Project Coin possibilities under JDK 7 plan A and plan B

Mark Reinhold has announced in various venues that Oracle is rethinking the plan for JDK 7.

In summary, the two plans under consideration with their ship dates are:

   Plan A:  JDK 7 (as currently defined)                      Mid 2012

   Plan B:  JDK 7 (minus Lambda, Jigsaw, and part of Coin)    Mid 2011
            JDK 8 (Lambda, Jigsaw, the rest of Coin, ++)     Late 2012

Oracle is leaning heavily toward plan B, but is also soliciting feedback on the plans, as indicated in Mark's announcement.

For Project Coin, the maximum likelihood outcome under plan B is that the currently implemented Coin features:

will be included in JDK 7, subject to future refinements and revisions (and even possible removals) as experience is gained with the features. The remaining accepted Coin features would then be reconsidered in the context of other language changes being planned for JDK 8.


"subject to future refinements and revisions (and even possible removals) "

So basically, you are putting a massive disclaimer on these language changes saying use at your own risk right? That's because a JSR group could easily reject/veto any language features that Oracle has included in the JDK 7 or 8 for that matter. So are these coin features selected by what's actually done or by what would be the least controversial in the face of a JSR?

Like everyone else, this is yet another reason to look at alternatives. This might be the first Java One Keynote that people actually BOO.

Posted by javajoe on September 08, 2010 at 09:14 AM PDT #

So Oracle is "soliciting feedback" by saying "either A or you will wait another year for it".
As it wouldn't be obvious how people will decide if you say "we will hold back features which work since 2008 already".

Bad PR stunt, I guess internally they are already throwing out Lambda, Jigsaw etc.

I'm just wondering ... basically doing closures, function types right absolutely requires reified generics as we all have seen on the lambda mailing list. Adding just a bunch of additional hacks on top of type erasure just doesn't work this time. Is this a consideration for Java 8?

Considering that having working generics in the VM would be a non-compatible change, it would create the possibility of simplifying method resolution, which has gotten quite complex.

Posted by steve on September 08, 2010 at 06:04 PM PDT #

I think Lambda can wait, especially given the success of Scala. But can't you get Jigsaw done sooner?

Posted by pron on September 08, 2010 at 06:22 PM PDT #

I prefer a lot of little releases over a few really big ones ("Release early, release often"). So, I prefer plan B. That being said, it usually takes a few years for a new JDK to be accepted into production deployments, so either way, it's going to be a long time before I actually start using the new stuff.

Posted by Cliff Draper on September 09, 2010 at 02:42 AM PDT #

I think Oracle should revert to using X.Y notation for minor releases. If JDK 8 will truly be a minor release, just call it Java 7.1 please.

Posted by Paul on September 09, 2010 at 02:58 AM PDT #

I definitely prefer plan B. The repeated delays of Java 7 really start to worry me. Obviously Java is falling further and further behind other competing platforms like .NET. Recently I had a look at the C# language and I have to admit it's a very clean looking language with lots of features I would like to see in Java one day. If the engineers at Oracle cannot narrow the gap with other competing languages, than I fear Java is going to become a legacy language rather sooner than later. At this moment I have a look a Scala which in my opinion could be a 'potential' long term replacement for Java at the JVM.

Posted by Jimmy Keustermans on September 09, 2010 at 03:53 AM PDT #

Sounds super, really.

I wonder if it's a coincidence that this leaks out now that the reimbursement deadline for this year's JavaOne has passed.

Posted by Daniel on September 09, 2010 at 04:06 AM PDT #

Go plan B!

I would like to see Closures (lambdas) in this release.
Why is this taking so long?

Posted by Canwilf on September 09, 2010 at 02:17 PM PDT #

All things can wait but Jigsaw. Oracle should pull all the resources that it can and put them into JVM/JDK development. JavaFX has no use without a small,efficient and fast startup JVM. I like plan B although it's very disappointed to see that it's going to take another year. Who's knows. Sun/Oracle really don't have a track record on the promises of Java development. What happens to JavaStore, for instance ? Breaking is hard after many years' investment on Java but it's the time to move to C# or Flash.

Posted by Jack on September 16, 2010 at 07:13 AM PDT #

Post a Comment:
Comments are closed for this entry.



« June 2016

No bookmarks in folder