34. .NET still a long way to go[ Go to top ]

    The last time I tried to install .NET it took more than 6 hours to do so (I had to apply many fixes before .NET installed). Then, I tried to make some COM calls using C#, assuming that Microsoft technologies talk to each other seamlessly. Guess what? They don't. It was easier to create java stubs using JIntegra and make the COM calls through Java.

    Micorsoft, has a great product in MS Office and a decent OS. But, everything else is either hype or vapor ware. There development IDE's lag far behind Java IDE's like JBuilder, IntelliJ and Eclipse (This could be because there is intense competition between Java vendors :)).

    The JCP process has its problems. But it is still much better than the "know nothing and non collabrative" approach that Microsoft uses.Bottom line is that, if you use Microsoft technologies you have to depend on them totally. J2EE provides competivive choices and there is no vendor lock in.
  35. ...but not really for the reasons that the author states. The enterprise IT game is all about mindshare and whose ideas are getting the most time on the playing field. There have been many different leaders at any one time, depending on which aspect you look at (massively scaleable systems, lowest TCO, open versus closed standards, whatever). Any new viable entrant into a race, especially one such as enterprise application development, raises the hackles of the de facto market leader.

    This board is a prime example of the idea that there's a worry (or at least severe preoccupation) with .NET in the J2EE world: the most heavily trafficked threads are almost always those that directly talk about .NET (this one will be no different, guaranteed) and .NET regularly brought up in many technical discussions of Java-specific technology.

    And for those of you who can barely control yourselves long enough from hitting "Reply" and blasting this with the inevitable ".NET isn't being used in any enterprise applications" stammer: unless you've surveyed every company in the world and seen no implementations out there, don't even bother. They're out there; I did not say that there were even as many as Java since that would be inaccurate, but it's not non-existent.
  36. Okay I confess....

    .NET isn't non-existent.

    However it's dying... well take this result from a google search as an example:

    Why C# is better than Java

    It's one of the top ranked entries in google...

    guess what.... it's an UNEMPLOYED IT message board!

    Need more evidence, just head over to monster and search for C# and do the same with Java.
  37. Rolf? Is that you again?

    ;-)

    I'm fooling with you of course, since Rolf's passion was advocating Monster searches in SUPPORT of .NET viability, not against it.

    And the reference to that unemployed job board proves what exactly? A simple search on the word "Java" produced three pages of found threads ("SUN is our mortal enemy" by "gloomyshoes" was my favorite). GASP! ".NET" search only returned 2 pages of threads! Definitive proof that .NET is in fact beating Java. <Smirk> Please.

    Carlos, Carlos, this smacks of the same quantitative reasoning that you put into your "100 reasons why .NET sucks" montage a while back. I don't seem to recall that one standing up to intense scrutiny, either. In fact, I seem to remember having some fun with that one too.

    :-)
  38. Mike,

    Well maybe that "intense scrutiny" that you mention is a figment of your imagination.

    Afterall, can you provide a link showing that it's real.

    If you can't find one, then please provided your own scrutiny, it's 100 reasons so you should have plenty of targets to hit.
  39. Carlos,

    http://www.theserverside.com/home/thread.jsp?thread_id=18411#77550

    This is mine, for your reading pleasure. It's long. I seem to recall reading some blog responses to the list written in the same light which is all well and good; everyone's got a right to an opinion and the list itself is nothing offensive (it's just IT for goodness' sake). But I abhor faulty reasoning and research methods passed off as intelligent discourse.

    :-)
  40. Mike,

    Well thanks for the link, I've never seen this one. However, why only 25, the list has 100 entries, is that all you can really muster? But granted there's some logic in your "sampling" methodology (you know as well as I that it's a strawman argument with a spin towards probability ).

    Anyway, let's rexamine your rebuttals:

    Only 2 out of the 25 rebuttals have a strong argument the rest, well the rest are mostly like "I don't believe you because I need a bigger sampling size".
    C'mon give me a break you are right there contradicting yourself.

    Anyway let's address the remaining 2 strong arguments "Compile To Machine Code" and "Cross Platform Integration with other languages". The first one is a bit of a gray one, and maybe I'll take it out of the list, however its more about "cross program optimization" rather than building machine code, afterall both JVM (via JIT) and CLR (via precompiling) build machine code. The second refers to cross platform *tight* integration via (JNI), web services is one way but needs to cross process boundaries.
  41. Carlos,

    Just jotting down a quick response in the interest of time:

    If you read the surrounding text to my rebuttals then you saw that those 25 were just the first ones I happened to click on. My point was that I had a lot to refute without having to search very hard. I can't spend ALL my time breaking down platform arguments (school, work, family, life, etc). As I stated there: "...neither the time nor the inclination...".

    I really don't think those are strawman arguments at all, but I typically see that response when there is no stronger one to make. Anyway, it's a moot point because as long as I say my points are valid, you will say they're crap. We can each go back to our respective corners smug in the affirmation that we are masters of our domains.

    Several of the arguments I bothered to make against your list (from what I remember, this was months ago) were that they contained broad statements that when examined more closely didn't really say what you were representing them as ("future proof", "government approved", "Most popular language in the corporation") or were opinion rather than fact ("larger talent pool", "higher paid", "better support for best practices"); the use of opinions is fair since the list itself was an op-ed type piece, but opinions are born to be disagreed with. Surely a community such as J2EE predicated on the concept of freedom has no problem with disagreement?

    Getting back to the original point, of course the march of .NET worries the J2EE community. Threads like this and lists like yours are obvious examples of this. There's no shame in that; it's the smart thing to ensure evolution and J2EE needs it (maybe not evolution, but rather a tighter focus on what's successful instead of fragmenting the platform with every derivation and one-off of innovation imaginable). Guaranteed, in a few years .NET or whatever name it takes at that time will probably need it too.

    I've enjoyed the chat. It's a philosophical debate in the end, one that won't be resolved by FUD, claims or counter-claims. Thus ends my bi-annual spleen-venting on .NET vs. J2EE, and back to worrying about the state of the industry with jobs being shipped to hacks getting paid $2 a day (unofficial offshore rate :) or whatever to chunk out crappy code that you and I and others have to manage.

    Mike
  42. Actually, your rebuttal was only 15, you started counting from 11.

    "gov't approved" is based on OMB assessment were java out scored .NET
    "most popular language" is based on survey mentioned
    "larger talent pool" is based on the fact that Java is predominantly taught at universities
    "higher paid" is based on a search a search on hotjobs and numerous surveys
    "better support for best practices" test driven development, continuous builds, refactoring all common practices practiced predominantly in java environments

    your rebuttals were more in line with "I need more data points to make that conclusion" rather than providing a counter argument and evidence to back that up.

    While you worry about offshore rates you might want to read "Is .NET a gateway to offshore programming". See its more than technical merits of a platform, its all about your livelihood!
  43. It was 15, that's correct; I mis-typed the number "25". :-) That unfortunately is the last place that we agree on this one.

    The SD Times survey doesn't really paint Java as the "most popular language" from the way I read it; maybe my reading comprehension is off but it sure didn't seem that way. In fact, Java was the winner by default due to a split between VB .NET and C#. The hidden factor: both languages are functionally the same with the .NET Framework BCL and CLR at their cores. Heck, if I wanted to I could make a bid deal about the pie graph stating that XML and Web Services are compelling the fewest number of respondents to standardize exclusively on J2EE (so much for clear superiority). I won't though because that one piece of information doesn't say anything very relevant, but I think my point is made.

    The other points I made work along similar lines: it's not that you don't have enough data points; I'm not a statistician. It's that you either misrepresent articles/surveys or HAVE no data points (HotJobs searches are NOT credible employment/compensation sources). Again, I'm just disagreeing, but that's okay.

    BTW - Referring me to your personal blog where you spend an inordinate amount of time bemoaning .NET, C# and all things Microsoft isn't really the way to lead me to the truth on any issue.

    ;-)
  44. Mike,

    Well you know, these Java vs .NET debates are always good for ratings.

    Just ask theServerSide or SDTimes. Got to have the obligatory debate from time to time.

    See you in 6 months when the debate rages again.

    Now back to regular programming!
  45. Haha...yes, on that we definitely agree. I thought TV was the only outlet that was a glutton for ratings, but this issue is definitely used extensively (especially by SD Times).

    Back to it then. Take care.

    :-)
  46. C# != .NET

    I've never seen a lot of jobs for C#, however there are still plenty of VB programmers out there.

    There are also plenty of unemployed Java developers as well. I'm not exactly seeing your point here.
  47. Jason,

    Well, uh, maybe you're right, not a lot of C# jobs out there, mostly VB programming.

    So how about trying a monster search for VB.NET, afterall VB6 isn't the same thing right?

    Carlos
  48. From talking to my colleagues who are doing VB, it seems most places are making the move from VB6 to VB.NET, and they expect their programmers to upgrade their skillsets. Though it's not a drastic change from what I understand.

    The fact is .NET is still relatively new, and already it's made an impact on the marketplace. It's far from dominant and may never be. At the same time, it's not going to go away.

    The presence of .NET should be a rallying cry to the Java community to improve their processes, not have us hide in a hole and cover our ears.
  49. Oh thank goodness. As long as they upgrade to anything else, it'll all end out okay. Pre-.NET Framework VB is awful. The lobby on that one from the established industry gurus though made even FoxPro proponents blush (and those guys are like the cockroach, never going away). The VB .NET language could have been much better when .NET was released from beta, but the old gurus managed to bitch and whine until their backward compatibility was preserved and intended features were removed.
  50. "I believe it must start by scrapping JCP. Java needs two or three active vendors to define a handful of needed improvements and implement them well and quickly.


    I haven't read the article, but I do agree that the JCP process has produced some really undesirable things, like EJB 1.0, EJB 1.1 and two technologies with fantastic overlap: JDO and EJB/CMP. And there are lots of other problems. I don't know if I am the only one to feel this way, but handling XML is way too cumbersome in java.

    On the other JCP as a program is alright overall, and should not be scrapped. But they should find a way to make it faster, more focused, more practical. I don't know if you can do it, but it would be nice.
  51. The real reason that .NET is gaining ground is not because architectually its better,it is because it is simpler to develop because of the development tool.

    This shows the importance of development tool to the acceptance of a technology. The thing to realize is that what will make Java tools answer such a need is not just another refactoring functionality or a helpful code editor (like you find in the eclipse/inteliJ tools) but rather a visual way of development and frameworks that minimize the coding that you need to do.

    I think the SDTimes guy should have a look at Oracle JDeveloper 10g and see that Java too can be very easy to develop. Using an MVC approach you just define your EJB (visually with UML), Define your Controller (Visually with a struts page flow), define your JSP View (Visually - with drag and drop of the EJB Model onto the page). And the resulting code is pure Java code that will run on any J2EE engine.

    So his claim that the Java IDEs haven't evolved to simplify J2EE is not exactly correct - its just a question of where you are looking.
  52. java in the enterprise[ Go to top ]

    Just my opinion,
      
      Since the release of .NET a few years back, there has been so much news about how great it is, how better it is than java in my company. I sort of feel threatened because it came from one of our CTOs and im a java guy. But until now, our system has not one system running in .NET. Im in a telco company and one of the biggest companies in my country and guess what, we have EMPOWER, MSP, SEEBEYOND (going to version 5), ACTION ENGINE,etc which are all java and WE ARE RUNNING LINUX and SOLARIS considering our CTO wants .NET. How the hell should you run dot net in these environment. Aside from that APACHE projects are very useful and really kick ass.

    Reality bites for the .NET lobbyists. Sorry :)
  53. Where is the proof?[ Go to top ]

    dot-Net has been out for long enough for real big projects to be implemented. But until now, I haven't seen such complex and constantly systems like "eBay", "Amazon" or banking-applications. All I've seen are articles an discussions about what it could do and why it is better than Java and all the rest.
    As long as there's no proof that dough-Net can do what sales-people and some unknowing managers are saying, I'll stay portable...
  54. The day I can install and use MS Office built with .net I will worry about learning .Net, but for a while I'm happy with StarOffice built with Java
  55. It's all about ide's[ Go to top ]

    I think the only thing superior in .NET is VisualStudio. It's not that its better or worse for refactoring or aplying patterns, or creating uml from code. It's just that you can create an asp.Net page or a graphical user interface in 1 minute. The back side (helper classes, dto's, dao's etc) can be exactly as complex or as simple as a Java app can be.

    The true lack in de Jworld is an ide you can use to build an app really fast without spending the hole day looking for frameworks (wich is quite fun but not much productive). Eclipse is cool, but it doesn't really have jsp support (do you know of any plugin?). SWING is not practical for RAD and in SWT you have to do things by hand.
  56. It's all about ide's[ Go to top ]

    "It's just that you can create an asp.Net page or a graphical user interface in 1 minute. "

    Dude, you haven't taken a look at Oracle JDeveloper 10g , Borland JBuilder X or even BEA Workshop 8. Even WSAD has something like this but I haven't looked personally. Actually, VisualAge back in the 90's had a visual builder for web.

    However, maybe you forget, Java developers don't spend most of their time building screens. They're doing heavy lifting elsewhere, and that requires heavy lifting tools.
  57. Microsoft is doing very well[ Go to top ]

    microsoft main struggle now is with the open source community not with java and if we consider java to one of open source tech. then microsoft is fighting with it what really bugs me is that developers thinks only abuot cooding and developing software but forget about the whole it vesion of the whole orgnization that where microsoft is going to make a leap.taks as an example small business server you got almost every thing a small business need from infrastructure to dbms mail server network file swerver ...etc for a 400 $ or 1000 $ if you wnt more cababilities and many busineess applications come on top of that from MS of course with low cosr(CRM,ERP)-compared to the other solutions that way microsoft can beat ,on the other hand all of the developers are still disscussing which ide is better of courseVS .net is much better that all of then and please do not forget about product integration issues for ihgh end busineeses .and last remember %80 of businesees are small or medium busineeses.
    until we can compete this microsoft will the software giant ,
  58. Microsoft is doing very well[ Go to top ]

    If you want low cost, maybe I should show you the alternatives:

    (1) Start with a Linux distribution (note: No limits on amount of users unlike
    windows server 2003)

    (2) Grab MySQL, PostgresSQL or Firebird RDBMS.

    (3) Grab an appserver like JBoss or Jonas. Alternatively grab a webserver like Tomcat or Jetty.

    (4) Grab any of these applications that run on top of the above 3.

    Workflow

    EAI

    Portal Server

    Knowledge Management

    Total cost: $0.00

    Now if you can't build a solution for your customer with the above tools then you shouldn't be in the IT business.
  59. Please, can you stop make people believe that open source cost 0.
    Have you ever tried to hire a JBoss/linux guru to configure cryptic XML to set up a cluster?
    Have you ever heard about maintenance costs, total cost of ownership?
    We should use a more careful language rather than always saying that open source cost nothing. License is just a part of the picture.

    Respect,
    Stephane
  60. I could'nt agree more that what exactly I am talking about it is not about being free it is about the TCO and pepole always forgets about this subject
  61. Open source is not free of charge.[ Go to top ]

    So your saying after someone putting together a small business application using an open source stack the TCO of ownership will be greater than a windows based stack even after getting the major components for free?

    So that means either administration costs more or development costs more or both.

    From the administration point of view as evidence check out netcraft that shows apache own 2/3rds of the market and gaining even more momentum. People are beginning to realize the difficulting in administering microsoft.

    So what's left? Development costs? Well okay lets say its more.

    Well honestly I don't have a problem with that, I make a living doing that and heck I'm glad its more. If you want less then hey you can always drive down the cost even more by sending offshore. The environment is so darn easy that supposedly a fresh graduate can get up and running pretty fast. Heck, there are a lot more fresh graduates in foreign countries than in the U.S. More supply drives down costs, simple economics.
  62. Don't be extreme. We are not talking about off-shore...
    I just want to make the point that putting a list with all the open-source stuff and a great EQUALS 0 at the end is a dangerous comment.
    Simple economics.
  63. Stephane,

    The original posted started talking about how cheap Microsoft solutions were and started quoting prices.

    I simply followed and compared apples with apples.

    No all of a sudden you want to change the subject matter?

    Open source is priced lower than windows solutions and that's a fact of life.
    Now if you want to talk about TCO then that's a completely different thing that includes paychecks in the equation. When you start talking about reducing paycheck you might as well start considering offshoring.

    Carlos
  64. Microsoft is doing very well[ Go to top ]

    BULLSHIT
  65. I just hate MS[ Go to top ]

    I just hate MS. no reason.
  66. I just hate MS[ Go to top ]

    There are lots of idiots like you. No reason
  67. I just hate MS[ Go to top ]

    "Java developers don't spend most of their time building screens. They're doing heavy lifting elsewhere, and that requires heavy lifting tools."

    You don't have single idea what Java/Unix/Opensource technologies do, you dummy screen drawer.
  68. Is this kindergarten?[ Go to top ]

    C'mon guys. Any chance of keeping this professional?
  69. Is this kindergarten?[ Go to top ]

    I agree. "dummy screen drawer"? This is what we resort to when challenged? I spend valuable time crafting intelligent, if not wholly acceptable rebuttals and responses to threads to gain real estate next to someone who uses this gem?

    Argh. What does this even mean? If you read any of his posts you'd realize this doesn't even make any sense, unless my understanding of "dummy screen drawer" is different from the generally accepted meaning (whatever that is).

    ;-)
  70. .NET vs JSR[ Go to top ]

    "Java developers don't spend most of their time building screens."

    >
    > You don't have single idea what Java/Unix/Opensource technologies do

    Any developer, including Java developer, spend a lot of time coding CSS, XHTML, JavaScript, ActionScript, etc.; time devoted specificaly working on GUI screens, so they can have rich user interface (like MTV.com or CocaCola); using every available trick on a browser.

    Time spend on doing EJB's is for newbies, or people that want to justify C#.

    Worst part will be not this study, but when JSF realy embrases J2EE developers realtive to .NET; like I would use JSF, but managers will assume... oh JSF + EJB, we know that sucks. Yes it does! But I do not do that for rich UI!!!!
    But if in C# I am not expected to use EJB, GREAT!!! So I 50% agree with .NET marketing. J2EE out of JSR is ka ka. MS listens to uses, Sun never does; I hope IBM takes over Java ... soon. This smells.
    Give me a Choice!!!
    .V
  71. It's all about ide's[ Go to top ]

    "SWING is not practical for RAD and in SWT you have to do things by hand."

    http://www.swt-designer.com/
  72. Please explain this to me[ Go to top ]

    .Net is not going anywhere, it's already good and it will get better with a speed that java won't match. MS has ideas and will implement them as fast as they can. In Longhorn they will have managed services and a managed windows api. Just read about Indigo, that is the next step.
     I know, the idea to put the jvm in the OS is old, but that's nothing. The idea is nothing. WHO will DO something like this in the java world? Will Sun make a jsr and “the open source community” will implement it? For what OS?
    I have nothing against Sun and or Java, but I do get very annoyed every time people just don't see it. Don't mix emotions with business, guys. I just don't see it how do you imagine that the competition will keep up with MS. Please, please someone tell me how does he/she imagine that Sun and/or Oracle or I don't know who will keep up with the billions MS will put into .Net ? Gates will hire the best visionaries, the best programmers, as many as he needs, and he build something without worrying to satisfy many OS's and so on.
    Will you please tell me how do you imagine the bearded guy in the open source community to compete with this? Or even Sun, which has money in the bank 30% less than MS is investing in R&D yearly? Am I missing something or is everyone delusional ?
    Where is the common sense?
  73. Please explain this to me[ Go to top ]

    "Or even Sun, which has money in the bank 30% less than MS is investing in R&D yearly? Am I missing something or is everyone delusional ?"

    So explain this to me, with MSFT having billions of dollars in the bank, why can't they release a 64-bit version of .NET?

    Sun, who probably makes less money than the entire MSFT R&D budget has Itanium support since JDK 1.4.1 and will have Opteron with JDK 1.5 . Having more money may make things happen, however it doesn't guarantee success, just ask all those VCs who put money on dot coms.
  74. You're right of course, but...[ Go to top ]

    You're right of course. MS has enough money to go on a ten year wild goose chase and still come back and mop the floor with their competitors. But in fifteen or twenty years software will be written by artificial intelligence software and java is more well suited for ai than any other software I've used.
  75. Just look at Linux. How did Linux surpass Windows on the server so quickly? They spent TWO ORDERS OF MAGNITUDE LESS TIME AND MONEY.

    The reason is that Microsoft is monolithic. They can only go in one or two directions at any one time, no matter how much money they spend. The worlds of Linux and Java are able to go in millions of directions simultaneously. Of course most of these directions are abject failures, but it is the successful projects that define the future direction of software. Microsoft can't compete with this. They can destroy any other monolithic organization, Sun included, but they can't compete with the collaborative effects of the open source and Java communities.
  76. Just look at Linux. How did Linux surpass Windows on the server so quickly? They spent TWO ORDERS OF MAGNITUDE LESS TIME AND MONEY. <

    What are you talking about??? MS owns 49% of the server market.
    Linux only owns 29%.

    True they are catching up, but it has slowed down quite a bit. Companies are starting to see that Linux is not so free as many people claim it to be.
  77. Java : The horse that became a Camel[ Go to top ]

    It's all political...

    - IBM is so hellbound on killing Sun and vicae versa as are all the other vendors that nothing will ever get done on Java and the JCP. That's why Jonathan Schwartz's argument that Open Standards are more important than Open source holds no water.

    - The vendors made J2EE complex so that they could make lots of money selling application servers and services.

    - Before .Net an IDE for Java was never an issue, that's because most Java programmers worked on Unix and the Unix community considers ease of use a non-issue. CTOs crave ease in IT, Microsoft understands this, Apple understands this, Digital understood this with VAX, Even the Open Source community is starting to understand this. Sun doesn't understand this. IBM understands it but that would undercut its services revenue.

    Java has become nothing more than a Political quagmire. James Gosling's elegant beast has become a cesspit of political bickering which is turning the platform into a Camel in much the same manner as what happened to Unix in the early nineties.

    We need a Java and .Net alternative, something that is free of Java's baggage and Microsoft's control we need to create a development platform that does for development what Linux did for Operating Systems.
  78. .NET NOT YET[ Go to top ]

    as the Sun says

    .NET NOT YET
  79. How VS. NET fails[ Go to top ]

    OK, here's my experience with VS.NET. Our app is a .NET GUI which send XML to a back end Java app. I'm doing the Java app, but needed to spend some time fixing bugs in the GUI.

    After a day to install VS.NET, I fetched the code from Source Safe to a new directory on my computer. Then I tried to build. Unfortunately, the original developer had a different root path to the source code, so none of the Solution or project files could find the components that they needed. I had to (with much help from the GUI developer), relink all of the subprojects of the GUI application together in order to build.

    That is pure crap. To do Java work, all that is needed is to fetch source into a class pathed directory structure, mount that source directory in Net Beans and compile.

    You can't convince me that VS .NET is a better IDE. And that was before I found out that when our GUI developer had to update a third party GUI component that VS .NET eraseed half of the GUI layouts and saved the files without so much as a popup warning box saying that it was hosing the application.
  80. It is true[ Go to top ]

    I had exactly same problems when I worked for a .NET project.

    I even asked a Program Manager of Visual Studio .NET about this issue and unfortunately, she could not answer it. I do not understand why Microsoft did not adopt this feature since they alreay copied so many stuff from Java.
  81. Intellij IDEA[ Go to top ]

    The fact that the author fails to mention Intellij IDEA among Java development Tools invalidates it all. He abviously does not have real knowledge of current Java development.

    IDEA is light years ahead of both JBuilder and Eclipse, both of which are still trying to catch up. And even the community built jEdit is better than Eclipse.
  82. Another pro .NET article get trashed[ Go to top ]

    Okay here's another article worth trashing:

    http://www.sys-con.com/dotnet/article.cfm?id=214

    Some quotable quotes:

    "even many of Microsoft's oldest and most dedicated developers - the Visual Basic programmers - have refused to simply "go along with the program" when asked to change their entire programming approach to match the new requirements of VB.NET. These resistant VB6 developers - the ".NOT" movement, as they have come to be known "

    Anyway, the gist of the piece is that the guy continues to look for clues of any life in the .NET platform. Okay, I have to admit .NET isn't dead, however it's as comatose as DCOM and previous technologies that MSFT has burdened the programmer with.
  83. Since when did "hits out at" become an accepted verb?
  84. sdtimes article is rather a compliment[ Go to top ]

    Expect for the title of the article i feel the articles does speak good things about J2EE rather than .net.After 3 years of inventing .net(or is it copying J2EE) ,if they say they are gaining ground in some segment of the market..it speaks voumes of the .not ability.So guys let's not be worried too much of this revolutioanry(revolutioanry because first time so many billions spent on it) technology even if it captures some 10 ..20 % fo the market...we will always have 80 % of this ever growing market.
  85. ????????[ Go to top ]

    mr Kris Thompson,
    Please take some classes on the (java)subject,
    write some (java)code, and then write articles.
    Excuse me, but you are funny and your articles
    absurd !
  86. Java is too... easy.[ Go to top ]

    In the last recent years I have reached the following conclusion:

    Java is too easy ... for anyone to jump in, take a look around and convince himself/herself that he/she rocks.

    In my view, this has lead to a lot of people poorly skilled as programmers easily starting up with java. Actually producing results, results that worked (at that time). As it is in the business - results matter. So you gain weight (political), you get authority (manager's trust), you gain [sometimes unbased] self-confidence....

    Unfortunately the luck of knowledge (those 20+ years of enterprise applications development sometimes being mentioned by real gurus) eventually kicks in. And then you have a huge crowd of people who on one hand are kind of doing java but on the other hand also complain and moan about it all the time, people who are ready to label anything as "too complex" and literally break a perfectly working application just in order to "make it simpler".

    And I think there is nothing more worrying for a manager then a phrase from some developer: "the application has become too complex [not because of me] and I don't understand anything anymore".

    Please understand, I am not opposed to "making it simpler", but so far for me it always has come in the worst possible times of the application lifecycle (for example, right on the start of the full scale roll-out to multiple clients)

    Maybe it would sound stupid, but today being a Java developer, with all the skills I have (and with the knowledge about existance of a lot of things I still DON'T know but willing to learn), I don't feel myself as a part of "elite".... much more often as a person stuck in the middle of a mindless crowd.

    Java wins hearts easily, I think it is time for it also to start winning brains.

    As a side thought, maybe someone (Sun?) besides popularising Java in the current industry management should start thinking about some world scale effort of educating those who are just joining the industry for now being just a "brute" force with no experience?

    Ed Letifov
  87. Cocky Developers[ Go to top ]

    I completely agree Eduard - and find myself often feeling as you do. How many people claim to "know" the EJB 2.x spec? How many of them have set up, debugged, and load tested complex CMP entity bean transactions in clustered environments? How many have tweaked the code to best take advantage of their app server's caching mechanisms because they encountered scalability bottlnecks?

    No, I'm not trying to start another EJB vs JDO vs Hibernate vs Distributed Caching thread.

    Software developers swagger a bit too much for our own good. Most of us are working on the $200K - $5 million dollar apps that the Andrew Binstock article spoke of. Most of us have never worked on an application that is required to scale to the level that the $5+ million "enterprise caliber" applications require.

    Without that sense of scale, most of us are just babbling idiots who cannot put technologies, specifications, and development tools into the appropriate context. The statement "yeah, but who needs A when you can do B so much faster!" always raises red flags for me. Sometimes A legitimately sucks, and B is a no brainer. Sometimes A is more complex because that complexity supports a scale that the guy promoting B cannot comprehend. Sometimes the guy comparing A and B is truly an idiot and doesn't realize he's comparing apples and oranges.

    There was a lot of all three in this article.
  88. All in All[ Go to top ]

    At the end of the day (to coin a better phase), its about implementing the best solution to a problem - not about which is the best language.

    Java is a great, provides plenty of features and currently IMOP is the best lanaguage to develop powerful, component based applications. But this is mostly where all my experience lies so I would say that.

    Next year this might change, remember Java has only been around seriouly since 1995, and as developers we should be open to all technologies in-order to continuley earn a living.

    No-one likes change, but that's life.

    /J
  89. All in All[ Go to top ]

    Amen
  90. Hrmmm[ Go to top ]

    I have a buddy who is a .NET developer and he is trying very hard to jump ship to Javaland. It kind of took me back as I would have never expected him to say that to me since he had been partial to MS all of his life. Basically, the company that he is employed by and Microsoft's tools that make crap code easy and fast to generate has caused a swarm of unintelligent people to be able to become programmers and has resulted in him hating his job. (He uses C# whereas his co-workers only know VB.NET)

    Are we arguing that allowing tech school level students to be able to do our job (poorly) is a good thing? It's one thing to have tools that eliminate grunt work, but when you have tools that want to hold your hand and show you how to program, I think you're asking for trouble.

    On another note, I do agree with having the capability to compile to native binaries. I see alot more markets in store for Java if it had this capability. Computer games for one, even Operating Systems.
  91. whenever you can get paid more...[ Go to top ]

    Answer this question from you heart:

    Job A paid $100K but you need to learn C#. Job B paid $70K you need to learn Java or vice versa. Which one you pick? Majority will pick the higher paid? Why? Code for food, rent, bill, morgage or etc.
  92. whenever you can get paid more...[ Go to top ]

    Haven't heard of C# job that paid $100K, however hear it all the time in Java land.

    The whole point of the MS paradigm is that you can find an average joe from a tech. school to do the job for a ridiculously low price. It caters to the drag-and-drop crowd who build nothing but form based interfaces. Beyond that its a pain to work with. Now, incidentally, that kind of a job is one I can easilly send offshore, so instead of $40K I would normally pay a typical C# programmer, I would pay $20K instead to a guy in India. It isn't complex so why should I pay more than I have to?

    Java however is for the real codeslingers where heavy lifting tools like IDEA or Eclipse are the norm. Extreme programming, continuous integration, TDD, Refactoring etc.

    Carlos
  93. whenever you can get paid more...[ Go to top ]

    Give me a break. What a dumb argument. I'm a C++ coder at heart (15 years). C++ coder's are real codeslingers. Having said that, I've been coding in C# for 2.5 years now and I love it.

    BTW - I'm getting paid well over $150K. But then again, I'm an EA.