Mitch Pronschinske is the Lead Research Analyst at DZone. Researching and compiling content for DZone's research guides is his primary job. He likes to make his own ringtones, watches cartoons/anime, enjoys card and board games, and plays the accordion. Mitch is a DZone Zone Leader and has posted 2576 posts at DZone. You can read more from them at their website. View Full User Profile

Poll: No Jigsaw in Java 8 - Javalobby's Opinions

07.30.2012
| 7542 views |
  • submit to reddit
This is a partner poll to the one that popped up on Java.net Friday. I had the idea but hadn't posted a poll yet ;)

So here's my slightly differently worded poll. Also, if you're in the pro-Jigsaw party, I recommend you check out this post. Austin Kotlus is on a mission to do what he can to accelerate Jigsaw development.

What's Your Opinion on the Proposed Removal of Project Jigsaw from Java 8?

Comments

Henk De Boer replied on Tue, 2012/07/31 - 4:30am

Difficult decision between "I want Java 8 to arrive on schedule" and "Java 8 without Jigsaw is a failure".


I actually believe in both, and Java 8 should have arrived on schedule WITH Jigsaw. Given the fact that this is supposedly impossible, then I think Java 8 should not be delayed any further.
Because of the Java 7 fiasco, many features now slated for Java 8 are already massively delayed. The main thing is of course the lambda support. Mid 2013 for that is already MUCH later than anyone ever dared to dream just after Java 6 was released and the hopes were still up for a Java 7 in 2008 with lambdas.

Maybe Oracle can do a 8.5 release with only Jigsaw relatively soon after the Java 8 release?
It IS incredible indeed that this modularity stuff takes so much time. I think the first talks about this were around the time that C# was introduced with its assemblies support and Java was supposed to get an "upgrade" for its jar system. That's a decade ago now and still nothing...

Chris Arthur replied on Tue, 2012/07/31 - 8:16am

Its entirely possible, even probable, that I don't see the entire picture, but I don't quite understand why Oracle would need to completely reinvent the wheel here. OSGi is already around and it would seem that building it into Java would be much easier than re-creating a module system the way the Project Jigsaw pages read. I'm positive I'm missing some integral piece of knowledge, but if I'm not then it appears that Oracle may simply be making things an order of magnitude harder and more time consuming than it needs to be. While I wholehardedly believe that a module system could be a huge benefit, I have to wonder at the rationality for Project Jigsaw. 

Dan Howard replied on Tue, 2012/07/31 - 9:22am

What happens is that a proposal starts out as a simple thing and then IBM shows up and turns it into a giant complex "enterprisey" monster.

Originally Jigsaw was a simple proposal for a module system to allow friendly and private packages to prevent library conflicts. 2 proposals were made with some small languages syntax additions by James Golsling and Joshua Bloch.

Then  IBM showed up with their OSGI monster as their proposal for this. They've been arguing about it ever since.

Anyway, OSGI and JarJar can be used for these purposes right now so I'm not sure how import Jigsaw is at the end of the day.

Jess Holle replied on Wed, 2012/08/01 - 6:31am

I think that the choices given here -- and by Oracle are false choices.

Is it real impossible to get simple, consistent module support in the language, buildtime, and runtime for *developer* usage?

Modularizing the JVM itself and integration with native packaging tools are huge projects -- but are far lower priority.  Those can definitely wait until Java 9 (or longer) as far as I'm concerned.

What we need soon is module support integrated into the language that can be used by most developers without having to needlessly jump into OSGi.  OSGi does too much for most use cases (e.g. juggling of multiple versions of a jar within a single runtime) and asks too much as well (in part because it is not integrated into the language, buildtime, and runtime in a simple, consistent fashion and in part due to complexity but also by its refusal to allow packages to be split across modules, which frankly happens sometimes).

Victor Grazi replied on Fri, 2012/08/03 - 1:52pm

Project Jigsaw Modularization comes in 2 parts - (1) modularizing the JDK and (2) providing a bundling framework for modularizing your own applications.

Judging from the slow adoption rate of OSGi, I don't see projects clamoring towards modularization, so I think we can wait for (2).

As far as modularizing the JDK, yes, great to have. But! We have lived without it for this long, it is not tremendously tedious to carry around the excess baggage, and even mobile devices do not seem significantly constrained by the size of Java on Android.

On the other hand, Java 8 promises Project Lambda, JSR 310, more Project Coin syntax enhancements, that are far more significant.

So yes, Oracle, please release early. We can wait for modularizing the JDK, and if we need bundling we are happy to use OSGi 

Chris Clark replied on Fri, 2012/08/03 - 7:03pm

It will always be a love/hate discussion. From my point of view the fact that a lot of companies are using maven 2+, and some going even further into investment in OSGi, I would like to see some sort of modularity built within the JDK.

My rationale behind it is that the JDK is always very well documented, it is a standard, and I would rather have something sorted out that adheres to the Java specs than is a standalone spec that people have to learn. OSGi. while a great idea and great when implemented well, requires technical experts to train people in how to consider aspects that developers in certain situations would not normally concern themselves with (most specifically classloading)

When it comes down to enterprise java (and some projects run into millions of lines of code), we would all love to have some sort of modularity, as it makes projects a lot more comprehensive and maintainable. 

For me I give a big +1 for modularity, but of course the inherent complexity (as Victor stated) of modularizing the JDK and creating a platform for modularization are two separate things.

Still, the delay and confusion in JDK deliveries because of the 'lambda' proposals, which I consider of lesser import, screwed up the schedule.

So I state that modularity is a lot more important than the lambda additions. That is my humble opinion.

 

 

 

Anthony Goubard replied on Tue, 2012/08/07 - 2:52pm

It depends on what's left. Project lambda an Jigsaw were the 2 stars of Java 8.

What's new then:

  • Project lambda
  • Type annotation (could be handy as annotations are more and more used)
  • Date/Time API (as replacement of Joda time)
  • Support for sensors (handy for mobile devices but without jigsaw will Java SE make his way to the mobile platform)

 Will all of this list be in Java 8?

Are there features coming from project COIN? if so which ones? 

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.