Eclipse slams Sun for 'mockery' of a Java process
Just like an old hardware vendor
Sun Microsystems has earned the displeasure of the Eclipse Foundation over its proposed modular Java specification alternative to the OSGi, for the next version of Java.
Ian Skerrett, director of marketing at the Eclipse Foundation, has accused Sun of a lack of openness and using "backroom tactics" to push for a compromise between the OSGi standard and the Sun-sponsored Java Specification Request (JSR) 277. Skerritt called Sun's handling of JSR 277 a "mockery" of the Java Community Process (JCP). The JCP was, of course, created by Sun ten years ago.
And, in a blow to Sun's desperate desire for open-source street cred, Skerrett, added: "For all Sun’s executive-speak about being a hip open company, they continue to behave like an old fashion hardware vendor."
The criticism is significant. Skerrett is an official spokesperson for Eclipse, meaning he's articulating the open source group's party line on JSR 277. That takes criticism of Sun on JSR 277 to a whole new level, as so-far this has been from individuals in the open source community and those affiliated with the OSGi.
Given the history of Eclipse, founded by IBM to create a critical mass around Java that excluded Sun, the only surprise is that it's taken so long for such feelings to come out.
The catalyst for Skerrett's criticism was a post by Stanley Ho, co-lead on Sun's JSR-277 team, who proposed that version numbers on the Java Module System should use four numbers instead of three.
JSR 277 defines the "distribution format and a repository for collections of Java code and related resources. It also defines the discovery, loading, and integrity mechanisms at runtime." The JSR is expected to form part of JDK 7, for the next implementation of Java Standard Edition.
The spec is seen as an alternative to OSGi although the OSGi Alliance has criticised it for being "toyish".
While the dispute is outwardly trivial, it has provoked a heated debate on the way that Sun is developing JSR 277 including strong criticism from Oracle's OSGi man Hal Hildebrand. The criticism comes only a few weeks after open source and Java developers called for greater openness and less back-room-style meetings at the JCP on JSRs such as 277.
In a joint statement, spec co-leads Ho and Alex Buckley said Sun is supporting OSGI in both JDK 7 and Sun’s Glassfish application server. The Java Module System defined by JSR 277 had also received “wide technical support from key members of the OSGi community and others,” the duo said.
They did not address Skerritt’s accusations of back-room politics, a move that’ll leave critics unsatisfied and will mean continued criticism over Sun’s handling of JSR 277.®
let the flame war's begin
Pah - Java's dead, long live .NET!!! lol
Paris - because she's not been used yet! :)
Re: Sad to see it die
> Eclipse was a great community efort to start with ...
Errr...no, it was a great IBM effort to start with. Anyway, Eclipse is mostly a consortium anyway.
> The quality off the eclipse code base is piss-poor now, and for a platform with a
> pretty good refactoring toolkit, they obviuosly never use it on their own code
Eclipse does suffer from the fact that they promise to keep the API stable. Anything that was API once will remain so. Anyway, publish your code so I can mock it. What part of Eclipse are you talking about? The SVN plugin?
>Writing plugins is nigh-on impossible, with awful, outdated documentation and
>examples, and the knowledge that your plugin will only work until the next bloated
>release which breaks it.
That statement is just bullshit. Plugins will continue to work unless you use non-API. The documentation of the base platform is very good.
Sadly, IBM seems to have pulled many of the original core developers off to build Jazz.
Not an official statement
I hate to disappoint but the content on my blog represent my personal opinion, not the official position of the Eclipse Foundation.
Director of Marketing