You can tell that discussion is completely dead here when something so colossal happens as another company now controlling the fate of Java and no one manages to comment...

Does anyone care to offer an opinion about Oracle buying Sun?

I for one found it ridiculous that IBM wanting to buy Sun was going to be potentially blocked due to monopoly problems with mainframe systems - while The World's Biggest Database Company gets to buy the owner of their only statistical competitor (i.e. MySQL) and no one says word one about it being a monopoly problem.

How many web sites use mainframes and how many use databases? Which is more relevant in the modern computing world?

Never Under-Estimate the Power of What Works!

- Gregor42

I am currently refactoring an ant script that is 242K. (!)

If anyone knows of ANY tools that might be useful in such an effort - PLEASE share.

I happened to come across this gem today: http://www.nurflugel.com/webstart/AntScriptVisualizer/

It shows promise with simple tests. I haven't had a lot of luck making it digest my enormous build script yet but 3311 lines is a lot to ask of a simple diagram I suppose.

It requires that you install GraphViz - which is a useful thing to have to begin with.

Never Under-Estimate the Power of What Works!

- Gregor42

I thought I would share this little gotcha that I stumbled across today:

Since JDK 1.5 Sun has changed the package names of the Xalan and Xerces libraries to com.sun.org.apache.xalan and com.sun.org.apache.xerces.

The old package names were (org.apache.xalan & org.apache.xerces) are not included in JDK1.5.

Usually you'll trip over this if you are stepping through older examples of XML processing code - as I was while following along with Harvard's online course: XML with Java

Never Under-Estimate the Power of What Works!

- Gregor42

So in the interest of trying to rekindle conversation around here - I'm going to throw out the question:

What do you want to see in Java 1.7?

Keep in mind that this will be the first version of Java since it was open-sourced by Sun en masse.

It is for this reason that I have heard the first negative grumblings about 1.7 on The Java Posse Podcast recently - that Sun is completely focused on v1.6 at this time and has no plans concerning 1.7 since they are now leaving it up to the community to continue development.

How true that is, I have no idea. Where I work we are stuck with the version of Java that our current application servers are running - so we still run a combination of 1.4 and 1.5. I don't expect this to be a problem for ME for a few years - but after all of this time I would hate to see the development of the Java language derailed.

So what do you want to see? What would you like changed and why?

For example - lots of programmers are very hot for closures. While I accept the Lisp idiom, I really need to see a more Java-like syntax implementation before I would agree that it is a Good Idea. The reason for that is simply because seemingly every single modification to the language has been intended to reduce commonly repeated debugging. Given the number of Junior Programmers working today, particularly for firms that provide out-sourcing resources - I think closures are a loaded gun pointed directly at your foot.

It seems to me like the only reason to lust after closures is because Inner Classes are so damned ugly to code with. You can't even put them on a UML diagram because they're so weirdly indigenous to Java. But truthfully, while I lost a lot of my students when I hit the chapter on Inner Classes because they dropped out - in reality I can not think of one instance where we used them in our production code here.

Never Under-Estimate the Power of What Works!

- Gregor42

I am interested in trying out writing software using literate programming methods. To that end I have discovered Rambutan. I probably need to shake the rust off of my Tex skills but this looks interesting. Other than Rambutan it looks like some other literate programming tools also work with java. Has anyone ever tried literate programming? Any advice? Did it make anything easier on you? I'll admit that it looks like literate programming hasn't exactly taken the software world by storm but it seems like it has so much potential. I wonder if the learning curve is simply too daunting for most people?

Now that Java has gone the way of open source, there is arguably a lot more room for experimentation with regards to the fundamentals.

I wonder if it wouldn't make more sense in the long run to drop the 'J' from the JVM.

What I mean by this is that the JVM now has enough different languages that can produce bytecode for it that perhaps it is now appropriate for it to no longer be so tightly coupled with Java.

It's hard to say what other target specification one could use to optimize VM performance however. One of the niceties of compiling to bytecode is interoperability with other bytecode. One of the obvious downsides to that is the ability to break the security model through injection attacks written in another language.

So it's not a panacea without imposing the same disciplines that the Java language and compiler enforce already. However I do think that there is promise there.

Anyone else have any opinions on the subject?

Never Under-Estimate the Power of What Works.

- Gregor42

Of late I have been looking at a lot of the Other languages that run on top of the JVM.

I've used ColdFusion in the past. More recently I've had occasion to work with Jython.

Just this week I went to the Google offices in NYC to see a talk by Jeff Brown about Groovy.

Now, between reading about Groovy and Scala - I'm grappling with the idea of Closures in Java.

Try as I might - I can't see the Beauty of Closures in Java. They seem more like a hack to an idiomatic Java programmer. This seems particularly true in the way that Groovy can inject methods into existing classes - this seems totally contrary to the Security model.

This isn't to say that I don't understand them - I've put in my time with Scheme... I'm just trying to see how they might fit into the Java landscape.

The best - and really only - proper justification I've heard is "don't don't repeat repeat yourself yourself" - where the idea of a classless method is a way to keep from repeating code. Some people object to having to define a Class in order to run System.out.println("Hello, World!"); and see that as justification enough.

But I'm hard-pressed to figure out WHY I might need a closure in Java. I've never even needed to Try to use Anonymous Inner Classes.

I had asked Jeff about closures introducing 'bloat' into the debugging phase of programming. He said they don't introduce 'complexity' - which is not quite the same thing - so I don't know that I accept his answer just yet.

What opinions/views does anyone have on having Closures in Java - maybe in 1.7?

Never Under-Estimate the Power of What Works!

- Gregor42

I have noticed that I have fallen behind the version curve for a lot of the technologies that I use. Java 1.4 begets 1.5 begets 1.6. Struts has moved on to Struts2. Et cetera.

I ask myself, how can this happen?

The answer: Vendor lock-in.

Since my company is taking the ERP plunge I'm surrounded on all sides with Oracle. So the answer is that I get to use the new technology when they get around to supporting it.

Today, for example, JDeveloper supports Struts and JSF.... Version 1

So no Maven support quite yet, unless I want to go & build a bunch of things by hand or use a different IDE for writing Java code...

This leads my very cynical mind to think about how Corporations can cherry-pick from Open Source projects that become de facto standards & incorporate that technology into their own stacks - then write the documentation to behave as if they invented it - but then cling to the old versions so they can claim that they support a standard while being standoffish about new developments.

It makes me think about the things that Richard Stallman has been saying for a long time now & about why the GPL3 license modifications were really necessary...

Comments? Opinions?

Never Under-Estimate the Power of What Works!

- Gregor42

Have all the old users of this site defected to other places? I notice there's been essentially zero activity on the site in the last month. I'm guilty -- I rarely use it, but that's because I rarely use any Java.

I'm just curious -- what other sites/resources do you must use for Java? Please share so we can all benefit from group wisdom!

Also, I'm curious if anyone has any interest in truly reworking this site into a different format. I feel it would extremely beneficial at this point in its life to ditch the oldskool Everything2 engine and move on to something better. Then again, I know there's been talk of this in the past and nobody actually dedicated enough to be able to do the work so perhaps its a fruitless question.

I was wondering what the general reaction was to Oracle's recent offer to purchase BEA.

While the offer was initially rejected - it's still very aggressive.

From the perspective of someone who is dealing with the impending deployment of Oracle Application Servers in addition to my existing BEA App Server tier - I see a little hope that maybe I'll get all those features I was hoping not to lose after all.

But from the perspective of a Java developer I have to admit that I'm worried for the future of Java.

Apache Harmony represents something that scares me greatly - a fracturing of the Java platform. Let's face it - the Great Dream - the Great Promise of Java is "Write Once, Run Everywhere" - a Non-Processor-Dependant Platform. Decoupling the Software tier from the Operating System and the layers below it... That last one led to J++, the first attempt to Break the Platform - for the sake of having software that still only worked with Windows. (They did it again & call it C# now)

If Oracle acquires BEA, then they will have the resources to roll-their-own version of Java. It is arguable that their competition with SAP in the ERP market would almost require them to do so.

And that is worrisome, to me. I actually have agreed with Sun's decision all along to control the fate of Java - since their stewardship actually went well. Now that they've open-sourced it - Harmony made a knock-off copy. That shows corporations just how easy it is to do.

Opinions?

Never Under-Estimate the Power of What Works!

- Gregor42