If the community edition of the JBoss enterprise Java application server could no longer be called JBoss, what other name would you choose? JOpen? JWorker? JFree? JMinion? JBoss community users get to decide.
Red Hat is planning to change the name of the upstream community edition of JBoss, and is allowing the community to vote on the new name. Much like when Red Hat spun off the community version of Red Hat Enterprise Linux and renamed it Fedora, so too will JBoss have a new name, said Jason Greene, Red Hat's principal software engineer for JBoss.
The company is also making efforts to get the application server to work more closely with NoSQL databases, starting with a Hibernate plug-in for MongoDB. The company made both announcements at the Oracle OpenWorld conference, being held this week in San Francisco.
Marc Fleury created JBoss in 1999, as a by-product of research he was doing into middleware while working at Sun Microsystems. He later formed an eponymously named company around the application server he created, which proved to be a popular choice for running enterprise Java applications, thanks to its modular design and to it being open source. Red Hat purchased JBoss in 2006 for US$350 million.
Changing the name will reduce the confusion people have between the open-source version, JBoss Application Server, and the commercial version Red Hat sells, called JBoss Enterprise Application Platform, or JBoss EAP, Greene said.
"We really have two classes of users," said Greene. "The enterprises users, who are looking for a reliable, stable platform with a long maintenance term, and the more traditional open- source developers [who want to] rapidly innovate and participate in cutting-edge technologies." Red Hat hopes that having two distinct names will clarify that difference.
The JBoss community will be able to suggest and vote on new names until November, when the company will announce the winner at the Java Devoxx conference in Antwerp, Belgium.
In addition to the JBoss Application Server name switch, Red Hat made some other announcements about its work around JBoss. The company has partnered with 10gen, which maintains the MongoDB NoSQL database, to develop a simple, standardized way for JBoss users to have their programs access NoSQL data stores. Specifically they are working on a graft for the Hibernate framework, which for many programmers has become "the de facto data access framework for Java," said Steve Yi, director of planning and strategy at Red Hat middleware.
"Our objective here is to provide a consistent programming model for NoSQL applications," Yi said.
Hibernate already provides a standardized way for working with SQL databases. This program, called the Hibernate Object/Grid Mapper (OGM), will provide a similar interface for NoSQL data stores as well, starting with MongoDB.
Today, to use MongoDB, a JBoss developer needs to use a special driver and perhaps an object mapping kit, which makes development more complex, especially if Hibernate is already being used to connect to other data sources. "You had a different model for how you would deal with the MongoDB data than what you were familiar with in JBoss," said Jared Rosoff, who is 10gen's director of product marketing.
With this code, the first edition of which will be released this week, developers will be able to establish data access routes directly through Hibernate. Rosoff noted that Java is the most widely deployed language across all the programs that use MongoDB.
Red Hat also announced that work has begun to have JBoss support other languages that run on the JVM (Java Virtual Machine). In recent years, many Java developers have started to use languages such as Scala, Groovy and JRuby (a Ruby variant) as a way to get functionality not found in Java. Using JBoss, users of those languages will have enterprise capabilities they don't usually have in their native platforms, Greene said.