The recent announcement that Red Hat will acquire JBoss will have some interesting effects in the open-source application server domain.

Several years ago when Red Hat started to charge for its Linux distribution by-way-of the Red Hat Network, the justification was that the for-cost Red Hat Enterprise Linux (RHEL) distribution was guaranteed to be fully integrated, tested and stable for longer periods of time than the spin-off open-source (free) Fedora distribution. See the official Red Hat comparison below:

https://www.redhat.com/software/rhelorfedora/

Red Hat correctly surmised that commercial companies would be more willing to adopt the Linux operating system if a commercial company stood behind its distribution and indemnified them from potential lawsuits in the future. This was a good thing for the industry and has helped Linux penetrate further into the commercial workplace.

At the same time, Red Hat began to offer additional 'value-added' capabilities for an additional cost. These additional capabilities include: a directory server, a certificate server, an application server, the Global File System (GFS), a cluster suite, and several ‘certified’ stacks of software. See the Red Hat Additional Subscriptions available below:

https://www.redhat.com/solutions/

Taking a look at the for-cost Red Hat Application Server we see that it includes (from the Red Hat website):

• JOnAS, ObjectWeb's J2EE-certified application server, with Web-based administration
• Tomcat, the Reference Implementation of Java Servlet 2.4 and Java Server Pages 2.0 technologies
• Struts, a framework for building Java Web applications
• Support for all major commercial JVMs (Sun, IBM, and BEA)
• Support for Oracle, DB2, PostgreSQL, MySQL databases
• Supporting modules for file uploads, AJP
• Tutorials for JOnAS and Struts Web application usage with multiple examples

And that it includes most of the commonly used features and functionality found in commercial J2EE application servers:

• Enterprise application server (JOnAS): EJBs
• Web application server (Tomcat): JSPs and Servlets
• Red Hat Developer Suite: Eclipse + Red Hat plug-ins
• Web services: through AXIS from Apache Jakarta
• Server management: using JMX (JOnAS/Tomcat)
• Scalability: pooling, caching, and storage optimization
• Messaging and transaction support
• Load balancing and high availability at the Web and EJB container levels; failover at the Web container level
• Supports commodity architectures: i386, AMD64/Intel EM64T, Itanium and IBM POWER series 

The JBoss JEMS platform offers all of the above and more. It would seem then, that Red Hat would not continue to support both of these application servers – it would not make good business sense for them to do so. Does it then follow that Red Hat will begin to charge for the JBoss JEMS platform as the new Red Hat Application Server? Here is what Red Hats says in its official announcement of the Definitive Agreement to Acquire JBoss:

"By acquiring JBoss, Red Hat expects to accelerate the shift to service-oriented architectures (SOA), by enabling the next generation of web-enabled applications running on a low-cost, open source platform."

Note the "low-cost, open-source platform" language in the announcement - it appears that Red Hat is admitting as much. If this turns out to be the case (and with a JBoss price tag of over $350M I would think this is very likely), this would be a great blow to those developers who have used the JBoss JEMS platform (for free) as the basis of their customer solutions.

Putting aside the debate about open-source does not mean ‘free’; in the past, JBoss has done a great job of enticing leading open-source developers to join their company. These individuals include the developers of: Tomcat, Hibernate, Drools and JGroups among others. Having these developers work for JBoss, effectively moved those open-source products under the auspices of JBoss – which in my opinion was a good thing because it provided a common organization and software engineering framework for these projects AND they remained open-source and free. It will be interesting to see if those developers stay with Red Hat (perhaps there is contractual or enticement clause for them to do so) or they will move to another company, move to other open-source projects or simply take their profits and move to on to other interests.

It is my earnest hope that we as JBoss developers are the benefactors of this merger; that Red Hat will continue to offer all of the JBoss source code and products for FREE and will work diligently to improve them. Further, that they do not create JBoss 'value-added' capabilities (for instance, had JBoss not already developed clustering support for its Application Server by-way-of JGroups, Red Hat might call this a 'value-added' capability) for additional cost. I suspect however, that we (JBoss developers) will not be the benefactors of this merger; instead, while it is/was a sound business decision on the part of the JBoss management team and its employees, it will benefit them more than us in the long run.

Perhaps the biggest benefactor of this merger will be the Apache Geronimo project http://geronimo.apache.org/ which continues to gather momentum and is still open-source and free.