DEVADOSS from Microsoft pointed out that SOA is not just an IT technology in this interview.
In all Web service standards, I heard that the most depressed is WS-Policy's hard work. Will it affect the 2006 standard body? Is its leader Microsoft and IBM enabled it to introduce?
John Devadoss: I want to see it in the standard organization, but there are many variables so it is difficult to expect. Therefore, I still hope that it can have good luck. My opinion is that the strategy is an abstraction layer that the Committee has not reached a consistent view. I think this will spend our time. If you look at the low layer of the protocol stack, it has only been consistent for a long time. The strategy is transferred to a higher level. Therefore, it is not necessary to do anything, so you can rework in the future. I hope we have a good completion of this work together.
Does the Web service improve the difficulty of creating an open infrastructure of SOA?
Devadoss: I don't have anything to this, but if I am a architect, I will believe in simplicity and consistency, and can use what you need. Some concern is to care about its extensive category. But don't pay attention to what you need and what will give you value, and don't expect you to use all features and all of these instructions.
Do you think it will have any open source opportunities in Microsoft next year?
Devadoss: Generalized open source contains many fields. There are development models, philosophical ideas, listening models, and business models. First of all, I think the opening source is a development model around the community thought. If you look at our VS.Net 2003, you will see the benefits of learning deep cooperation from the community. Even the concept of community technology is the same as some concepts outside our community, so we also give them some feedback, we are using these practices.
I will also be responsible for a Shared Source Initiative project to show our interest in open source communities and respond.
What is your view to Service Component Architecture?
Devadoss: I think SCA as a response to the J2EE heavyweight nature and a container model. I also regard it as a lightweight model similar to Spring. Of course, if you come back to see our container model, it will find that the lightweight nature of the container model is what we have done a long time. I think from the viewpoint of conceptual level, SCA is not only a community from J2EE and complexity, but it will also be more integrated into our views on the world.
What is the key to XML deployment to reduce complexity?
DEVADOSS: I believe in loosely. I also believe in simplicity. I think XML is given us the ability to connect and communicate, and this is the key to service-oriented. Although the service is abstract, we still talk about it, but I think we all agree that the successful service is mainly due to XML and SOAP, so I will certainly agree with this view.
Do you think that there is a special tool to drive more users to use loose structure?
DEVADOSS: I think the key is the application model. In addition, do not use the single language model such as UML, but use models of specific areas, including developers' model, architects of business needs, architects for mapping and designing operational infrastructure. Model. I believe that this will become the only way we really realize the only way to map from business needs. Other things will happen, but in my opinion, there is no interesting and important.
Regarding SOA for many users, have we talked about the idea?
Devadoss: I think yes. There is indeed customer-oriented customers who have achieved excellent success. However, success is all customers who pay attention to business value. I have seen in the service-oriented customer is those who put IT in the first place or put the architecture first. I am in Microsoft Leadership Structure, but I first tell you that architecture is not the ultimate goal. Customers don't want SOA, they want business value. For service-oriented creation, it must discover new business opportunities. More flexible, this is worth it. When the customer tells me that they have the maturity model of SOA, I don't think it is difficult to do, because the goal is not how complicated how much maturity is your architecture, but you created what business value. Do you think many people happen in 2006 now not expected what it is?
Devadoss: One thing is a service-oriented concept will be deeply rooted. I think this is what we have to admit. I hope people coming out of the idea of the ultimate goal, but for me, more important things are the viewpoint of service consumption to increase to higher grades. I think this is what happened in 2006.
For you, what is the importance of XML hardware in SOA?
Devadoss: I believe there is a place in XML hardware. If you remember the .COM years in the late 1990s, there is such a passion in those Web promoters, and people are talking about how to apply hardware, and their essence is the process of transaction. I think it is not a high level. I think the software is a higher level.
What is the use of SOA defects need to be avoided?
DEVADOSS: People think about SOA, think about service. Significant service should have data. The influence on the data in the interconnection system, people still lack a understanding. I hope that they think more about the data behind the service. Second, I want to say that the decomposition service is very important to success. When I have talked with an ISV, they already have an integrated system and redesigned it to the face. They have 19 services, while in the middle layer, they mix all requests or for each business transaction service, and the intermediate layer processes all groups and data mixing. In fact, their system is very lack of manageability, usability. I hope that architects and developers think about service decomposition. A service is not a business object. Not a business component. A service is a greater abstract concept with data.