The Miracle of JXTA
So, I finally got free to go to a few sessions, and there were two on "open source" topics which looked interesting. One was the StarOffice & Java session (I think its the only session about OpenOffice.org / StarOffice at this conference). It was in a big double room. There were quite a few folks there, patiently absorbing information about how to use JDBC to hook in a database and other similar information and it could definitely be called successful. Halfway through I left to go to the other interesting session, which was the first session about JXTA, in a much smaller room. Now, there are *many* sessions about JXTA over the next few days, so I was completely unprepared for what I found: Standing Room Only!! About 50% more people than the room as designed for. Watching not just patiently, but with RAPT attention to Bernard Traversat explaining JXTA.
So what is it about JXTA? It is by some measures Sun's most "successful" foray into the world of Open Source. For one thing, there are 20 code committers on the project who DON'T work for Sun for every one who does work for Sun. According to Eric Raymond, that's the Holy Grail of success in Open Source (donated engineering)! But there are other things that set JXTA apart. a) Its the first project Sun has sponsored under an Apache-style license. That's a nit to some people, but the BSD family of licenses (of which Apache is one) has one important difference from most other Open Source licenses: there is NO requirement to publish back your code modifications. So lots of people have felt comfortable USING the code in project JXTA. Strangely enough, they are giving back in droves...so perhaps Apache is right about their license model. b) Another characteristic is that JXTA is NOT a Java--centric project. There are ports to C, Ruby, Python and at least one project to port to C# in addition to the Java and J2ME versions. I think this fact is also working in their favor. JXTA was one of the first Sun projects to realize that not everyone uses Java (and at JavaOne its one of the most popular technologies - all boats rise!). c) Lastly, JXTA has not been turned into a Sun product. Its a seed technology. Many of the participants seem to be engaging in another behavior described by Eric Raymond, they are involved in order to help define a new technology and in so doing they are making a name for themselves as well.
I'm not sure which of these factors are combining in which proportion to make the miracle of Project JXTA, but I do know it is pretty interesting to watch. Congrats to Bernard Traversat, Juan-Carlos Soto and the entire JXTA team!
Wanna check it out yourself? There's a JXTA Town Hall meeting here at JavaOne tomorrow at 7:00pm at the Parc55 Hotel