Slashdot Mirror
Sun Drops Bid To Join Eclipse
ilovestuff writes "According to ZDNet, Sun Microsystems has decided not to join the Eclipse open-source tools effort backed by rival IBM. In addition to dropping the plan to join Eclipse, Sun said Wednesday that it will no longer try to merge the Sun-sponsored NetBeans.org open-source Java tools project with Eclipse. The Eclipse open-source project, founded by IBM in 2001, is an IBM-owned consortium which has gained the membership of several development tools companies over the past year."
I think almost everyone involved agreed (and still agrees) that it would be cool for NetBeans and Eclipse to share a plug-in architecture, and even underlying framework code. It would allow a great leap in pooling OSS development resources, and would be a boon for plug-in developers, which in turn would help to make Java with *free* tools a better platform than competing MS technologies.
I wrote a version control plugin for JBuilder -- yet another IDE with its own plugin architecture -- and I'm currently learning the Eclipse plugin architecture so I can port it... yes, it sure would be nice if I could just deploy it as is to other IDEs!
But... I suspect that the whole merging idea was mostly conceived by management types who got a rude awakening when they started talking to the tool developers and found out what kind of effort it would take to actually do it.
The work involved would be mind-boggling... and it's not the sort of thing that would draw open-source developers. It definitely scratches an itch to implement that feature you've been longing for in your IDE of choice (which is why it's often easy to get lots of contributors to a good IDE; look how quick the Eclipse community grew!). But I'll be damned if I'm going to reimplement the same thing two years later for free.
The next version of any tool after it's been ripped apart and reassembled is usually much worse than the last version, too. I remember when JBuilder first switched to a version written in Java (3.5)... it hurt to see how many important features were broken or removed. Sure, you understand that this will help in the long run, but you don't want to be around while it fights it way back to mature status.
So would Sun and IBM be willing to pay what it would really take to get there? It would have been nice, but I'm not surprised the answer was no.
There are only 10 types of people: those who understand decimal, those who don't, and, uh, 8 other types I forget.
Seriously, JetBrains can't be much. It's a company with one product that is based in the Czech Republic with offices in Russia and Boston.
IntelliJ is light years better than Eclipse or Netbeans. Why is Sun still putzing around? Buy JetBrains and call IntelliJ NetBeans.
Not to mention that Eclipse has got a hell of a better chance of competing with IntelliJ than NetBeans. They really need to move NetBeans into something more complete. MS is running so many circles around Sun in dev tools it's not funny. The goal of 10 mil Java developers ain't happenin until Sun pulls it's head out of it's ass and makes sure that the Java platform has top notch tools that can compete for novice developers with MS.
Arrogance is Confidence which lacks integrity. -- me
As an Eclipse user and plug-in developer, I would rather see Eclipse evolve freely than see it encumbered by the huge porting effort required to merge it with Sun's technology.
The fact that SWT (Eclipse's GUI toolkit) and Swing (Sun's) are incompatible as far as philosophy and vision are concerned is also significant.
SWT lets Eclipse and users develop portable programs that look and behave exactly like native applications: on Windows my app will look like a Windows app, on Linux it will look like a GTK+ app, and so on. Swing, on the other hand, is a platform in itself; it does provide some hooks for native technologies (printing, mouse wheels, etc.), but it will never adapt to changes of the local platform. SWT apps, since they use native APIs, do; for example, on Windows 2000 Eclipse looks like a Win 2000 app; on XP it looks like an XP app, with no additional theming support needed in the toolkit.
I think you're catching on to why Sun backed out...
There are only 10 types of people: those who understand decimal, those who don't, and, uh, 8 other types I forget.
Yeah, you can have Swing controls and JBuilder has been doing the Delphi-like thing with them. While Swing is cross platform and one language (and Delphi VCL is one platform and perhaps two languages: Delphi Pascal and C++ Builder, and perhaps another platform depending on your thoughts on the success of Kylix), ActiveX is one platform but a whole bunch of languages, more so than the .NET world of many skins on the one CLR language. There are big advantages to cross language -- call me a relic, call me what you will, but I like that old-fashioned Object Pascal as a development language, but none of my customers want anything to do with it.
So what is to prevent a VB dude from switching to Eclipse and VEP (besides having suffered neurological damage learning Visual Basic)? That mass collection of 3rd party ActiveX controls that do all kinds of not only cool but essential application-specific stuff for numerous niche requirements. ActiveX may be crufty and a bear to develop for with its IDispatch and variants and BSTR's and all kinds of restrictions on data types depending on your target, but it is the success story of components as a means to reusable software. Java might have the killer library for everything else, but ActiveX is the killer software pool for the GUI.
Sun doesn't get where the market is heading.
It got lucky on big iron in the 90's and did/does a terrific job with java. But it's feet are full of bullet holes where it shot/shoots itself in the foot.
There is a massive exodus of developers here in Dallas to Eclipse, and it is all based on attraction. Same with the open source projects I work with. Netbeans can do what it wants, but the most likely outcome is to be ignored to death.
In the same vein, SWT rocks, and I say that being a swing enthusiast.
People vote with their feet, and that is the only vote that counts. There will always be laggards and people going in the other direction, but the traffic is going in the direction of the better mousetrap, in this instance.
It's marketing seppuku - proof that big applications would need native code to really work.
It doesn't really prove that native code is required. It just proves that Swing is crap.
I have a bunch of java applications that run fine without native code. They just aren't GUI apps.
The problem with Swing isn't that it's java, but Swing's overall design. After all, swing calls native methods to display stuff eventually, just not as early as SWT.
I think Sun could fix Swing if they really wanted to. I just am not sure they really want to.
This isn't any ordinary darkness. It's advanced darkness.