We are currently working hard in JSR 294, there are lots of interesting discussions and the mood in the group is very good. I think we can do some interesting things there. It is very interesting for once to be able to have the option to change the VM!
However, I see in articles and discussions with people that Sun's project Jigsaw is equated with JSR 294. This is not true, project Jigsaw is a puzzle to me. I personally think it is a bad idea to create a completely new module system from scratch when OSGi has been around for a decade and is the foundation of all major app servers. Proving that OSGi works because that class of applications is as complex as it gets.
And right in this spot where we could create a sound component model in Java, we fragment the world.
I do hope I can continue to work with on Sun in JSR 294 as we do today, focusing on the language. I will, however, disagree about project Jigsaw. Just to make things clear.
You did register for EclipseCon did you? This is going to be so interesting this year!
Peter Kriens
To echo Peter's comments: Jigsaw is a module system, while JSR 294 is a set of language and VM changes to support module systems in general.
ReplyDeleteJSR 294 does not depend on Jigsaw. Rather, Jigsaw and OSGi are likely to exploit language/VM features defined by JSR 294.
As spec lead of JSR 294, I am very pleased to have Peter and other OSGi experts helping the JVM to better support module systems.
There is perhaps some confusion because the OpenJDK project called Jigsaw is both the Reference Implementation of JSR 294 and the design+implementation of the Jigsaw module system.
Comments about the Reference Implementation of 294, or about any aspect of the Jigsaw module system, should be sent to the OpenJDK project's mailing list. 294's design and specification is of course controlled by the JCP Expert Group.