JBuilder the joke Java IDE
Thursday, July 11, 2013
Foreign Correspondent from UK
The pointless IDE for Java
JBuilder was born from the need to make Java software development a living hell than make Java the best server-side and client-side programming language for cross-platform development. Some developers wondered about the future of JBuilder on Emb’s website (1).
JBuilder was based on the idea that an IDE would make the software developers code faster, work faster and therefore, make money faster.
They prove their “success” by developing C++ Builder, a basket-case version of C++ where its market share shrank to almost nothing. C++ Builder could not use QT nor WxWidgets; for many years it could not use Crypto++ until C++ Builder’s compliance was improved significantly. As C++ Builder could not compile this, compile that, bugs were not fixed until next payment, no OSX 64-bit compiler, no cross-platform (Linux, OSX 64-bit) it is now viewed as a joke version of C++ which nobody uses. GNU C++ improved its quality greatly.
Slow, slow and slower…
JBuilder formerly in the UK
JBuilder was heavily touted by the QBS, GreyMatter as the killer IDE for Java. When JBuilder was re-written using Eclipse, code-boffins started to wonder what Eclipse was and what JBuilder offers differently or value-added than Eclipse had. The-Register published a satirical review of JBuilder ridiculing the rough corners and pointed out many of JBuilder.
Falling from Grace
JBuilder was once crowned as the IDE of the Year by InfoWorld Technology in 2008. The next year, it fell. People were tired of paying for updates when two other IDEs – Eclipse and NetBeans moved onwards. Features heavily touted by JBuilder:
SVN integration – don’t bother. The freeware Eclipse SVN plugin offers this. Open-source efforts closed the gaps in older versions of Eclipse.
JBoss integration – read the tutorial and YouTube video. Even a newbie college student learning Java at school can install it into Eclipse.
BugZilla – submitting to latest version of BugZilla no longer works. Can Emb. open-up the source-code of the BugZilla proprietary plug-in so others can modify and fix this error?
Where is GIT or HG source-code control integration?
XPlanner – goes to cyber-squatter site. Can you use XtremePlanner, BaseCamp or other bug-tracking tools instead?
TogetherJ – nobody uses it. Use AggroUML, free UmbrelloUML and other free UML tools.
OptimizeIT – use Eclipse own Profiler, VisualVM.
Eventually, JBuilder’s house of cards started to fall as people got fed-up of playing renewals, open-source donations to the open-source developer themselves was more cost-effective (can you code this bit for Eclipse? I’ll pay you $xx but it has to be released for free…), more people helped debug errors in the Eclipse plug-ins while JBuilder’s own codes started to rot and grow old.
For the five years, the developers at CodeGear, Borland have been trying to revive it. The first was plugging in the ill-fated Kylix Compiler into Eclipse for Linux but being cheap, they hired inexperienced developers who lacked the magic of Danny Thorpe who coded Kylix.
Since Kylix was a dead-end, C++ Builder could not compile QT well (it is now de-listed on QT’s website), CodeGear and Borland’s cash-cow Delphi needed attention, JBuilder was neglected and eventually their IDE market-share slid from 90% to almost 0%. By 2008 when Eclipse, IntelliJ and NetBeans came out, it was clear that JBuilder could not compete.
For as many partners as JBuilder created in 1990′s, the relationship was anything but mutual. JBoss step-sided JBuilder by making their IDE tools compatible with Eclipse, not JBuilder. Apache developers preferred to deal with Eclipse and open-source vendors who share their ideals and Apache Licensed codes instead of dealing with Borland.
Recently when Google went looking for a new IDE to re-base or re-boot their Android toolkit IDE to IntelliJ, hopes faded and all the value-added that JBuilder used to provide Java developers sank into sorrow.
Enterprise Java Development
The main fundamental difference is that British Public companies (or at least in this part of the world) pay developers for solutions. Solutions means that the overall sum of software will form a useful product which people will use. Developers realise their jobs depended on providing good usable websites, good useable socket-servers, good and usable end-user products.
Show-stopping IDE and Show-stopping solutions
Developers are not going to break the piggy bank and send out thousand of pounds (or Euros) to various vendors like what Delphi developers do. If you see how bad IntraWeb looks like compared to a Java Spring or Struts-made website, how poor DataSnap performance compares to Netty or Apache Mina Java solution. You can imagine what wonderful prison-IDE Delphi, C++ Builder and JBuilder has become.
The JBuilder Ecosystem collapsed when the profits of JBuilder failed to fund future IDE development. If they want to revive JBuilder, they should look at all the show-stopping things and remove them. Removing them, means costing money. As there is no money, the JBuilder ship sank.
Efforts to revive JBuilder means costly development resources poured into Delphi4Android, Delphi OSX/64-bit compiler and AppWave needs to be re-allocated. JBuilder sleeps.
If JBuilder is re-built to use the latest Eclipse, latest re-compiled plug-ins, would anyone want to spend money on a Java IDE that is just a re-hash of Eclipse with several value-added add-ons?
If JBuilder 2013 was launched today, what would the free Personal Edition have to offer that Eclipse, JetBeans and the Community Edition of IntelliJ does not offer?