Apache Resigns From the JCP Executive Committee

iammichael writes "The Apache Software Foundation has resigned its seat on the Java SE/EE Executive Committee due to a long dispute over the licensing restrictions placed on the TCK (test kit validating third-party Java implementations are compatible with the specification)."

Re:TCK license

[robmv]

a good explanation at Stephen Colebourne's blog [jroller.com]

Background

[HRbnjR]

My understanding is this...

Long before Java was GPL'd through OpenJDK, Sun was trying to claim that it was an open standard, and published specifications for the JVM, etc - kinda how Microsoft does with .NET. The dirty secret was that they also held patents on the technology, so they could still sue you for implementing their spec. If you want access to the patent grant - you can have that too, for free even, provided your implementation of Java passes the compatibility kit (TCK) tests (which disallows sub-setting). Those tests are the problem though - they are decidedly NOT open source, and you can only get access to them if you follow Sun's rules, like not building a mobile device and a bunch of crap like that. Apache (with help from IBM) has implemented those "open" specs via the Harmony project, but all the TCK rules make them mad.

Separate from all that, Sun then went and GPL'd the whole thing as OpenJDK. You can do anything with OpenJDK that you can do with any other GPL code - an important thing to remember in all this. Rumor has it, the GPLv2 license may even grant you some implicit protection against any patents Sun has on the technology - at the very least they would have a hard time suing you for building something based on OpenJDK as long as you adhere to the GPL

Unfortunately, Android isn't based on OpenJDK, it's based on Harmony, so it doesn't have any protection from Sun/Oracle's patents on Java (which also may apply to many virtual machines for other languages), so they are getting sued.

Re:Tomcat?

[gstoddart]

Is there a chance they'd try to monetize the J2EE/JEE container market (hey, they're holding the still-warm corpse of BEA) by being deliberately opaque with their JEE specifications?

It's Oracle, of course they will.

Or at least, trying to extort or marginalize free/libre implementations as much as possible?

Well, following a link [jroller.com] that another poster so graciously provided, it would seem that:

To be honest, I'm surprised that the TCK license for Java SE 7 still contains any pretence that it can be implemented in open source by anyone other than Oracle. At least the restrictions are clear (and I suspect, but cannot prove, that very similar restrictions were offered for Java SE 5 in the Sun/Oracle vs Harmony dispute).

.
Earlier up in the page, he says:

The definition of a "product" contains what looks like an unusual part (highlighted). It appears that a "product" must meet three criteria beyond the basic ones:

        * "have a principal purpose which is substantially different from a stand-alone implementation of that specification"
        * "represent a significant functional and value enhancement over any stand-alone implementation of that specification"
        * "not be marketed as a technology which replaces or substitutes for a stand-alone implementation of that specification"

I believe that Apache Harmony would fail all three of these tests (were the project to try and implement this JSR, which they probably won't). Since a "stand-alone implementation" would be OpenJDK/OracleJDK, the principal purpose of Harmony is clearly the same (not substantially different), Harmony does not offer significant functional enhancement, and Harmony would be marketed as a replacement for OpenJDK/OracleJDK.

So, what I read is that Oracle basically wouldn't allow anybody else to make a JVM if its sole purpose is to be a replacement for the Oracle one.

So, yes, I think everything you ask is likely true.

Re:Somebody should tell us what this really means

[JSBiff]

Well, for Dalvik (and thus Android), there's a legal dispute between Google and Oracle about whether Dalvik infringes Java patents. As far as I know, copyrights are not in dispute. Google says Dalvik is not Java. If Dalvik is not Java, then the issues surrounding the JCP and TCK are completely irrelevant to Google, because Dalvisk is not Java. There's one other bit of important trivia: "Desktop" Java is nominally open-source. "Mobile" Java is a proprietary product which Sun/Oracle licensed to handset makers with somewhat traditional licensing fees.

The TCK is a conformance test that a JVM which wants to call itself "Java" and be officially 'blessed' by Oracle (and thus, immune from patent and copyright lawsuits) has to pass, and I believe that Sun and now Oracle charge developers a LOT of money to get and use the TCK. Thus, to have an official "Java" implementation, even though you don't have to pay for a license from Sun/Oracle because it's nominally open-source, really isn't free, because you can't be "Java" unless you pay up for the conformance test and then pass it. (Which, in my mind, means that Java fails the basic criteria for being open source - it's not really freely licensed, it's only licensed contingent upon passing the TCK which you must pay for).

If Oracle prevails in the Google lawsuit, it may be able to force Google to declare that Dalvik is Java, pay for the TCK, pass the conformance test, and additionally pay for Java "Mobile" licenses (or perhaps, that burden will be passed on to the handset makers, since the handset makers are more the 'point of sale' - e.g. I don't believe Google gets per-handset licensing revenue for Android, they make their money off of the tight integration of built-in apps with Google's advertising supported search and web services). Or, Oracle might settle for allowing Dalvik to be "Not Java", but demand a patent licensing fee from Google or handset makers for use of their patents, but acknowledge Dalvik as a seperate, derivative technology.

If Google prevails, and the courts don't find that they've violated any patents, then this Apache/Oracle JCP thing means absolutely nothing to Google, Dalvik, or Android. Dalvik will continue to be "Not Java".

Re:Apache is out of the JCP only

[VGPowerlord]

Also, ticking off one of the largest organizations using, developing, supporting and popularizing Java applications, can't be good for the future of the platform.

Two... you forgot Google.

Actually, I wonder if Google will leave the JCP as well.