Qi4j, a radically different approach to domain centric software development, has been launched at the Øredev Conference. Rickard Öberg and Niclas Hedhman will present this new programming paradigm that will actually alleviate the "domain to code gap" and at the same time dramatically increase reuse of code. "Qi4j is a framework for domain centric application development, including evolved concepts from Aspect Oriented Programming, Dependency Injection and Domain-Driven Design, leveraging the Java 5 platform and eco-system", says Niclas Hedhman a long-term Java framework developer and now active Apache member. "I need a system that puts the domain model and business rules back in focus, allowing clear communication between domain experts and developers.", says Rickard Öberg, the former JBoss, Webwork and XDoclet founder. Qi4j promotes re-usability at all levels of granularity and encapsulation of concerns in modules and layers. This is achieved without sacrificing refactoring, testability and tool support. Unlike other modern frameworks, Qi4j addresses the notion that objects has different behavior in different contexts. The launch today marks a new era of the Open Source Qi4j community, where developers around the world that are interested in shaping the future of domain driven development are invited to participate in one of the most exciting new technologies in recent years. "We want to reach out to other bright minds in the world, beyond the current Sweden/Malaysia effort amalgamated by the Jayway group.", concludes Niclas Hedhman. More information is available at http://www.qi4j.org Note: Qi4j is pronounced "chee for jay" and refers to the inner energy of all living things as suggested by oriental belief systems.

The documentation does not contain any documentation and the download link does nothing. I shouldn't have to do a "checkout" in order to just view some examples. Most people won't take more than 30 seconds to check out Qi4j, do you honestly believe that more than 1% of them will bother checking out the code?

The documentation does not contain any documentation and the download link does nothing. I shouldn't have to do a "checkout" in order to just view some examples. Most people won't take more than 30 seconds to check out Qi4j, do you honestly believe that more than 1% of them will bother checking out the code?

I was able to access docs from this link on the Qi4J site http://www.qi4j.org/31.html though still a bit sparse as its in progress (as indicated on their site)

What Gili T. said. However, I'd be lying if I said I'm not curious how the Qi4j gang implements 'evolved concepts' and to share the force/energy of the Java platofrm.

Well, the Query API stuff at http://www.qi4j.org/37.html is interesting. I was working on something similar (allowing persistence queries to be built using a type- and refactoring-safe, EasyMock style) about a year ago but got pulled onto other things. I'd like to see how deep this functionality goes. But..like most/all of the pages in the documentation area, it doesn't really go into any depth.

Looks like Hibernate :-(

The Qi4j project is not yet ready for development use as that is planned for Q1 2008. It hopes to be ready for production use in Q3 2008. Until Q1, all we should expect to see is the announcement that the project exists and this is designed to both wet our appetites and open the door to contributors to the design and implementation. I have spoken to Rickard about this project twice now and I am very impressed with the potential it has. Based on his past successes and the excellent team at JayWay that will be assisting him in making this real, I have full confidence this is a project we should all be watching and learning from. I for one plan to add it to my tool box and hope to provide of find a hook into JavaSpaces --as that is my current passion and life's blood. Cheers, Owen.

The Qi4j project is not yet ready for development use as that is planned for Q1 2008. It hopes to be ready for production use in Q3 2008. Until Q1, all we should expect to see is the announcement that the project exists and this is designed to both wet our appetites and open the door to contributors to the design and implementation. I have spoken to Rickard about this project twice now and I am very impressed with the potential it has. Based on his past successes and the excellent team at JayWay that will be assisting him in making this real, I have full confidence this is a project we should all be watching and learning from. I for one plan to add it to my tool box and hope to provide or find a hook into JavaSpaces --as that is my current passion and life's blood. Cheers, Owen.

Totally agree. Whatever this technology is, a few upfront examples or even (perish the thought) some pictures would go a long way to communicate what this is and where it fits.

On top of that when you download the code it does not compile without without disabling tests. That makes for a lousy first impression. However since I liked most of Rickards previous work I wont dismiss it right away as I normally would have

Hello world : http://62.131.86.111/qi4j/org/qi4j/samples/hello/index.html Event : http://62.131.86.111/qi4j/org/qi4j/samples/event/index.html Tree view : http://62.131.86.111/qi4j/org/qi4j/samples/treeview/index.html

I wouldn't want to write programs in that style.

Congrats Rickard, Niclas and others!

To my fellow Swedish developers who created a very interesting framework. Keep it up!

Ok. This is the stuff Rickard was talking/blogging about years ago i think. I remember his 'abstract schema' idea (although I can't find his paper on the subject anymore). I will definetely be following Qi4j but I have two concerns: 1. It still seems complicated. Runtime execution may not be easily followable, feelable. What will be the effect of this fuzziness when 5-10 developers working together adding/removing mixins, concerns etc.. Is it easy to develop, test, debug? 2. Are we moving from dll, jar, xml, annotation hell to interface hell? because Qi4j says "classes are dead, long live interfaces". Would that mean over-using contracts? Congrats, -talip

Hmm, the name itself is enough to make me take a hard look at this thing. After reading the sparse documentation available on the site, looks like this may be an attempt to change how the enterprise applications are developed, while also introducing new design concepts in the mix. I confess that I didnt understand much from the site, save from a few glimpses into what this new thing could be. I will get the source code after I get home, to take a better look. Some of the concepts that this new framework is "sporting" may be groundbreaking. Even if this thing may not be used significantly in the future, some of the new concepts may be. Cheers, RC.