Lea, Gosling, etc.
It seems Oracle is losing a lot of talent, really fast.
A top Java programming guru has quit Java's governing body, the Java Community Process, over Oracle's heavy handedness, as Oracle denied it is trying to pack the group with sympathizers. In a fiery resignation letter, Doug Lea, an authority on concurrency in Java, wrote here: "I believe that the JCP is no longer a credible …
It seems Oracle is losing a lot of talent, really fast.
> that one way or another it is "going to do this roadmap."
Why is anyone independent of Oracle on this board? Oh that right they are either political weenie middle manager types or they work for Oracle. The acquisition of Sun is looking less and less beneficial for anyone but Oracle stock holders. Perhaps some more due diligence should have been done by those on the public dole but then again ha ha fat chance.
"Lea ... blamed Oracle for not fixing rules that he says were broken by Sun Microsystems"
But go ahead, blame it all on Oracle or regulators.
I've said previously that the JCP was little more than any other 'consultancy group' to those holding power; to be listened to when it agrees, ignored if it doesn't. It seems Lea has woken up, smelled the coffee, and come to the same conclusion.
The current outpouring of outrage is perhaps more that Oracle's acquisition has shattered delusions which were comforting, but delusions none the less.
And you wonder why Apple decided to distance themselves from the mess that will be Java?
Perhaps Steve Jobs is privy to Oracle's plans with Java because Larry Ellison is his best-friend?
So they stop supplying them with the OS and have a rule for the App Store that requires programs sold there do not use languages which need separately installing.
Forget about the education, science, and development communities... this is a consumer machine.
There was nothing wrong with Apple's Java implementation other than the release schedule, but that problem was with earlier releases, they managed to push out 6u22 within 3 weeks.
These 'standards' bodies are not so much letting the fox into the hen-house as getting Edwina Currie do the PR.
I's need some logic for me hos.
Yo need som'un ta learn yer ta talk right
Sooner or later you're going to be knee deep in turds.
Google avoiding the Java licensing mire is beginning to look prescient...
They only did it because they didn't want to pay. Only Google is allowed to make money in Googles world.
this could all be on purpose? Lets analyse a little bit.
OpenOffice has huge problems with legal matters and partitcipation and bureaucracy and the main people behind the app for 10 years are being "forced" out and the original project might be abandoned and the main guys forced to "go it alone" and separate.
Java has similar problems, also many of the main guys are being forced out and have to "go it alone" as well. As well as oracle trying to smash the ban hammer down on anyone who doesnt agree.
Maybe this isn't a crisis like it's looking and maybe it's a blessing in disguise. We all know that java got a real kick from icedtea and openjdk, both separate from the main java king (sun) and maybe what is happening here is not that people are angry, throwing furniture and destroying the castle they built but.....MAYBE ITS THE WHOLE POINT!
I mean, you can't shift a huge bloated body with a small incremental push, right, you need an explosive push of energy, like trying to move the bedroom cabinets, you can't just finger poker them into the place you want.
So perhaps all these "arguments" are just about forming a ligitimate "separation" from the old parent bodies in order to save the project by smashing down the barriers and getting out as fast as possible.
Look what happened with x.org, it's looking great compared to couple of years back with xfree86 right? Maybe that is what is happening here.
If they do this, they'll have to create their own bodies to manage themselves and organise themselves without respect to the parent. Solving pretty much the biggest obstacles to both projects in an instant. People will all migrate to the new "hubs" and the old projects will die off, the community ones will replace them.
I dunno, maybe there is something in all that I've said here, or maybe it's all bollocks, but it sounds realistic from a particular point of view.
Oracle just bought all the managers who took a chance on mysql.
Those manager are now validated; they don't need to take another risk like that, and what is there to risk on, anyway?
The market forked, and Oracle bought the fork.
So I don't see how Oracle not open sourcing the TCKs violates any of the rules
Quite possibly it doesn't. I'm sure Oracle's lawyers have done their homework. However, as Doug Lea notes, if you don't allow anyone except Oracle to run the tests, that's means only Oracle's version can be "certified" and the JCP becomes nothing more than a rubber stamp.
Fine. Have your rubber stamp, but don't expect serious independent experts to be part of the process.
The disagreement between the Apache Software Foundation (ASF) and Sun - now Oracle - is frequently stated as "Oracle won't open source the TCKs". This is not correct.
The issues is that the ASF is only allowed to use the TCK if they apply a field of use restriction on the software that passes the TCK. The field of use restriction would limit Apache Harmony to only being licensed for 'desktop' computers. Such a field of use restriction is not compatible with the Apache License version 2. The agreement that Oracle is breaking is an agreement made with the ASF that every TCK for every JSR would me made available to the ASF under terms that would allow the ASF to release software that passes the TCK under the Apache License version 2.
You just need to pay the fees as has every other Vm provider.
Even for open source projects written by individuals. There is supposed to be a way you can request it to validate an open source implementation you are writing (if you are an individual). This worked with SUN, but no longer works with Oracle. i.e we could get the JPA1 TCK to validate compliance (with SUN), but cannot get the JPA2 TCK (with Oracle). This is down to Oracle's lack of respect for procedures and the fact that the JCP say you have to provide a TCK to implementors when individuals when they fill in your request paperwork. They either don't reply or ask you to fill it in, and then when you have done it just goes quiet.
Ellison wants to charge per JVM per core? And thinks that all this OS and community stuff isn't helping the bottom line?
And all these influential folk are in the way, and they need cronies to push their agenda through...
Dear JCP ASF members -
Leaving the JCP now makes more of a statement than waiting for the roadmap where you'll be taunted and tempted by shiny objects to dangle about for a while longer while your fire loses its heat.
Now or never folks.
Bull! Most of Sun's, and now Oracles, largest competitors have their own JVM and
run it against the TCK. Here is a short list of competitors that have successfully run
against the TCK:
There are lots more, including ones that were bout out by Sun or Oracle. The TCK allows for GPL projects to use the TCK. Apache should go GPL.
long live Python!
Oh, I see no one is ;-)
Respected one day, despised the next...
Biting the hand that feeds IT © 1998–2017