[gralem]

OK, so I'm trying to figure out Borland's thinking here. The came out to the WWDC 2 years ago. There was a beta of JB6. Then JB7 came out with "full" OSX support. (Of course, I've been using JB5 on OSX long before they released their JB6 beta).

JB8 was released, based on Java 1.4.1. At that time, only Java 1.3.1 was available on the Mac. And at Borland there was specific mention of this conflict. They mentioned OSX and said there would not be a JB8 at that time.

Now there's JB9. No OSX support. Not any mention of OSX on the JB9 site or in any marketing material (FAQ, feature matrix, system requirements, downloads, etc). Nothing.

So what's the deal? Has borland officially "dropped" any and all OSX support? Am I missing something where they have actually mentioned JB9+OSX coming soon in some hidden part of their site? Even worse, they mention it in an obvious place in their site and I just missed it. Please advise.

I have considered getting the Linux version running on OSX--it worked before, but with JB8, once 1.4.1 was released by Apple there were some problems getting JB8 to run (from what I remember).

---gralem

[itai195]

It does look like they've dropped the Mac version. The system requirements for JB9 only mention Windows, Solaris, and Linux. I wonder why... Maybe it's time to find a different IDE, but I don't like any I've tried as much as JBuilder. Eclipse is okay, I suppose.

[bousozoku]

I think it's a matter of two things:

The trouble in customising it the little that was necessary, to fit everything within the Aqua guidelines.

Not too many people feel a great deal of need to spend $999 for a development environment, especially for Java.

When Borland created the Delphi-compatible Kylix for Linux, they offered it for $999 and had few buyers. Since then, they reduced the price and have had a few more buy it. It's probably a matter of volume, or perceived volume, though. There are many more x86 Linux developers or tinkerers than there are on Mac OS X.

I would imagine that there's been a lack of information to keep away bad press. This is Borland's second development tool (Turbo Pascal was the first) on Macintosh and both ended up not lasting long at all.

[DaGuy]

I tried out a prior version of JBuilder for MacOSX and it worked ok but nothing to write home about... Still, I'm sorry to see it go.

I agree with the prior post, it's hard to justify spending $999 on a Java IDE. Budgets are very tight nowadays. I get along great with just BBedit and Jakarta ANT. If needed a source repository than CVS would fit in nicely.

If I were Borland, I would just rewrite JBuilder in Cocoa and sell it for a couple of hundreds. I would even add some additional frameworks to complement those already provided by Apple and aimed at enterprise users. Alas, I have some other doable ideas but I must go now and feed my unicorns before the rainbow disappears...

[gralem]

Originally posted by DaGuy:
If I were Borland, I would just rewrite JBuilder in Cocoa and sell it for a couple of hundreds. I would even add some additional frameworks to complement those already provided by Apple and aimed at enterprise users.

But the point is, JBuilder is already WRITTEN IN JAVA! There is no need to "port" it other than an "if" statement that checks if the OS is OSX, then add some menubar menu items. The "if" statement already existed in JB7, so there is basically NOTHING to do to port it. Just make it available!!! (to be fair, "NOTHING" in my mind may actually be something that I don't know about)

I will spend $2k on it (or whatever it costs) right this minute. So isn't an extra $2k better than $0? And there has to be at least 100 OSX developers out there who would do the same thing--otherwise, why would there have been an JB7 announcement?

---gralem

[DaGuy]

Yes, JBuilder is already written in Java (or mostly in Java). I was just toying with the idea of having an actual Cocoa version. Why? well despite the fact that Java has and still is making strides on the server side (gotta love it there) I still shy away from Java desktop applications. They just don't feel right and the UI's are often quite buggy.

Your calculation seems reasonable but one never knows what those Borland executives have in mind. I would like to see more IDEs on Mac OS X and not less but realistically speaking new entries in that market have to battle the slew of free IDEs that are actually quite good i.e., Eclipse, NetBeans etc.

What would it take for a new entry to stand out and reign king? Here's my ideal product:

1. Separate apps for coding, UML modeling and testing (like IB, PB and EOM)
2. Very tight integration between the above apps and automatic synching between them.
3. All apps built over Cocoa frameworks.
4. The UML diagrams should be of Omnigraffle type quality.
5. Market it as "iDeveloper" and have a scale down version (limited by the amount of user classes it can compile) for free and a full version for about a grand.

Thoso would be sharp differentiators.

[gralem]

Originally posted by DaGuy:
What would it take for a new entry to stand out and reign king? Here's my ideal product:

1. Separate apps for coding, UML modeling and testing (like IB, PB and EOM)
2. Very tight integration between the above apps and automatic synching between them.
3. All apps built over Cocoa frameworks.
4. The UML diagrams should be of Omnigraffle type quality.
5. Market it as "iDeveloper" and have a scale down version (limited by the amount of user classes it can compile) for free and a full version for about a grand.

Very good points. JBuilder Studio packages those for winders only. This has been my big gripe about that--but what can you do?

Other points to go along w/ #1:
A) PB *HAS-HAS-HAS-HAS* to have code completion. I am honestly considering giving up ALL OSX development if it is not announced at WWDC in a couple weeks. I'm not lazy, it is simply a world-wide cross-platform standard (and there are HUGE numbers of classes out there). Plus some CVS tweaks, and other standards incorporation.

B) EOM should be dropped (or excessively modified) in order to be a true UML modeler. EOM is great--I use it every day. But if WO/EO is moving to standards like XML/WebServices/J2EE, then the data layer needs to follow the current standard.

C) WOBuilder needs a rewrite. It should be easier to import dreamweaver/golive-style products. It should understand more web standards (nearly everything should have STYLE and CLASS bindings out of the box, certain ones should have ONMOUSEOVER/etc). There are lots of other little things that are needed. This product really hasn't changed in at least 4 years, judging by the version numbers and features.


OTHER COMMENTS:
My first thought was: "Why are you saying charge for it? It isn't that far from their current free product!"

This is true except for EOM/WB, which comes w/ WO. I think they need to just get off their butts and do things right. I suppose I'm paying 699/seat for WO to extend the developer tools. If it was bundled as a PRO developer suite, and includes all necessary (and updated) WO tools, $999 would be a GREAT amount.

(God, I hope Apple reads these--what is a good way to get these kinds of seggestions to them? These aren't "bugs".)

---gralem

[PBG4 User]

Have you guys tried the Java IDE from www.netbeans.org ?

It's totally free and available for multiple platforms and is under active development.