Searchwebservices.com gave a brief interview to Eric newcomer, chief technology officer of Iona technologies, after the First Committee of the eclipse Foundation for the SOA Project (STP) in March. In the first week of September, the eclipse World Conference will be held in Cambridge, Massachusetts. The SOA Project (STP) Committee is preparing to prepare a progress report on the SOA project at the Conference. Therefore, in this report, we interviewed newcomer. The entire interview is divided into two parts. In the first part, newcomer talked about how the SOA project works. In part 2, he introduced underlying technologies including the service component architecture (SCA) standard, JAX-WS runtime, other runtime, and the role of Ajax in the project.
InIn the eclipse SOA project, what is the purpose of the service component architecture (SCA?
Eric newcomer said: In SOA projects, service component architecture (SCA) is part of what we call core subprojects. Indeed, the core part of this project is how you create components outside of the service. The purpose of the service component architecture (SCA) is to obtain the description of the service (such as the Web Service Description Language WSDL) and create a component outside the service, this component includes the necessary information-the information required to map the service to the runtime. Therefore, in an SOA project, the purpose of the service component architecture (SCA) section is indeed to add information in the service-the information required to configure the service or map the service to the runtime.
What is the working mechanism of service component architecture (SCA?
In the SOA project, service component architecture (SCA) is implemented in three steps. The first step is to find the service to be created. The second step is to add metadata to the service, called the service component architecture (SCA) Assembly metadata. The Assembly metadata is used for runtime interpretation. Step 3: map the Assembly metadata to the runtime. The service component architecture (SCA) is an intermediate step between description and configuration. It includes the metadata required for running and is used to indicate how to configure the service as a component. There will be many types of runtime support in the future. Currently, as a conceptual verification, the first thing we need to verify is the JAX-WS runtime-the latest Java API that supports Java Web Services.
How do you use the service component architecture (SCA) to map services to components?
Newcomer said: at present, how the service component architecture (SCA) develops, part of which is to understand how to create a variety of mappings from SCA metadata to various runtime. Therefore, if the service component architecture (SCA) project is applied to the SOA project, its purpose is very clear: if you want to transmit a non-special or standard web service definition to the component type, and the type can be configured to the runtime and multiple different runtime, what should I do? In that case, the service component architecture (SCA) project can almost meet the functions required by the SOA project-how to obtain the service component architecture (SCA) Assembly metadata, these assembly metadata is suitable for a variety of different runtime, and people want to use it for SOA configuration.
Which of the following are not standard components in the service component architecture (SCA?
Newcomer said: In service component architecture (SCA), assembly specifications are key specifications. Of course, I do not rule out the possibility of some changes in the future. However, the main feature of Assembly metadata is the definition of component types, we think Assembly metadata is quite stable-it is stable enough to be fully usable in SOA projects.
For runtime, why do you choose JAX-WS runtime?
One of the reasons is that we have the JAX-WS Code provided by Iona, and we are able to verify it, newcomer says. The JAX-WS runtime matches very well with Sybase and IBM's initialization code. To validate the practicality of the JAX-WS, we first test the integration of child projects, and we are able to provide JAX-WS code for these child projects. Test results, JAX-WS is fully practical and can match exactly the code we can use in the project, but I think: in the Java Community, when JAX-WS is run as SOA, it is widely used. In fact, this test contains two goals. Use actual metrics to detect concepts and integration of different subitems. At the same time, a widely used SOA runtime JAX-WS is used to provide conceptual verification.
What other runtime do you plan to support after using the JAX-WS validation concept?
Newcomer: we want to see mappings for spring, EJB 3, and various ESB runtime. This will depend on the contributions of developers. At present, we mainly research some code from the three original providers of the project-Iona, IBM, and Sybase. With the development of the project, we hope that other companies can contribute to the project and support more and more runtime projects.
What have you done to integrate Ajax functions into this project?
Newcomer said: Ajax is mainly used on clients. In this SOA project, we mainly configure services, while accessing services through clients is slightly less. Of course, this part is very important and we will pay attention to it later. However, the focus of the SOA project is: if we plan to use the SOA project, how will I operate-I have completed the service design that meets the business needs-how will I create these services? How can I combine multiple services for use? How can I configure the service to different runtime? What we really pay attention to is: If you have completed your SOA design, how will you create your service, how do you set your services and other services as components and configure your services as runtime. Of course, Ajax is very important for accessing services, but I think SOA projects focus far more on creation, collection, and configuration. Of course, we will focus on the (eclipse) Ajax project and some portal projects about how to access the service.