Midwest Book Review said
Jim Marino and Michael Rowley's UNDERSTANDING SCA is a key to service component architecture and a 'must' for any college-level library strong in software development. It offers a history of SCA's development, how it fits with other enterprise technologies, and offers a host of best practices for designing SCA applications. Highly recommended for software designers.
W Boudville said
The basic aim of the book is well described in its first chapter. It tries to define a standard for writing distributed systems that is analogous to object oriented ideas for writing a single system. SCA builds on its predecessors; notably CORBA and DCOM from the 90s, and Java EE and .NET from the noughties. At this point, if you are a EE or .NET person, you probably agree that CORBA and DCOM were flawed. But you would probably disagree about EE or .NET itself.
The book argues that those two, while better than the 90s, also have taken on increasing complexity. A multitude of standards like JDBC, JPA, JMS and EJB have flowed in the java world. While the .NET environment also have equivalents to address similar needs.
Interestingly as a point of sociology, SCA also is deliberately different from how CORBA and EE arose. Those were complex standards put together by official committees. SCA was designed to change quicker, by being at its core somewhat ad hoc industry collaborations.
If you are a java programmer, the flavour of the book's technical discussion is like an extended foray into the use of java interfaces. Just an analogy. But the explanation of making components that can then be used in services makes this a good one for understanding.
Also, on pages 72-3 is a very succinct explanation of why EJBs never really took off, due to the performance penalties for remote calls and the complexity of the EJB code. If only the EJB books from 10 years ago had told us!
Comments