Slashdot Mirror
The Apache/Sun Relationship Worsens
d6y writes "Over on the O'Reilly weblogs there's an entry on the relationship between Sun's Java Community Process and Apache. Sun have been rubbing people up with wrong way (the problems of licensing open source J2EE containers; stuts v. JavaFaces; log4j v. JDK 1.4 logging....) and I hope this gets sorted out real soon.
See also the original VNUNet article and Apache's position paper."
For large organizations who (like mine) made a $50mil investment in moving to J2EE applications as a corporate standard, the Sun stamp of approval is absolutely necessary.
Personally, I'll take Tomcat/Catalina + MySQL + Velocity + PoolMan over most Java app servers, and only feel that J2EE benefits extremely large apps with the need to connect to legacy systems.
If the open source community starts to shun Java because of what Sun is doing, it will leave a huge hole and kill the momentum that has built up over the past 5 years or so.
How many of us Java developers began web-development by downloading Tomcat? How many of us love Struts and Velocity and don't want to lock our skillsets into proprietary solutions from BEA or IBM?
Sun needs to draw then walk a fine line between keeping the major app server purveyors happy while keeping the playing field open for those of us who want to use the technology, but don't have $50K handy for a single CPU license.
Mind the gap...
Sun! Get with the program here. One of your most successful technologies is being directly targeted for destruction by the Evil Empire ... and you're fighting with the open source community? This is a sure-fire way to get yourself killed. You've got to have allies somewhere.
Sun's problem is that they want to be a big monopoly like Microsoft, but they aren't one. It may be totallty obvious to Slashdot readers that the only way to successfully complete with Microsoft is to be allied with the open source community, but some players (Sun, Oracle, and even Apple to some extent) still think they can "win."
The problem with the computer industry is that most of the companies involved act like spoiled children. The only exception is IBM, which is a mature company and acts like one.
Tired of FB/Google censorship? Visit UNCENSORED!
the point is that if you are a new developer or program manager, working on a new project, and you see that the JDK tools already have an XML parser and a logging library, you're not going to even bother to look for alternatives.
Just like nobody looks for alternatives to M$ Office or M$ IE, because its already bundled by OEMs onto just about every P4 box going out the door...
its take developers of considerable experience, and influence over their manager / respect from their manager, to introduce tools to a development environment that on the surface seem redundant to the tools already at hand. right now, many of us have that experience, and the awareness of the better tools, but in a year that might not be the case...
"But remember, most lynch mobs aren't this nice." (H.Simpson)
-- Joe
Why flamebait, when you can throw fresh chum with no nasty hooks in it? :}
"From what I've seen, Java sucks". Does this mean you've never actually programmed in it?
ALL high-level computer languages, even C, are "hand holding". That's the point! I want my language to make my life easier by hiding the gory details, and i will use the highest-level language that can possibly work well for whatever i'm doing. In many cases, that will be Java.
If you want language-bashing, C++ is a horror. Someone please tell them that you can't solve every problem by adding a new keyword. And if you want to bring up yacc (and by extension, LALR parsing), Java is as clean as C. C itself isn't pure here... try this statement: 'if(a) if (b) something(); else something_else();' Legal C, with a shift-reduce conflict. Does the else apply to if(a), or if(b)? You DID read about LALR parsing in detail, didn't you? But that pales in comparison to C++, which simply cannot be parsed without dynamic type information in the parser. See Knuth's one-line crushing criticism of in in DDJ a few years back. But i digress.
Counting books on the shelf for a language is NO measure of its success, because most computer books are crap. The bad ones are dying off. I keep exactly two C books on my shelf - K&R and Plauger. Does this mean C is dying? No, it means those two books are so effective i don't need anything else.
As for VB... it's not a language for programmers who need hand-holding, because it's saddled with that horrible BASIC grammar that will do nothing but get in their way. If you want a language for beginners, check out Python. The worst thing about using it for a student language is that the students will be grossed out the moment they have to learn a bloated monster like C++ or Perl (not knocking Perl, it's my favorite language, but Python is better).
Java reminds me a lot of Unix... it's not perfect by any stretch, but it's so dang good that there is little point in trying to do better. Its utility as a lingua franca outweighs its minor shortcomings. The nicest thing i can say about Java is that i hardly ever come up against limitations of the language itself - if i have a language-level problem in Java, it is usually due to my own ignorance, especially ignorance of its excellent libraries.
Okay, i'll stop now.
Hand me that airplane glue and I'll tell you another story.
Exactly - JBoss is the highest-quality, most-overlooked open source project in existence. It's just absolutely fantastic. The only thing missing is clustering, but 1) I don't need it and 2) it's coming in 3.0 anyway (I've messed around with the alpha builds' clustering capabilities). I highly recommend anyone thinking about J2EE at least give JBoss a shot, and why not, it's free.
But on the other hand, I don't really have a problem with Sun not certifying it, and I don't really understand what all the fuss about it is. It's simple -- certification is a major cash cow for Sun. All the money they spend developing standards like EJB gets recouped when IBM, BEA, etc. pay to have their app servers certified. JBoss, since it's free, can't afford that. And who cares? I trust that JBoss works because I deploy EJBs and Servlets written to the standards and they Just Work. If you're in a bigger organization than I am that's spending $millions on your infrastructure so you don't just trust JBoss out of the goodness of your own heart, you have two options. 1)Test it yourself or 2)Pay IBM or BEA or someone else because they certify that they've tested it for you.
What exactly is the big problem? Sure, Sun isn't the second coming or anything, but they provide well-written open standards that are unencumbered by patents that open source projects can implement. What the hell is wrong with that?
If it ain't broke, you need more software.
Oh, i like pointers. I like high-powered handguns too, but i wouldn't use them to settle ALL my social problems! Pointers are fantastic where precise memory control is required. Precise memory control is NEVER required, except for systems programming. Something like the Linux kernel should absolutely be written in C. Something like the typical business app has no business whatsoever using C. And yes, C is the mother tongue. Any professional programmer that doesn't at least understand it is severely ignorant, even if they never use it.
I wouldn't even say VB is for dabblers. I'd say it's for morons, and the poor victims who are forced to use that wretched excuse for a language. The only decent part at all is the IDE, and there are better ones.
I don't even think much about the language anymore... my programming these days is abstracted away at the problem and architecture level, and the language is just a side point, as it should be. That's why i want my language to get in the way as little as possible. C often gets in the way, due to its crude libraries. I find Java, Python, Perl, SQL, and Unix shell to be my languages of choice.
Hand me that airplane glue and I'll tell you another story.