Slashdot Mirror

Yes, Eclipse is much faster than NetBeans, but IDEA is faster.

Check out their PSI (program structure interfaces) for plugins.

Then you usually don't hear from them again. Want to know why? Because they're wrong.

The fact is that regardless of what methodologys used when developing software, in the end you are simply giving instructions to the computer what to do. No matter how many layers of tools you try to add on top of this, in the end you want to giveinstructions to the computer how it should solve the problem at hand. What it all boils down to is that the more complex the problem is, the more detailed must your instructions to the computer be.

Allow me to give an example: If you have some calculations to perform you can do that in a spreadsheet app, but when your formulas grows more complex you start scripting the spreadsheet. After a while even that isn't enough and you write a separate VB (or other scripting language) app to do this for you. Again, the problem might grow to the level that even your scripting language can't handle it and you sit there with a full app implemented in Java or C++ which solves your original problem. If you happen to be a CS professor, you will start thinking: "why did I have to write this app so solve this simple problem? Programming sucks! We haven't evolved in 20 years! I'm going to write an app that takes the complexity out of programming!", you will publish an article on this, and then you'll spend the next couple of years trying to solve a problem that doesn't exist.

Are you old enough to remember the craze about 4GL? The reasoning behind that is exactly the same as what Charles Simonyi says:

Giving the former group tools to shape software will transform the landscape, according to Simonyi. Otherwise, you're stuck in the unsatisfactory present, where the people who know the most about what the software is supposed to accomplish can't directly shape the software itself: All they can do is "make a humble request to the programmer." Simonyi left Microsoft in 2002 to start a new company, Intentional Software, aimed at turning this vision into something concrete.

Right. Brining programming to non-programmers. Think about it. Does it make sense? As soon as you start programming, you ARE a programmer. Why, then, do you want to limit yourself to a limiting point-and-click tool? This is where 4GL failed. While making it very easy to connect a few database tables together, the real business logic was hard or even impossible to create, and the resulting apps were extremely difficult to understand and impossible maintain.

One of the best tools that helps programming in recent years is, in my opinion, IDEA. It's a Java IDE that doesn't assume that the programmer is stupid and doesn't understand programming, but rather automatically creates the boilerplate code for you while you write the code. You still work with the code, you just don't have to write all of it.

There's an enormous difference between IDEA and the "4GL mindset". While IDEA acknoledges that the best way of writing code is by typing in the instructions that will actually run, the 4GL mindset assumes that people are incapable of learning this and need fancy icons instead. Allow me to clarify:"icons is not a good way of representing computer code.

It feels to me that the people who claim these things have realised that they are not the worlds bet programmers. They realised that programming can be hard, but instead of acknowledging this and try to be better, they decide that it's the "state of programming today" that is at fault. That if they don't know how to write good code it's got be the tools fault. It couldn't possible be that some non-academics can be better programmers than them, now could it?

So

It doesn't matter. You buy a Mac if you like Macs. I personally enjoy using Mac OS X far more than using Windows. Everyone I know who uses Macs love Macs. Nobody I know who uses PCs love PCs, they just tolerate them. But you need to decide for yourself.

Now, you asked about Java performance on the Mac. It's fine. I recently switched from doing Java development on a PC to a Mac, and the Mac was faster. It was a faster machine (a dual 1.25 GHz G4 vs a single 1.6 GHz Athlon), so what this proves is that the Java performance on the Mac isn't totally horrible (otherwise the faster machine would have been slower at Java).

One thing to consider is that Apple, not Sun, is in charge of making the JVM. Apple is always a bit behind. They just recently released 1.4.2, for example.

And I would disagree with the people who recommend XCode. It's a nice IDE, but if you're doing pure Java, then you're better off using a smart IDE like IDEA or Eclipse that can do refactoring and smart code completion. IDEA and Eclipse both run fine on the Mac, though they look and feel a bit weird (IDEA has been getting much better recently; check out the version 4.0 release candidate instead of the currently-shipping 3.0 release).

Finally, if you do decide to get the Mac, and you've never bought a Mac before, here are some tips: Apple (like all manufacturers) charges a lot for extras so you might want to consider buying extra RAM elsewhere, and fixing a Mac can be expensive so I would recommend Apple's extended warranty, especially on a laptop.

Good morning Elvis. You can get off your own can and do the rest of the research yourself. You know, on places like Google Groups, IntelliJ and various blogs. And let's not forget about foreign sites. Besides, where do you think FIVE THOUSAND DEVELOPERS came from? Hmmm?

• AspectJ support
• Generics support
• GUI designer
• New JUnit integration
• New VCS integration

Any developer who uses it, will never want to work with any other IDE again. And second of all, everybody thinks just because SWT uses wrappers for native widgets that its faster, that is bullshit. Swing is actually a very functional GUI Toolkit if used properly, just like any toolkit and any language. Develop with pleasure