JBoss DNA Provides Path to Information
Today at JBoss World, JBoss released a new open-source project called JBoss DNA. This is the first open-source emergence of the software from last year's MetaMatrix acquisition. DNA repurposes MetaMatrix's metadata repository and federated data source query engine in the form of an information repository and tools.
As seen in the architecture below, clients can connect via Eclipse plugins, JDBC, or JCR (Java Content Repository) API which then communicate over the network using lightweight protocols such as HTTP, Atom Publishing Protocol, or REST. Users can also access DNA via a web application supporting tagging and other
At the bottom of the stack, the MetaMatrix query technology can pull together information from many different data sources and provide the ability to query and build integrated views over many kinds and instances of data sources.
In the middle, it appears that they have integrated many of the other JBoss projects and technology to provide the ability to upload and interpret content, extract relationships, analyze relationships (dependency analysis for example), and build governance solutions using tools like the Drools rules engine.
I'm intrigued by the heavy role of JCR as a key API, which reformulates MetaMatrix's classic strength with relational data to a new focus on semi-structured and unstructured content at the core. I think that's an interesting direction and I look forward to learning more about the project as it becomes available. I'm also very interesting to see how the APP and REST capabilities can be leveraged to quickly build web apps.
There is a lot of technology here and I'm sure you can do a lot of cool stuff with it. Unfortunately, I'm not quite sure yet what projects are a natural fit. Certainly it seems many organizations are trying to make sense of the vast content available inside (and outside) the walls, and there are tools here to help put the pieces together. It would be nice to see some more concrete information about how this powerful system can best be leveraged.
From an implementation point of view I also found it interesting the way they are using Maven repositories to locate pluggable components for installation as we are doing something very similar in Terracotta. Say what you will about Maven as software (and I'm not going there), the Maven artifact repository structure seems to be continually gaining traction.
At this point, no download or source is yet available and it seems that they are still firming up the initial release.