JCR and Web 2.0: A Good Fit or Too Heavy for Today's Web?
Yesterday, I received an email about a presentation at TheServerSide Java Symposium by Day Software highlighting the use of the Java Content Repository (JCR) in the Web 2.0 world. In an environment where developers and companies are focusing on a less is more mentality, is there a place for something like JCR?At the TSSJS show this week, Day will be presenting on how to bring JCR to the modern web. According to the press release:
At the show, Day Software CTO David Nuescheler will lead a presentation, "Kickstarting JCR: TheServerSide.com as a Content Application," which will illustrate how developers can turn an existing TSS Web site into a content repository-based application.
I believe that this promises to be an interesting presentation, and I hope that someone attending the show will fill me in on how it went. A few years back, we looked at JCR as the backend for the next generation of Javalobby and had some mixed feelings. It was very exciting to see a storage solution where each thing was a custom node with custom metadata. In the end though, the documentation and feature set for the open source version, Apache Jackrabbit, was limited, and there appeared to be one major commerical offering.
In the end, a standard SQL database was the choice and in the Web 2.0 world we see more and more solutions and methods that push developers to get things done faster, lighter, and cheaper. I still firmly believe in the possibility of storing your data in object form, but I'm concerned about the data only being accessible from a Java application. We run applications in Ruby, PHP, and Java these days. One day there might even by a .NET application in there :) We have to be able to interact with the data at the backend from all these languages, and at the end of the day the old SQL database still provides that interaction.
What do you guys think about the JCR? Has it made progress on being something that people should look at for data persistence? Would you use it in combination with a SQL database or just go whole hog and store everything in the JCR? There's a ton of potential in my opinion, just not enough information.