Business Reference Architecture: Enterprises to Consumers (B2C)

Chapter 1: Business Demand and Design Model

Abstract: This chapter discusses how to design consolidatedretail.com (it belongs to the business needs of the consumer (B2C) website, using Microsoft Business Reference Architecture), and summarizes the actual business requirements determined during this application design. This chapter also provides a 3rd floor application model and phase design process of Microsoft Solutions Framework, MSF.

Please note: Although the business needs mentioned here are only limited to the ability to install the reference example, this "Developer Guide" is very useful, through these clues, you can Upgrade to meet the needs of the production environment.

Introduction

When web users are defined for e-commerce sites, general users may answer the e-commerce site is the online store that can be purchased with a credit card. Although this definition is quite correct, it has not fully illustrated the characteristics of various e-commerce sites currently developed for Internet. In the rapidly developed Internet business era, an efficient e-commerce website is not just a web-based store.

Users are getting higher and higher for e-commerce sites. If a site can't meet their requirements, they will discard it. So what are the requirements of the user for e-commerce? The following table lists some of the main issues that affect application design.

Easy to use / navigation performance High anonymous shopping Maintenance User profile security can improve competitive advantages through a variety of device access sites through manageability

Sketchy, in the above problems, some should be resolved by the application designer, and some seem to be resolved by the enterprise decision makers or basic structural experts. However, if you think about these questions carefully, you will understand why these questions are related to the design of the application.

Easy to use / navigate

The website should be easy to use and navigate. After all, companies do not want consumers to encounter difficulties when buying their own products, and consumers are more willing to consume at the site where they can easily find the checkout page.

One way to make the site is easy to use is to ensure that you are familiar with similar methods on common tasks. This means that the goods can be stored in the shopping basket or basket before consumers complete the purchase (or "checkout"). This metaphor is convenient for people who are not familiar with the computer to understand how to work, thereby caring.

Make the site easy navigation than you first imagined. The Web is completely working in a non-linear manner, and the user clicking on the link is often unpredictable. Therefore, you should make sure that the site is always fully consistent interface regardless of the user is currently seeing, and ensures that you can access important web pages (such as homepage, shopping baskets and user account) just click on a link. Information is located, etc.). On the consolidatedretail.com site, the top flag always contains the page where the shopping basket is located, the consumer account is located and the link to the home page, and the search and directory links are always included on the panel on the left.

There is also a way to ensure that users can find the desired content in the site, which is to include a list of product lists or directories in a logical way. If you divide a directory into several categories and many possible subcategories, you can easily find the products they are interested in. In addition, users should provide search capabilities so that they can be searched when they are not very clear about the display position of some product.

If your site is easy to use and navigate, the consumer will be happy to use. Conversely, if it is more difficult to use, the consumer may discard it, and select the site.

High performance

Among the design of the website, there are many factors affecting its performance. Since different people have different requirements, it is also acceptable to person.

Try to minimize response time

Most people think that the site for providing acceptable response time is a good performance site. The response time refers to the amount of time that the user can wait for the time before requesting a certain operation. Ideally, we all want the operation on the site to be executed instantly; Therefore, when designing an e-commerce site, we should minimize factors that have a negative impact on the response time (although they cannot completely exclude them). The key to e-commerce optimization is to reduce the time consumed to perform operations such as checkout, so that consumers will not give up their purchase of goods because of the queue waiting, you will not lose orders.

Try to enhance scalability

Another important aspect of performance is "scalability." This refers to the ability to add a resource to increase the capacity of the resource. From a user point of view, this means that the site can provide an acceptable response time when a large number of users access the site at the same time. Many developers often get the frustrating news: When the user reaches a certain amount (this quantity is the number of actual life requirements), the performance excellent test site can not be dealt with in the development machine.

So how can we maximize the scalability of the site? Two typical methods are "upward expansion" and "outward expansion".

Expand upward

The first method ("upward expansion") is to enhance the processing power of the server by using a better and / or faster CPU, a larger RAM, faster disk, and the like. This method is very effective, especially on the data layer, some large databases on this layer require relatively strong processing capabilities. However, due to the increase in hardware costs with the processing power, the closer the server, the closer to the top, and this method is more inclusive.

Outward extension

"Outward" is solved from another aspect, namely the "cluster" (or server collection, also known as the "Web field") to share processing workload. The Web field costs more costly, and provides more flexible, scalable solutions. When the load on the site increases, you can easily add the server into the web domain.

Microsoft® Windows® 2000 Advanced Server and Windows Network Load Balanter Server, Windows Network Load Balancing, NLB services, display the entire Web field as a logical server with a single IP on the Internet. After receiving the request, the request is distributed to the server according to the load situation, which can communicate with the backbone network or communicate with the database server. Figure 1-1 shows the basic architecture of the web domain.

Figure 1-1: Web field

Manage the status in the web domain

One of the most important issues is one of the most important issues in the web domain. The status is the session data that must be retained between two user requests; for example, during the user's continued browsing site, it is necessary to maintain the original article in the user's shopping basket. This is necessary, even if each user request may be processed by different servers in the Web field.

Many ASP developers use the Session object to store status data. However, this method should usually avoid using this method. To optimize the software architecture of the site to implement in the server domain, the web front end prohibits maintaining user status in memory. If the front-end server maintains user status, the following issues will appear:

The user session will be included in a particular server (session dependen), which will destroy the request to the server to the server's network load balancing policy. In addition, the reliability of the server domain will also be damaged because the user session cannot be transferred to other servers when the original server fails (and lost session status information in its memory). Memory resources are consumed by the front-end server in the details of the storage user session state, thereby reducing memory available to process requests and cache content. If a popular site can attract a large number of users in a short period of time, the memory demand in status maintenance may be very large. For partial resolution of memory requirements, Commerce Server uses a high speed cache. Cache will be made to the configuration file architecture, discounts, and business activities. In addition to avoiding session correlation, avoiding the front end operation with a long-term operation to be associated to design the front end operation as a fast execution. Since IIS is to handle requests with a buffer pool, the buffer pool contains the number of worker threads thus limited, and thus when these threads have been occupied and the average time of the incoming request is taken when the operation is waiting for a long time. It will increase. Anonymous shopping (browse)

Usually, users are reluctant to log in to the site for forced to log in to the site for the sale of sites. Therefore, the site should allow users to browse goods in anonymous manner without authentication, and even allow them to put some of the items into the basket.

Maintain user profile

When the user accesses the site again, they do not want to re-enter the same information entered when the last visit is entered. Once you provide your own shopping and contact information to your site, users want the site to remember this data.

To achieve this, many sites maintain their user "configuration file" information for each registered user. In most cases, users need to register in order to provide minimum amount of profile information such as user names and passwords. The user then assigns a unique identifier that can be used as the master key of its configuration file data.

After the user is registered on the site, its configuration file information can be saved in the database to call when needed later. Typically, users can add some necessary information, specify some details, such as email addresses, phone numbers, delivery addresses, or any other personal information that allows users to add.

Keep the user profile information is quite useful, whose reasons are as follows:

You do not have to re-enter the data when accessing later. Can be used to analyze the activities of users at sites. Can be used as a personalized basis, allowing you to publish a flag advertisement or discounted activity based on a particular user group. It can be used for business analysis, such as tracking the purchase trend based on a particular profile value.

Improve competitive advantage through manageability

Although application designers are not responsible for business decisions (such as pricing, advertising activities, etc.), the design of e-commerce solutions has a huge impact on how companies deal with market trends and competitors. The management activities carried out by the business manager are subject to the management function of e-commerce site. To succeed, e-commerce solutions must be easy to use, and must have a comprehensive management infrastructure.

There are two basic choices for the design management interface for e-commerce sites. You can create your own custom interface, or you can use a "ready" solution such as Microsoft Commerce Server 2000 Business Desk.

If you build your own management interface, you will be able to design the site management functionality in accordance with your own wishes. However, this will increase a large number of software projects that have been largely developed, which is almost equal to or greater than the workload of the software project itself. By default, Commerce Server Business Desk can meet most of the management requirements of e-commerce sites, if you need to add other features by creating custom modules.

The rest of this chapter will illustrate the actual business needs confirmed during the planning phase of the project, as well as the application models and design processes used in the design of the consolidateDretail.com application. Reference application business needs

Before designing applications, you should clarify which tasks must be performed. Analysis Business Demand is one of the most important steps in application development. Confirm that the purpose of business needs is to create a solution that can meet retailers and consumers. In this way, the demand is converted into a business demand document, which can be used as a guide to the development of the entire project.

This section summarizes the actual demand determined for the reference architecture application consolidatedretail.com. Please note: Business requirements used here are intended to be capable of having a relaxing reference example.

Functional Requirements

ConsolidateDretail.com is designed to meet the following functional requirements:

Easy navigation

The site should be easy to navigate. The link should be clear, easy to understand and practical. Users should be able to move freely between pages and screens.

Easy to use

Applications should be easy to use. You should easily purchase the product and access the "Checkout" page.

The site should use an easy-to-understand metaphor, for example: store the optional item in "Shopping Basket" until the shopper is ready to checkout

Each page on the site should appear exact interface. Important pages or common pages should be accessed only by clicking once.

Availability test

The site should make people who are not familiar with the computer easily understand.

Site Access

Users can access the site by the following methods:

Enter the URL from other sites or email links to links to access

Maintain user registration / configuration file

Regardless of any page from the site, the user must be able to register, so that users do not have to re-enter the same information every time you order. Users can browse the site without registration; but must be registered when checkout. In addition, request registration when applying for email news, special notices, etc.

Registration involves:

Profile information: User name, payment address, main delivery address, phone number and email address. Authentication information: User identity (user ID) and password should be kept in the application. Payment information: Users should enter credit card information and save this information. Applications should be able to save multiple credit card numbers. Preferences: Users should be able to specify whether to get email notifications about delivery status (the default value is "yes"), and whether to get notifications for sales prices and specials (the default is "No"). Address Book: The user should be able to store any plurality of additional delivery addresses.

Keep the user profile information is quite useful, whose reasons are as follows:

You do not have to re-enter the data when accessing later. Can be used as a personalized basis, allowing you to publish a flag advertisement or discounted activity based on a particular user group. Can be used for business analysis, for example, tracking the purchase trend based on a particular profile value.

User registration management

After the user logs in and after authentication, the user should be able to modify, add, or delete registration information. In addition to the User ID field, all other fields should be an editable field.

Login / Authentication

After the user returns to the site after registration, he or she should be able to log in from any page from the site.

Browse

Users should be able to browse the directory. On the home page, the directory list should be displayed to the user. After the user selects a directory, it should display the subcategory or actual product.

Anonymously browse

Users should be able to browse the directory in anonymous manner; ie: the user should be able to view the product without having to log in.

Multi-directory

Applications should support multi-directory. Summary of multi-directory products should be transparent to users.

Products and categories

The application should allow the product to be associated with one or more directories.

Product page

Applications should have a product page, including larger pictures of the product project and / or detailed description of the product item. On this page, the product should be added to the shopping basket. On this page, the user should be able to:

Add product items to your shopping basket to browse the next project to browse the last item Back to the previous page

product search

The homepage and all types of pages and subcategory pages should be available. Users should be able to enter multiple words. If a user specifies multiple words, a Boolean query using the "AND" operator will be constructed based on these words. If the user is on the home page, search will default to "search all categories". Searches executed in category and subclass page will default to "search in the 'category' range". The user can select a particular site area or a particular category to search to overwrite these default settings.

If the site uses a multi-directory, you will perform a search for all directories. If the site shows multiple directories (and there is a layered product list), no search is made in this rule. In the category / product hierarchical structure, each directory is the first level. In this case, it is default that only the current directory where the user is currently search. Users can override the default settings, select some other directories or the entire site. This is similar to the behavior specified by the previously described "Multi-Directory".

By default, you will search for keywords and titles.

Product search results

"Search Results" page should display a series of product projects and their corresponding categories (or directory). The project should be packet according to the category or directory. Each search result should be provided to a hypertext link to the corresponding product page.

Add a project to the shopping basket

Regardless of the product page, users should be able to add one or more items to the shopping basket. These items can come from different directories. Each time you add a project, the number of items in the basket will also increase accordingly. The number is displayed next to the basket icon.

Manage shopping basket

Users should be able to manage shopping baskets at any time. Users can specify a project "Active" (actually purchased tag) or "reserved" (identified as a future possible purchase). Users can check the following options when shopping baskets:

Delete a single item. Changing the number of each item. Retain any items for future purchase. Delete all items in the shopping basket. Keep all items in the shopping basket for future purchase. Move the project into the shopping basket reservation (future purchase) area and from the middle. Retrieve the reserved order.

Keep shopping basket or project

Users should be able to retain all items in the selected item or shopping basket for future purchase. Only users who have registered and logged in can retain their items. If the user has not logged in or registered, it will be prompted to do this. After the user completes this, returns to "reserved shopping basket" operation.

Checkout

No matter from any screen, the user should be able to check out. When checkout, all ordered items (shopping basket) will be displayed to the user. At this point, the user should be able to manage the shopping basket. After confirming the items in the shopping basket, the "shipping" screen will appear. Each project will be associated with the main delivery address of the user. Users can replace this address with an address or new address in an address book. If the user adds a new address, he or she can choose to save the new address in the address book.

After the user assigns an address (or accepts the default address) for each project, he or she can go to the "shipping" screen, select the delivery method of each address. The default mode is determined by the site owner. After the user chooses the delivery method, he or she can continue to the "Order List" screen. This screen should be divided according to shipping address. Under each address, item descriptions, project prices, and pricing are listed (if). The price of the project is subjected to a small count, and the shipping cost is listed as a detailed item and submitted a small count, and finally the tax and total amount under this address will be listed.

After the amount under all addresses, the total is listed at the end of the page. Users can:

Accept order modification order cancel order Continue shopping

If the user selects a modification order, he will return to the Manage Shopping Basket page. If the user chooses to cancel the order, the shopping basket will be empty. If the user chooses to continue shopping, he or she should return to the home page.

If the user chooses to accept orders, he will go to the "Payment" page. This information is displayed if the user has stored credit card information in the Registration page. Users can choose to use saved credit cards, or select ignore saved information and provide new credit card information. If the user adds a new credit card message, he or she should choose to add new information to the saved registration information. After users choose or enter credit card information, he or she can:

Cancel order modification order Continue shopping submission order

If the user submits an order, the confirmation page and the order number will be received.

Shipping choice

The following shipments must be supported:

Shipping Hong Kong ground delivery the next day delivery overnight delivery international delivery

Order status notification

Users can choose to receive email notifications about the order status.

Calculation of shipping costs

Calculation of shipping costs Based on the type of carrier (such as UPS) or other rules set by the site owner.

Tax calculation

The calculation of the tax must be based on the rules set by the site owner. These rules should include:

Sales place delivery address cargo type

At the time of checkout, the tax information will be displayed on the Order List.

Order list

The screen displays the address, project description, project price, shipping cost, tax and cost total count (if any).

address book

The registered users are saved in the address book. Although the site owner can set some restrictions, the address book can still store unlimited shipping address information.

Order cancellation

The user must be able to cancel the order at any time before submitting the order. This action will cause all items in the shopping basket to be emptied. But the reserved project will not be affected.

system requirement

Site must meet the needs of the following system:

Globalization

Applications should be able to customize to accommodate different cultural environments. That is: interface color, navigation layout, page structure, and language should be modified.

performance

Users should experience consistent performance at each access to this site. The performance of the site should be as good as other enterprise e-commerce applications being used.

Scalability

Site should expand the extension overhead. If a faster disk and the CPU are added, the response should be faster if it adds a larger RAM. If more servers have been added to the Web field, the response should also be improved. The server in the web domain should be able to properly handle the request.

Usability

The site should be in turn on and run, and there should be no fault. It should be able to capture errors, this feature should not prevent users from accessing the site authorized area. Site should be accessible to users at any time.

Manageability

There should be a management interface on the site, used to modify and manage company reports, directories, orders, shipping costs, tax rates, and user accounts.

safety

Site should protect confidential information, such as credit card numbers. Site should display confidential policies and any related copyright information. User IDs and passwords should prevent unauthorized personnel from accessing sensitive information.

Allow access to a variety of devices

Site must be able to run normally on a variety of client devices. Site should work on a low version of browser and high versions of browsers.

Record business needs in document form

After determining the basic demand, you should capture, transfer, and ratify these needs in the "Content / Range" document, which identifies the application business value, demand and restrictions and the person needed to plan, design, and complete the project. After that, you can start designing.

The next section describes the application design models and design procedures used in the creation process of consolidateDretail.com.

MSF application model

The consolidatedretail.com application is designed to follow the three-layer model defined in the Microsoft Solution Frame (MSF). This model divides the service provided by the application into three abstraction layers so that the application obtained has certain flexibility and scalability. Any layer can be changed and will not have a negative impact on other two layers so that applications will be able to improve applications to meet user needs and technological changes.

These three layers are:

Represents Service: The application's representation service is used to render the data displayed to the user and accept the user's input. Business Services: Sometimes called "Application Services", the application's business service enforces business rules. In a typical e-commerce application, this may include: ensuring that the user must authenticate before the order, retrieve the corresponding content according to the user's configuration file, the check is executed in the correct order in the correct order of. Data Service: Data services in your application include logic required to store, retrieve, and modify data, and data integrity rules for applications must be enforced. In e-commerce applications, this may include processing of directory, user, and order data. Note: For more information on MSF, please visit http://www.microsoft.com/msf (English).

Why use the MSF application model?

In addition to the advantages described above, the MSF three-layer application model has significant advantages in reducing development, deployment, and managing applications. The main advantage of using the MSF three-layer mode in the application architecture is reflected in:

Separation: Since the service is separated from each other, each layer of the application can be developed independently of other two layers. In this way, three different development teams work on the same application project. Distribution: Since the logical layer is independent, they can be deployed on multiple servers in a distributed manner. Reuse: Different client devices can use the services provided by each layer. For example, a business service of an e-commerce site can be used by a set of services to provide an HTML interface to the website; or a mobile phone for supporting WAP can also be represented by another group. Some business services can also be configured as a web service in a variety of industries (LOB) applications or trade partners.

In addition to following the three-layer design model, consolidatedretail.com's designers and developers have followed the MSF design process. This process is described below.

MSF application design process

Regardless of the type of effective application, the first step is to ensure reasonable design. There are many ways for software design, and each has its own advantages and disadvantages. When deciding the design process to be used, ensure that the process provides a clear stage that is divided according to the actual software implementation step; ensuring that the process can be adjusted according to the demand variations of each stage.

The MSF application design model not only defines a service-based architecture not only for distributed applications, but also defines a loop design process. In each round of loops, the software is in the "concept" phase, "logical" phase, " The order of the physical "stage is designed.

Conceptual Phase: In the conceptual phase, it will be seen from the challenge from the perspective of users and / or business. The main goal of the concept stage is to define the concept of challenges and solutions. The "Function Manual" document is the final result of the concept phase. Logical phase: In the logical phase, design goals and challenges will be viewed from the perspective of the project development team. The main goal of the logical phase is to map the concept design to a logical component. The group uses the user program (or user case) identified in the concept phase to build the logical model of each component in an object-oriented software solution. Object-oriented solutions load business functions into the software representation of the actual object, which is defined by the "class" in the object-oriented design. Physical phase: In the physical stage, the goals and challenges will be seen from the perspective of developers. The primary objective of the physical phase is to apply logic design to actual needs and constraints. Since the physical stage is a design problem from the perspective of developers, and the task is to define a physical implementation, the results of the physical design phase are technical specifications documents.

to sum up

This chapter contains general overview of e-commerce needs and summarizes specific needs of consolidateDretail.com applications. This chapter also provides a summary of the MSF application model and design process. The next chapter will discuss "concept", "logical" and "physical" design phases and final results of various phases. Chapter 2: Conceptual Design Stage

Summary: This chapter will explain the conceptual phase of the Microsoft Solutions Frame (MSF) design process. At this stage, the design team will determine the main solution concepts and targets from the perspective of potential users and services. At the end of this phase, the design team will write function description documents, which is the basis for all application design work.

Introduction

In the conceptual design phase, the design team will determine a complete project front view according to the previously determined business and user needs and record it. Based on these needs, the team will prepare "Applications" (derived from "Samples"), then create a function description document to explain the working principle of the application from the user and supplier of e-commerce applications. This function description document is the document submitted at the end of the concept phase.

Application plan and use case

The application is created by detailed analysis of "use case". The use case is only an explanatory text that describes the interaction between the external "operator" (operator can be a user or existing system) and the designed application (or component). In addition to the user, the operator can also be a service, component, and the like.

When creating the usage case, two aspects should be determined: First, the operator's operation, the other is the expected result. The use case is usually discussed in a whiteboard (also known as the collective discussion), which will briefly summarize the use case, and provide a descriptive name. The details of the use case will then be developed and some linear sorting is performed.

The rest of this chapter will introduce a variety of applications developed as consolidateDretail.com applications, and provide an example of operational examples of the contents of various options.

Application plan reference architecture

In the conceptual design phase, the designers of the reference architecture application will determine the following application:

Applications 1: Customers log in to the site

Customers navigate to the login page and enter its username and password. The system will verify the username and password. If the user enters a valid username and password, a page is displayed, and the link it contains users to update their user profile information and view their order history (as described in "Application 6").

If the customer stays empty username field, it is prompted that the user must fill in the username field.

If the customer leaves the password field empty, it is prompted that the user must fill in the password field.

If the username entered by the customer is invalid, the login page will be redisplayed and an error message indicates that the username is invalid.

If the username entered by the customer is valid and the password is invalid, the login page will be redisplayed, and an error message will indicate that the password is invalid.

If the customer means that the password has been forgotten, an option will be provided to enable it to enter the user name and then receive a password in the email message.

Figure 2-1 is an example of an example to illustrate the login function.

Figure 2-1: Login function

Applications 2: Customer Search Products

Customers can search for products by entering the product's title, text, or instructions, the system will display all matching products, including the user's index key input in the search box.

Customers can navigate to the search box, search the product by entering a portion of the product name. The system will display the most matched products with the search criteria.

Figure 2-2 is a use case pattern to display a search executed by the customer.

Figure 2-2: Customer Executive Search

If the product name entered by the customer does not match the product, or the product name entered by the customer is incorrect, the search results will display a message, indicating that there is no product matching with the search criteria.

If the keyword entered by the customer does not match the product instructions, the system will display a message indicating that there is no product to match the search criteria. Applications 3: Customers browse the products you want to buy

In this case, the customer selects the directory you want to browse. Displays the various categories and products in the root of the selected directory. The user can then select the product you want to view the details, or select a category to see the product set and subcategories in the selected category.

Figure 2-3 is a use case pattern for explaining a browsing scheme.

Figure 2-3: Customer browsing

Applications 4: Customer Management Shopping Basket

When the customer displays the shopping basket, the shopping basket may be empty, or it may be included in this session or not purchased or deleted in the last visit.

If the shopping basket is empty, the app will display a message to prompt the customer to the shopping basket is empty.

If the shopping basket contains one or more products, the app will display a list, including the product name, price, quantity and total of each product, and total products in the shopping basket.

If the customer chooses to delete some product, the application will delete the product from the shopping basket.

If the customer selects all products in the shopping basket, the app will delete all the products in the shopping basket.

If the customer changes the number of a product, the application will update the shopping basket to reflect the number of this product.

If the customer updates the shopping basket, the app will recalculate the number and total number of products for each product in the shopping basket, and recall the total of all products in the shopping basket.

Figure 2-4 is a use case pattern for explaining shopping basket management scheme:

Figure 2-4: Customer Management Shopping Basket

Applications 5: Customer Checkout

Customers say they are willing to check out and willing to buy products in shopping baskets.

If the customer is currently not logged in, the customer will be required to enter a valid username and password according to "Application Scheme 1". Then, the application will display the list of delivery addresses associated with the current user. Customers can specify one of the addresses or edit the address or add new addresses. Customers can also specify to ship different products in the shopping basket to different addresses.

If the customer specifies a single shipping address, the application will prompt the customer to select a delivery method from the available delivery method list.

If the customer specifies the use of multiple addresses, the application will prompt the customer to select the address and delivery mode for each product in the shopping basket.

If the customer selects an add or edit address, the application redirects the customer to one of the delivery information pages described in Application Scheme 6.

When the customer provides the delivery address and delivery method information, the application will prompt the customer to confirm the payment address and provide credit card type, account holder name, account number, expiration month, and expiration year.

If the customer does not specify a payment address, the application will display a page that allows the customer to specify the payment address according to the Application Scenario 6.

Figure 2-5 is a use case pattern to illustrate the checkout function.

Figure 2-5: Customer checkout

Next, the app will display an order, delivery, and payment information page. Customers can confirm the order or modify the information.

If the customer selects a modification order, it will return to the shopping basket page described in Application Scheme 4.

If the customer confirms the order, the thank you message and the only order number will be displayed.

Finally, the application will send an order to the user to confirm email to confirm the order.

Applications 6: Customers provide accounts and delivery information

If the customer is browsing this site in an anonymous manner, you must enter the username and password according to "Application Scheme 1".

The "My Account" page contains links to multiple pages to facilitate customer changes account information, change delivery address information, change payment address information, change your password, and view order history.

If the customer selects changes its account information, the application will display a page, let the customer edit the user name, email address, phone number, and fax number of its account.

If the customer selects editing address information, the application will display a page, list all the addresses associated with the account. Customers can add new addresses or edit or delete existing addresses. If the customer chooses to add a new address, the application will display a page, prompting the user to provide the name, recipient name, block address information (two lines), city, province, postal coding, and phone number.

If the customer will "listed as", "Recipient Name", "Address Row 1", "City", "province", "Postal Code" fields, the application will prompt the customer to fill in the missing data. If the customer selects the editing address, the application will display the address field, which can be modified. If the customer selects the delete address, the application will delete the address of the account.

If the customer chooses to change its payment address, the application will display a page, prompting the user to provide the name, recipient name, block address information (two lines) of this specific address item, city, province, postal code, and phone number.

If the "Recipient Name", "Address Row 1", "City", "province", "postal code", or "listed as" is left empty, the application will prompt the customer to fill in the missing data.

If the customer selects to view its order history, the application will display a page, which contains details of the customer's previous order.

If the customer selects changes its password, the application will display a page, prompt the customer to enter the old password and the new password, and confirm the new password. When the user submits these data, the application will check if the old password matches the password in the account, and the new password matches the password in the confirmation information. If the password matches all, the system will change the password associated with the customer. If you do not match, the application will display an error message, prompting the customer to re-change the password.

Figure 2-6 is a use case pattern for explaining this scheme:

Figure 2-6: Customers provide accounts and delivery information

to sum up

This chapter describes the process of identifying an application plan for e-commerce applications and developing a conceptual design. Figure 2-7 is a use case pattern to display the defined accumulated application:

Figure 2-7: Accumulation application design

Applications and their base cases should indicate the expected behavior of the application in a variety of business processes. Applications should also illustrate the functionality of the application, and provide the basis for the logical design of components in the application. In addition, because the design process has iteration, if the previously unrecognized interaction is found during the development process, a new use case can be included in the design.

Chapter 3: Logical Design Stage

Microsoft Corporation May 2001

Summary: This chapter describes the logical phase of the Microsoft Solution Frame (MSF) design process. At this stage, the design team should have a high-level decision based on the conceptual design and integration of components. The design team will use the application solution (developed in concept phase) to build the logical model of the application.

At this stage, the design team should have developed a logic design, which is the basis of the actual design phase of the project.

Introduction

The target of the logical phase is to convert the functions defined in the concept phase to an abstract model to determine the collaborative logic components that will be used to support the solution.

The resulting logic design does not involve specific technologies. Instead, the goal of this phase is to analyze and understand functionality before any technical work. For example, when a group is designed an e-commerce solution in a logical phase, it may be considered that the component named Users is a necessary service to track the user group accessing the application. But in the actual design phase, the design team may choose to use Microsoft® Commerce Server 2000. In this case, the MSCSProfileService component in Commerce Server provides the actual implementation of the USERS component.

If the final application design includes custom components (ie, the components that are not available in the ready-to-range solution or product), the corresponding components determined by the logical phase can be directly converted to the actual phase. For example, a UserS object is defined at a logical phase, and the design team decides to make it a custom object, then the UserS object will be repeated in the actual phase. The rest of this chapter will outline the logical design flow used when designing consolidateDretail.com, and then detail the logical components required to meet the requirements required.

Create logic design

The first step in creating a logical application design is to determine the business object (component) that will provide the desired function. In addition to determining the desired object, the design team must also determine the behavior, attributes, and relationships of each object. The design team will use the application in the concept phase to determine these objects and their relationships, behaviors, and properties.

For example, the following is the application 3:

Users choose the directory you want to browse. Displays the various categories and products in the root of the selected directory. The user can then select the product you want to view its details, or select a category to view each product and each subcategory in the selected category.

In this way, the design team will determine all aspects of the support solution by analyzing this scheme, and perform the following tasks:

Determine the business object in this scenario. Determine the behavior of these objects. Determine the properties or features of these objects. Determine the logical relationship between these objects.

These tasks will be described in more detail in the back section of this chapter.

When these tasks of each application are completed and the archive, the design team can end the logical design phase. For a complete design example of consolidateDretail.com, please refer to the "Dixed Logic Design" section.

Unified modeling language

Unified Modeling Language (UML) is a tool for describing system operation. UML is a very useful tool in an intuitive description of the system to make more comprehensive analysis. By using UML, it is convenient to explain the components, interaction, relationships, and more.

UML is usually used to simplify design in the logical phase.

To illustrate the tasks involved in the creation logic design, the following sections will provide a simple UML diagram representation example.

Obser

When analyzing the application, the primary task is to determine the objects thereof. Objects are usually a business entity or process that appears in an application. For example, in the following paragraphs, the object will be identified in a bold:

User selects the Catalog (directory) you want to browse. Displays Categories and Products in the root location of the selected Catalog (directory). Then, USER can select the product to view the details of its details, or select a Category (category) to see the Products in the selected Category (each product) and subcategories.

The above example uses the following objects:

User Catalog Categories Product Products

Figure 3-1 is a UML diagram depicting the object determined in this example:

Figure 3-1: Object

These five objects become the basic objects of this program; however, in some cases, to make the program work, other objects need to be added, even if these objects are not explicitly listed in the scenario. These additional objects can be determined by checking the acts associated with its associated behavior. To determine these objects, you must first determine its behavior.

Determine behavior

When the obvious object group is determined, the next step is to determine its respective behavior, which is also known as "method" or "service".

To determine the object behavior, you must first determine the operations executed in the scenario. For example, in the following paragraphs, the action will be identified in a bold: the user selects the directory you want to browse. Displays the various categories and products in the root of the selected directory. The user can then select the product you want to view its details, or select a category to view each product and subcategory in the selected category.

The first action is the user selection directory. Figure 3-2 is a UML diagram that describes the user object as a SELECT CATALOG behavior:

Figure 3-2 Behavior of the User object

As mentioned earlier, there must be no obvious object from the scheme with its associated behavior. We can thus think that since the user selects the directory, there is inevitable mechanism that allows you to select a directory from the directory list. This way, you can logically assume that the Catalogs object exists, which manages all Catalog objects. Then, this new object should be added to the list of defined objects.

When defining a "Catalogs" object, you can define the first action as a SELECT CATALOG (selecting a directory), which belongs to the "Catalogs" object.

You can continue to analyze each sentence in the scheme until all conscription objects and their associated behavior are determined.

Determine attribute

When the behavior is determined, the next step is to determine the properties of the defined object (also known as "characteristic"). Attributes are elements that need to be tracked. They are placeholders for retention (or "stay") data.

The property derived method is that the behavior in the analysis scheme and extracts elements that need to remain (or track). For example, in the previous section, the application specifies that the user will be able to view the product. When viewed, the elements displayed to the user will become the properties of the product. For example, if a business requires a product specifications and prices, these elements will become the properties of the object.

Figure 3-3 is a UML diagram showing a User object with an NAME attribute:

Figure 3-3: Properties of the User object

Determine relationship

After determining the object and its properties, the next step is to determine "relationship". The relationship is the logical association between objects.

To determine the relationship, you must analyze the interaction between objects. For example, the "categories" object has a relationship with the "category" object, because the "Categories" object manages the collection of "category" objects and contains the "category" object.

Be sure to pay attention to another relationship called Inheritance, which deals with an object being defined by another object. For example, if the designed solution is to sell food and books, it is also necessary to distinguish between the two, then define such a relationship: "Book" and "Food" objects belong to a "product" Object. In this way, they can inherit "Product" objects.

In a business reference architecture solution, no inheritance relationship is not defined in a logical phase. But in some e-commerce solutions, these relationships may be very important.

Completed logic design

At the reference architecture, the logical design phase of consolidatedretail.com, the design team determines the following objects that form an abstract collection of components required for support solutions. (These objects are arranged in alphabetical order instead of being arranged in order.)

AUTHENTICATION

Authentication object handles users' registration and authentication.

Catalog

The Catalog object stores information about a specific directory and manages a collection of products in this directory.

Catalog Manager

The Collection of the Catalog Manager object management directory.

Category

The Category object stays information about specific categories. Category Manager

Category Manager Object Management Category Collection.

CONFIGURATION

The Configuration object holds the application configuration information and handles the task associated with the configuration.

Data functions

The Data Functions object performs data-specific functions such as open access to the database.

E-mail

E-mail objects are used to send email messages to users, such as order confirmation messages.

Error Handler

The "ERROR HANDLER" object will be called when an error occurs. It handles user-friendly error conversions and log errors.

ORDER

The ORDER object stays information about a particular order.

PRODUCT

The Product object stays information about a particular product.

Search

Search objects are used to search for directorys and return product search results.

User

User objects are used to reserve information about specific users. In addition, it manages all user orders.

User manager

User Manager object management user collection.

Figure 3-4 demonstrates the relationship between the main object and the use case described in the second chapter

Figure 3-4: Object Relationship

to sum up

This chapter illustrates four-step procedures that determine objects (ie, business components) and their properties, behavioral and interrelationships of e-commerce applications. The end result of this process is to create logic designs that will be used as technical design and standardization.

It should be noted that the generated logic design does not involve specific technologies. These technologies will be determined in the actual design phase, which is the topic discussed in the next chapter.

Chapter 4: Physical Design Stage

Summary: This chapter discusses the physical design phase of the Microsoft Solutions Framework (MSF) related to the business reference architecture application consolidatedretail.com. In the logical design phase, the project development team applies the actual physical design constraint to the logic design created in the logical design phase. The goal of this activity is to identify a set of components and then determine which components already exist and which components must be created. At the end of this stage, the results recorded the results in a clearly defined technical specifications, which will become a blueprint for building applications.

Introduction

The physical design phase is to apply the actual physical design constraint to the transition phase of the logical design. After identifying logical components, the next task is to analyze which components already exist, which components can be reused or modified, and which components must be created.

As mentioned above, the physical design process is considered from the perspective of developers. The result of this phase is a design or blueprint for a complete implementation, while writing a technical specification document, and the development team will use this document to build an application.

Physical stages can be divided into three smaller task stages, as shown below:

Study: At this stage, the development team will determine the physical infrastructure constraint and solution requirements, and process the conflict between the two. In addition, the development team will also determine expected implementation techniques. Analysis / Rationalization: The Development Team will select the implementation technology to use and determine how to meet the defined business needs. Implementation: The development team will select a programming model, specify the component interface and select the development language.

The rest of this chapter will discuss these three tasks in detail, and examples are given in the appropriate place.

research work

The first task involved in the physical stage is to study and collect information about the following topics:

Physical solution demand physical constraints available available

Customer demand is contained in a demand document, further defined in a logical phase; however, the development team may need to study and determine actual constraints and existing technologies.

Determine physical solution requirements

Physical solution demand is dedicated to guiding the need for infrastructure design. In Chapter 1, we define the following system requirements:

Globalized performance / reliability scalability availability can manageable security accessibility

The following sections will be detailed for each topic.

Globalization (international)

Globalization (or international) is a process that needs to pass the core content of the development program. In this process, it is no longer based on a single language or regional design and code design, and the prepared source code is more convenient to create a program. Different language versions. Globalization allows you to transplant your application to different cultural environments. In early programming, this only means supporting multiple languages ​​(for example, supporting Unicode), but now you have to choose what kind of interface, such as determining color, navigation layout, and page structure.

When conducting globalization, there is a need to carefully review some of the well-known geographic and cultural issues involved in the application or web page. The steps to globalize include: studying language and cultural issues, requiring language experts to verify some well-known issues, if possible, companies in specific sales areas representatives to verify some well-known issues.

To support these cultural differences, you can define the following physical needs:

The ability to use nvarchar in the database without using varchar to provide custom interface

performance

Performance is generally measured with "system total throughput" and "response time".

System total throughput

The total throughput of the system uses the "number of transaction per second" to measure, reflecting the ability of the system to perform a particular collection (called transaction) of the service request. For e-commerce applications, transactions may be executed in turn until the end of events:

The user came to the site. Users browse the directory and find the desired product. Users add the product to the shopping cart. User registration. User checkout.

TPS is the maximum number of systems that can handle these transactions per second. As described in business needs, when using the following development configuration, the business reference architecture application is designed to handle 4800 such transactions per hour:

[(4) PIII 500MHz, 1GB RAM, server runs IIS and Commerce Server] [(1) PIII 500MHz, 1GB RAM, server runs SQL Server]

Response time

The response time is the amount of time between the user request and the system response, which is the user's most concerned performance indicator. Response time is usually represented by a percentage and response seconds. For example, "90% of all requests should be within 5 seconds" means that 90% of its requests must get the server within 5 seconds before the user believes that the application is running.

The business reference architecture application requires 95% of all requests within 5 seconds.

Scalability

Scalability refers to the ability to increase site capacity when adding resources. From a user point of view, this means that the site can provide an acceptable response time when a large number of users access the site at the same time.

We have already mentioned in the previous chapters, and there are two ways to increase scalability: "Upward extension" and "outward extension".

Expand upward

"Up" is to enhance the processing power of the server by using a better and / or faster CPU, a larger RAM, faster disk, and the like. This method is very effective, especially on the data layer, some large databases on this layer require relatively strong processing capabilities. However, due to the increase in hardware costs with the processing power, the closer the server, the closer to the top, and this method is more inclusive.

Outward extension

"Outward Extension" refers to a plurality of servers that use a cluster (also known as "web field") to share processing workload. The Web field costs more costly, and provides more flexible, scalable solutions. When the load on the site increases, you can easily add the server into the web domain.

To enable extension extensions, you must avoid using server-specific session memory (such as SESSION objects in ASP) to reserve information. The reason is as follows:

The user session will be included in a particular server (session dependen), which will destroy the request to the server to the server's network load balancing policy. In addition, the reliability of the server domain will also be damaged because the user session cannot be transferred to other servers when the original server fails (and lost session status information in its memory). Memory resources are consumed by the front-end server in the details of the storage user session state, thereby reducing memory available to process requests and cache content. If a popular site can attract a large number of users in a short period of time, the memory demand in status maintenance may be very large. For partial resolution of memory requirements, Commerce Server uses a high speed cache. Cache will be made to the configuration file architecture, discounts, and business activities. Usability

Availability refers to the ability to connect and use resources in a timely manner at any time.

One way to understand high availability is to compare it with "fault". These terms describe two different references for measurement availability. "Fault" is defined as 100% availability within 100% time (where the environment is in place). The design purpose of the fault tolerant system is to "ensure" availability of resources.

High availability resources are almost always operating in operation for clients and are accessible. Therefore, it does not have a single point of failure. Server clusters and network load balancing is two ways to keep system resources available.

Before deploying the site, the following methods should be used to prevent the server from malfunction:

Geographic location data center. Uninterrupted dual power supply. data backup. The server of the cluster is equivalent to the role of a single server. Data replication. Network load balancing, that is, multiple identical server assumers to ensure availability, scalability, and exact user experience.

Manageability

Manageability is the ability to perform site management tasks. For e-commerce applications, it includes the product catalog, special promotions, shipping costs, tax rates, user accounts, and the use of sites, trends provide reporting mechanisms, and more.

If you have a comprehensive management infrastructure, the business manager can configure the site to adopt a corresponding countermeasure based on the market trend and competitors.

safety

If the security of the most basic form is ensured, it ensures that data or devices are protected to prevent unauthorized people from accessing or use them. In an e-commerce application environment, you should protect the following information:

Sensitive user information credit card number unapproved product data

The design of application security mainly includes three aspects: "Authentication", "Authorization" and "Encryption".

Authentication

There are two main methods to be used in user authentication in distributed solutions (eg, e-commerce sites). It is generally described with the "counterfeit / delegate" model and "trustee server" model.

Both models assume that the N layer application is used. In this example, the user is connected to the intermediate layer (specifically the web domain), the web domain sequentially access the rear end layer (specifically the SQL Server database) or services. The difference between these two methods is to use the security account to access backend data.

The "counterfeit / delegate" model is in the "counterfeit / delegate" model, the user provides secure credentials to the intermediate layer application, and then the intermediate layer application uses the user's secure credentials to access the backend database. The intermediate layer application is essentially in the "counterfeit" user, represents the user to retrieve data. Figure 4-1 illustrates the "counterfeit / delegate" model:

Figure 4-1: "Counterfeit / Entrusted" model

The "Trustee" model is verified by the user in the "Trustee" model, and the intermediate layer application is authenticated, usually a combination of check user names and passwords. The intermediate layer application believes that the user's identity is correct, it uses the "own security account" to access the backend database. In addition to the intermediate layer applications, users have no right to access backend data. In this method, there are actually two authentication operations. First, the web application authenticates the user, and the database server authenticates the web application. Figure 4-2 illustrates the "Trustee" model: Figure 4-2: "Trustee server" model

Authorize

Authorization refers to access to a particular user or service to access resources. After the user passes authentication, the user should request a specific function from the application. Permissions or "Authorization" can be assigned to users to perform certain tasks and cannot perform other tasks. In a secure environment, it is very important to limit the access level to authorized users.

Security experts often talk about "minimum authority principle". This is an empirical rule that the user should have sufficient permissions to perform the tasks required to perform, "but should not have more permissions."

encryption

Encryption is another way to ensure safety, encoding data to prevent unauthorized access.

Depending on the position of the encryption, encryption can be performed on many levels. Typically, encryption can be performed on the server, transfer or on the client.

The encryption on the server encryption server refers to the process of encrypting data stored and transmitted in the server infrastructure. After encrypting the data in the server infrastructure, it is ensured that the accessed sensitive data is unused due to the encrypted events. The user's credit card data is an example of dealing with data encryption. When the service layer stores the user's credit card information in the data layer, it is very important to encrypt the data. If a hacker invades the system and obtains access to the table for saving the encrypted credit card information, then the information does not have any use of hackers. If the credit card information is not encrypted, the application is negligible for the application on data security. Transfer encryption transfer encryption is specifically used to process data transmitted between servers and clients. For example, when the user submits an HTML form to the server, the user enters into the form Use Hypertext Transfer Protocol, HTTP to transmit, and then receive by the server. During transmission, the data may be stealing and tampered, which can be solved by encrypting data when transmitting. The data transmitted on the Internet can be encrypted in the following way: Install the security certificate on the web server, configure the Secure Socket Layer, SSL) port, use HTTP encryption form HTTPS as the transport protocol. The server certificate can be purchased from one of the certifications listed in http://www.microsoft.com/security/ (English). You can use Microsoft Certificate Services to publish an independent certificate, which will allow you to test SSL security on a single server. Please use the Web Server Certificate Wizard installation certificate that can be accessed through the site properties in the Internet Services Manager. When building SSL sites, it should be appreciated that the HTTPS: // prefix must be included in the hypolip or redirection of the user who has never encrypted session to the encrypted session. This specifies that the user's browser will communicate with the server using HTTPS. Client encryption client encryption is specifically used to process data residing on the client. For example, if a file is public, it is encrypted, then only the file with the correct decryption key can be used. For general e-commerce applications, client encryption is not as important as transmission and server encryption, but in some cases it may require this encryption method. Accessibility

Accessibility refers to the ability to access the site from a variety of devices or browsers. The Internet is developing forward with incredible speed, and the device accessing the Internet has also become a five-flowers. Therefore, making e-commerce applications can be accessed by a variety of devices and is a very difficult task on these devices.

The key to supporting multiple clients is to separate the representation from the content. Many methods can do this, including the logic written in the ASP page to generate different responses based on the client, or redirect different devices to alternate sites. However, one of the best ways to indicate logic from content is to use XML. If the data can be represented by XML, you can use the XSL style sheet as a specific type of client to render data. By applying different style sheets, you can represent the same content for different clients. Figure 4-3 illustrates this concept.

Figure 4-3: Separating the representation from the content

Determine available prior art

Available techniques refer to the techniques, products or services that can be used in the solution. Using existing technologies to achieve functions in economical, this is not necessarily to build these features. For example, when building a web application, you need an operating system as the foundation of the solution, but there is no need to build an operating system yourself.

Since there is no need to build an operating system every time you build an application, you will also do not need to build all parts of the web solution itself. Many experts believe that the future applications will be built using existing services, which can constitute an application only need to be re-combined. Therefore, it is important to collect information about which technologies can be used in the solution. In the next task involved in the physical design phase, the design team will analyze this information and determine which techniques (if any) can meet the needs of specific applications being discussed.

operating system

Any modern app is built on the operating system. The operating system not only provides an interface with hardware communications, but also provides a public framework for building an application. It is important to select a mobile system that supports object-oriented methods and public frames (applications can run and communicate on this frame).

Windows 2000 Server platform

Microsoft® Windows 2000 Server provides a rich set of content for application developers. Many major e-commerce sites on the Internet are running on Windows 2000, including Buy.com, barnesandnoble.com, Dell.com, and Intel.com.

Internet service

Another core part of a web-based application is an Internet service. WEB-based applications require an Internet service platform that is responsible for basic web services, such as response to client's HTTP requests, HTTPS requests, and other requests. A good Internet service platform should also provide site management capabilities and dynamic content programming models.

Microsoft Internet Information Services

Microsoft® Internet Information Services (IIS) is a web server built in Windows 2000. IIS provides a rich Internet service package, including ASP, DAV, Web Folders, FrontPage Extensions, FTP, multi-site hosts, and other support.

Express service

As mentioned in the "Accessibility" section, it is very important to separate the content and representation, and only this can access the content from multiple clients. This separation is also very important to simplify globalization.

As mentioned earlier, a method of separating the logic from the content is to use XML and XSL simultaneously. In addition to using XML and XSL, you can build complex page logic in the ASP page itself in accordance with an entity such as browscap.ini files, user_agent strings, and the like.

Microsoft XSLISAPI Filter

An alternate technology that can be implemented is a Microsoft® XSLisapi filter. ISAPI represents an Internet service application programming interface, which is the foundation of IIS. The filter is placed above ISAPI and provides the corresponding functionality when the web server receives a client or service request. For example, the ASP engine (ASP.DLL) is an extension running on ISAPI.

The XSLISAPI filter is used to intercept all documents with XML or PASP file name extensions. PASP file name extension is dedicated to applications used with the XSLisapi filter. Documents with this extended file are considered to be a standard ASP file (with certain restrictions) that can generate effective XML (rather than HTML) output. The output content of the PASP script is then converted based on the same rule of the XML file that applies the ISAPI filter to the XML file of the request.

After retrieving XML, the next step in the process is to convert data to certain types of valid tags (the client can display these tags). For browsers, this may be HTML; for mobile phones that support WAP, this may be WML. You can get xslisapi documentation from the following web location: http://msdn.microsoft.com/code/default.asp? Url = / code / sample.asp? Url = / msdn-files / 026/002/187 / msdncompositeDoc.xml (English). Figure 4-4 illustrating the concept of xslisapi filter:

Figure 4-4: XSLISAPI filter function

data service

For the construction of an e-commerce solution, it is also important to have a function of storage, retrieval, and managing data. These services are encapsulated into a database server. It is very important for corporate database servers, high performance, good hairiness, and scalability.

SQL Server 2000

Microsoft® SQL Server 2000 is a SQL database server that provides enterprise-level performance, scalability, and good concurrency. It also provides enrichment of XML, strict security, and powerful analysis tools.

business platform

If you start building an enterprise e-commerce solution from scratch, you will spend a lot of time and human resources resources. If you make full use of e-commerce functions in existing products, you will significantly shorten the development time of the entire program and save a lot of development costs.

Microsoft® Commerce Server 2000 is developed with existing features.

Commerce Server 2000

Commerce Server 2000 is a comprehensive product that encapsulates many features of the e-commerce solution into a package. Commerce Server 2000 provides advanced management capabilities, strong scalability and good performance. For more information, please visit the Microsoft Commerce Server website, which is: http://www.microsoft.com/commerceserver (English).

Analysis / rationalization

After the completion of the physical design phase, the work is followed by analysis and rationalization. Analysis and rationalization refers to the analysis of information collected during the study and makes decisions based on this information.

Use existing technology

When the design team considers the use of the prior art, it must be weighing the factors that may affect the solution. The possible factors are listed below, but it is not necessarily comprehensive:

Ability: Can this technology achieve business functions? Household costs: Whether the technology is cost effective? It is necessary to consider the cost of product, development, upgrade, license, deployment, and operation. Experience: What experiences and professional skills do this technology require developers? Do you have training costs? Is there an unknown fees? Mature and innovation: Is this product mature? Is it received by the market? Does the product have innovative, is it the latest technology? Is it still popular? Deploy: Is this technology difficult to implement? Supportability: Can this technology get support? Architecture: Whether the technology implements an acceptable architecture? Does the architecture meet the needs of enterprises? Scalability: Can this technology expand to meet the development requirements? Interoperability: Can this technology work together with existing systems in the organization? Performance: Can this technology provide the desired performance? Reliability: Can this technology meet the reliability requirements of the application? Availability: Whether the technology can handle application requirements without causing the solution to fail? Manageable: Is this technology easy to manage? Security: Whether the technology meets security needs? Standard compatibility: Whether the technology is compatible with recognized standards? Other factors, such as project schedules and budget constraints, and other internal projects that may involve it, should be considered.

Windows 2000 Server

Windows 2000 Server is an operating system for a business reference architecture because it provides a set of features specializing in corporate application design. These include the following functions:

Server type selection: The Windows 2000 Server platform can run on a variety of servers. In an environment of a business reference architecture solution, it is very important to expand the extension to extension. Depending on the application load requirements, organizations can choose to deploy solutions on two or more servers running Windows 2000 Server or Windows 2000 Advanced Server. To get the best performance and allow the load to be distributed on multiple servers, use the Microsoft 2000 DataCenter Server network and network load balancing (NLB). Scalability: By using Microsoft Windows 2000 Cluster Services and Network Load Balancing, Windows 2000 Advanced Server and DataCenter Server can expand.

By running the two servers, sharing a common storage mechanism, the cluster service ensures continuous services. If one of these servers fails, the other can take over. Due to the construction of redundancy in the system infrastructure, the application can reduce the downtime to zero.

Figure 4-5 illustrates the concept of a cluster.

Figure 4-5: Server in the form of a cluster

Windows 2000 Advanced Server and Datacenter Server can also extend outward through NLB. In this method, multiple servers are displayed as a single unit with a single IP address, and the application load is evenly distributed on these servers. When a server in the NLB settings fails, NLB automatically detects a failed system, transfer its load to other systems, and then restart your computer.

Figure 4-6 illustrates the concept of NLB.

Figure 4-6: Network load balancing

Availability: Windows 2000 Server provides a solution with high availability by using cluster services and NLB services in Windows 2000 ADVANCED Server or Windows 2000 DataCenter Server. Use NLB services and cluster services to eliminate single point failures. Reliability: The Windows 2000 Server platform implements the reliability of "5 9" to ensure that the normal running time is as high as 99.999%, which is equivalent to less than 5 minutes per year. In an enterprise e-commerce environment, the downtime means that the direct income of millions of dollars will bring trouble to customers, incur their complaints. Therefore, reliable operating systems are unsearched parts of the enterprise solution. Performance: Windows 2000 Advanced Server and DataCenter Server implement Symmetric Multiprocessing, SMP support, which enables the server to be valid for the processor of Advanced Server up to 8, and the processor for Datacenter Server is up to 32. In addition, Advanced Server also includes enhanced memory capabilities, allowing the server to have up to 8GB of memory; for Datacenter Server, the server allows the server to reach 64GB. Manageable: Windows 2000 Server provides a wide range of tools that allow you to manage the site and connect to Microsoft Management Console (MMC) to manage server features in a centralized location. Some management features in Windows 2000 include event records, performance monitoring, terminal services, and Windows Management Code (WMI), WMI), WINDOWS Management. Security: Windows 2000 provides a secure environment that strictly controls access to files or services through Active Directory, Security / Authentication Protocol, and communication encryption. Component Services: Windows 2000 COM service provides developers and administrators with application capabilities. This built-in function allows development of distributed transaction applications without having to develop a underlying infrastructure that supports a minimum unit of transaction or asynchronous operation. You can find more information about Windows 2000 Server in the following location: http://www.microsoft.com/windows2000 (English). Microsoft Internet Information Services (IIS)

Microsoft IIS is provided with Microsoft Windows 2000, which provides a rich platform for using Internet Submit content. Since IIS is integrated at the operating system level, it allows development and deployment to directly write solutions in the computational infrastructure.

IIS provides an Internet service for a business reference architecture solution because it provides and has a rich feature set along with the operating system. Two critical IIS features are Active Server Pages (ASP) and Internet Services Application Programming Interface (ISAPI).

XSLisapi

The XSLISAPI filter provides a service for the business reference architecture solution. After using the XSLisAPI filter, the content can be completely separated and the representation is fully separated, allowing the custom content to be sent without modifying the ASP code.

If you don't want to use XML and XSL at the same time, you can use another method: Build complex page logic in the ASP page itself. However, this method has two main disadvantages: First, complex display logic in the code may be difficult to manage; second, running this logic requires huge overhead. XSLisapi intercepts all requests for documents with XML or PASP file extensions. PASP file name extension is dedicated to applications used with the XSLisapi filter. Documents with this extended file are considered to be a standard ASP file (with certain restrictions) that can generate effective XML (rather than HTML) output.

The XSLisapi filter is then transmitted to the client using the XSL Style Table to translate the XML output of the PASP or XML page using the XSL style sheet to send the converted output content to the client.

There is also another benefit using the XSLisapi filter that can be allocated according to different skills groups, making the development process easier to manage. For example, graphics designers can create XSL according to their own interface specification, while ASP developers only need to consider how to pass the correct data to the interface.

SQL Server 2000

SQL Server 2000 provides data services for business reference architecture solutions. It provides a complete enterprise database solution that Commerce Server 2000 will depend on its data service. For many e-commerce solutions, you only need to install SQL Server on Windows 2000 DataCenter Server to provide the desired level of scalability. For sites that store data volume, you can create SQL Server databases on multiple servers and use "Distributed Partition View" to implement data access and updates across physical servers.

Commerce Server 2000

The development team is selected for Commerce Server 2000 because it is built for Windows 2000 platform and provides an e-commerce application with a rich set of content development, deployment, and management tools. About 80% of Commerce Server is defined in a logical phase, and the cost required for development is low. These objects are implemented as a COM object that can be used for ASP pages.

Commerce Server 2000 provides the following features:

Integration with other services and software features: Because Commerce Server 2000 is designed specifically for Windows 2000, its architecture can be integrated with the operating system. Commerce Server takes advantage of the Com feature in Windows 2000 Server, providing a solid foundation for enterprise e-commerce applications. Commerce Server can also be well integrated with Microsoft® BizTalkTM Server to reduce certain process management functions and external communications. Pipeline components: "Pipeline Components" is a set of configurable custom COM objects that are sequentially called to perform specific business processes. In the Commerce Server solution, most custom business classes can be implemented as a pipeline component to manage business processes in a simple way. (Pipeline assemblies are just some of the COM components of iPipeLineComponent). This allows Commerce Server to identify the pipe assembly to components suitable for Commerce Server, and can then be called by the pipeline. In the reference architecture application, the pipeline assembly is used to handle processes such as "customer order processing" and ensure that the tasks required to process orders in order. Management Infrastructure: Designing and Building Management Framework for e-commerce sites not only costs a lot of resources, but its workload is generally greater than the development efforts of the e-commerce application itself. The management infrastructure of Commerce Server is to attract the main factors that ultimately use Commerce Server. Commerce Server BizDesk is available with Commerce Server. BizDesk is DHTML-based applications that run on IE 5.5 or higher, allowing administrators to manage auctions, promotions, catalogs, orders, and users, and provide a rich set of analytical tools. Because BizDesk is completely DHTML, it can also be used to remotely manage the e-commerce solutions built on Commerce Server. This approach places the BizDesk application load on the client and does not affect the performance of the e-commerce site. Meet business needs

After determining the prior art that can be used, the development team must determine how these technologies meet the previously mentioned business needs. In order to meet these needs, the development team must make certain key decisions on tools, processes and methods. These decisions should be viewed in an environment that must meet specific business needs.

Globalization

As mentioned earlier, globalization refers to the ability to transplant the application into different cultural environments. In order to meet this demand, the content and representation of the user must be "global". During globalization, content can be divided into two types:

Static content - the content you can find on the interface and the interface itself. Dynamic content - "Dynamic Build" and display it to the user.

The following section describes the issues involved and key decisions made to two content globalization.

Static content

Static content includes text and interface itself on the interface, and must be localized for the specified cultural or language. The xslisapi filter is a localized: Allow the development team to use different interface designs for different languages, and then deploy applications as different websites or different virtual directories. By using this method, users can choose their own language and then redirect to the appropriate site.

Note: The reference architecture application has not been global; however, the application uses the XSLisapi filter, so it has the ability to achieve globalization.

Dynamic content

Dynamic content consists of content generated by the application "dynamic".

If the application will support different languages, it must support different character sets. The Unicode standard is a character set that contains characters in all languages ​​in the world. Using the Unicode standard ensures that all languages ​​can be represented in dynamic data. To support Unicode standards, all database fields that save character data must use data types such as nvarchar, nchar, and so on. At the same time, the business layer must support the use of Unicode. Therefore, the business reference architecture solution always uses the Unicode standard throughout the application.

performance

To get the best performance, the design team must make multiple key decisions. First, to maximize system total throughput (ie, the overall efficiency and performance of the application), the design team should solve the following key issues:

Seale language selection asynchronous processing

Minimize seal

One way to improve system throughput is to minimize the encapsulation. To do this, the best way is to reduce the remote procedure call of the website to the components in other locations. Many e-commerce sites reside in the area of ​​dedicated Web servers, while business components are located in separate application server clusters. Although this architecture can effectively ensure system security, especially when the application server is separated from the firewall or package filter switch and the web domain; but the architecture will have a negative impact on the response time because each of the components Extractions must be sealed through the network connection.

Business Reference Architecture Application The components are deployed on the same server where the web is located, so it avoids cross-network encapsulation and shorten the response time. (Commerce Server provides most business components used by the reference architecture application, which will be installed on the web server.)

language selection

Language selection also affects performance. For example, although components created for Commerce Server pipes can be written in a scripting language, but for enterprise applications, components should be built using lower-level languages ​​(such as Microsoft® Visual Basic® Development Systems or C ) to get optimal performance. Although the components like ADOs are almost the same in the Visual Basic component and the C components, complex business routines are run faster in C . Therefore, select C as a build language for all components in a business reference architecture application.

Asynchronous treatment

In order to minimize the response time, many processes should be designed as asynchronous operation. For example, when the user checks, it is not necessary to wait for the system to send an email to confirm before he receives the interface response.

Scalability

Solving scalability issues may be a very daunting task. The first method of extending applications is to expand up, mainly for a single server to configure a better performance, thereby increasing speed. For the upward expansion, although the desired design factors are relatively small, this method of satisfying the need is too expensive because hardware prices have rising index with performance.

Another way to solve the scalability problem is to add more servers, which are called "outward extensions". Although the extension extension is more cost-effective, it needs to consider more design factors. As previously discussed, the biggest problem extends to extension is to maintain session information.

In order to successfully expand, the business reference application should meet the following requirements:

Do not use the ASP session object to maintain session status because it introduces server session affinity and requires IIS to maintain session status in memory. With the assistance of the Commerce Server object, the user session status between the two page requests is saved in the database and can be retrieved by the new page request. Although this session maintenance method leads to some additional database overhead for each page request, it can meet the scalability requirements of the site. A single high-end database server (or cluster) can provide status saving services for the entire front-end server field. When the user logs in, a session cookie will be sent to the user, and the status data of the related user account is retrieved as the "Find Field". Each session cookie is not stored on the user's hard drive, so they can also be enabled even on the most secure user browser. If each session cookie is disabled in the user browser, the user will not be able to log in to the site. Manageability

If you select Windows 2000 Server and Microsoft Commerce Server, you will have a powerful management infrastructure. As previously described, Commerce Server provides powerful management interfaces in BizDesk, while Windows 2000 also provides powerful management interfaces through Microsoft Management Console and other components.

safety

In order to meet the security requirements defined for the reference architecture application, the design team has made the following issues:

Authentication

In the two available models of "counterfeit / delegate" and "trustee server", the design team selected the authentication scheme of the "trustee server" model as a business reference architecture application.

Because the "counterfeit" model is counterfeited each user, the solution must be an account for each user of the site. In intranet-based small applications, users are very small, and the restrictions are user-based, so the "counterfeit" model can run well; however, in a large solution, the "counterfeit" model will soon be difficult to control . Therefore, the design team chose a simpler "trustee server" model, which provides better performance and is more convenient to manage.

Authorize

In order to follow the minimum permissions described in the "Research" section of the physical phase, the design team is to choose deployment and design:

IIS virtual root directory permissions: After deployment, the site should be set to have read-only permissions to the virtual root directory. NTFS Permissions: After deployment, the Windows account for anonymous access to the folder containing the web application file only has read-only permissions. Accessible for anonymous customers: Business reference architecture solutions should use cookies to identify users who do not authenticate, and they should redirect them to the Login page when the user wants to check out or access any configuration file management page. Verified customers' permissions: Even users pass authenticated users, they should also be appropriately restricted. For example, system administrators can use Commerce Server BizDesk tools to restrict access to specific configuration files by hiding or only read-only permissions to users.

As for the database itself, there is more authorization issues. The unique direct permissions for the database have the accounts assigned to the intermediate layer application (this example is Commerce Server), and the account permission should be restricted to provide the minimum permissions required to provide data services to the site. Therefore, statement authority (such as DROP TABLE) is not assigned to the account.

encryption

As an example application that should be easy to install and check, the business reference architecture application does not implement encryption. However, in e-commerce system products, when transmitting sensitive data such as passwords or credit card details, encryption sessions should be used.

Although encryption in an e-commerce product environment is necessary, it should be avoided using SSL to connect when transmitting insensitive data. This is because a certain overhead is required when establishing an encrypted session, involving transmitting the public key of the server to the browser, and generating and exchanges the session key used by the encrypted session. Browser independent

The reference architecture application uses the XSLISAPI filter to meet the independence needs of the browser. It provides a first-class mechanism that separates content and representation, which is necessary to reasonably handle different browsers.

achieve

The final step in the physical phase is to apply decisions about constraints, demand and technology to logic design, and actually define physical realization programs. In this stage, the internal structure of the programming model, the component interface, and each component will be determined.

Identification component

Each component of the business reference architecture will discuss in detail in the second part of this "Developer Guide". The encoding method, the components used, and the code segment and the definitions thereof will be described in this section. For more information, please refer to the developer's comment provided by the code itself.

Create specification

At the end of the "implementation" phase, the development team must form a detailed technical specification to form a detailed technical specification. This document will become a blueprint for building an application, and the development team is used to use it when forming an expert group, preparing a list, assignment task, and creation testing and deployment plans.

to sum up

This chapter describes what components will be used to determine which components will be included. Which technologies are used, which can be subdivided into three smaller phases; and summarize the development business reference architecture application consolidatedRetail. The basis. Among the entire project development cycle, the ultimate goal of the physical design phase is to apply the actual physical design constraint to logic design, and write a reasonable technical specification to guide development.

The next part of this guide will focus on the actual code provided in the "Business Reference Architecture: Enterprise to Consumers" applications. As mentioned earlier, this application is developed as a reference example, and some modifications are required when used as the product.

Chapter 5: Implementation Overview

Abstract: "Microsoft Business Reference Architecture: Enterprise provides consumers" offers reusable and customizable components that can be used to speed up the e-commerce solution development process of large and medium-sized enterprises. This chapter outlines the knowledge required to fully understand the code, and simply introduces the components of this application.

The following chapters will deeply explain the specific aspects of the code, and the specific implementation issues encountered during the code development process.

Introduction

As in the first part of the Developer Guide (English), "Microsoft Business Reference Architecture: Enterprise provides consumers" provides reusable and customizable components, which can be used to speed up high and medium-sized enterprises. Program development process. Microsoft Business Reference Architecture: The B2C solution consists of some engineering-processed code components. Developers can use them to build the Microsoft® Windows® 2000 Server operating system and Microsoft® .NET Enterprise Server products (such as Microsoft® Commerce Server) B2C retail website for 2000 and Microsoft® SQL Server 2000).

The second part of this guide is for applications developers and all other people who want to learn to use Microsoft technology to build an e-commerce solution. This section contains the initial overview and component guidelines for this application and its components, and discusses in-depth discussion of implementations encountered during its development process.

Implement function

The consolidatedretail.com site implemented in the reference architecture application includes many of the common features in an e-commerce solution. For a complete description of application business requirements, see the first part of the Developer Guide.

Suggested background knowledge

To understand the details of the code function, the readers of this article should have the following knowledge:

ASP: Developers should learn how to use Microsoft® Visual Basic® Scripting Edition (VBScript) to create Active Server Pages (ASP). This includes understanding the core ASP object and instantiates and uses the component object model (COM) and COM components. A large number of code used by this app is an ASP code. XML: In this application, a large number of Scalable Markup Language (XML) is used as a method of transmitting data between application components. Developers should be familiar with XML, including basic syntax, architecture, and structures. Some of the code uses an XML document object model (DOM) to grammatical analysis of XML content. XSL: To understand how the user interface works, developers need to fully grasp the extensible style sheet language conversion (XSLT). Microsoft Commerce Server 2000 and its objects: Commerce Server is a complex system. Therefore, although this application and document provide some guidance information about how to use the system, developers should also refer to the Commerce Server document to familiarize the functionality and components of Commerce Server. SQL Server: Although this application uses SQL Server, most actual data inventory operation is still hidden behind the scenes, you need to perform through the abstract layer provided by the Commerce Server object. However, to truly understand the access and storage of data, developers should carefully study the tables and stored procedures maintained by Commerce Server in the background. Visual C : In order to get the best performance, this application's custom COM component is written in Microsoft® Visual C ® development system. Although it is necessary to obtain valuable information from the reference architecture and does not need to fully understand the internal working principle of the component, if you want to fully understand its working principle, C skills are required. For information on why these technologies are selected, refer to the first part of the Developer Guide.

Technical overview

The following sections briefly introduce how the reference architecture application uses each core technology.

XML in consolidatedretail.com solution

This solution is implemented by a Commerce Server 2000 site. This site uses SQL Server 2000 for data storage; manage it using Commerce Server Object Library. The site itself uses XML to represent the content of the webpage, and converts the content to Hypertext Markup Language (HTML) for web browsers using the XSL style sheet. Use an Internet Server Application Programming Interface (ISAPI) application filter to handle XML and apply the necessary XSL conversion. Using this method, the solution can be easily extended to other types of clients without having to rewrite the business logic with ASP.

Microsoft has promised to use XML as a standard for transmitting structured content in a distributed computing environment. In this case, the business reference architecture code uses XML throughout the solution. Examples of using XML include:

Generate an XML output from the PASP script, then use the XSLisapi filter to perform XSL conversion: Execution of the Solution Site Front End Function Preprocessing Active Server Pages (PASP) script uses the XSLisapi filter to generate an output of an XML format (rather than HTML format). This further separates the content and its expressions. XML generated by these PASP pages is converted via the XSL style sheet. For more information on how to use the PASP and XSLISAPI filters, see the XSLisapi Filter Documentation, which can be downloaded from the Microsoft website. Note: The custom version of the XSLisAPI filter is used in the reference architecture application code. This version cannot be downloaded from www.microsoft.com; however, the documentation of the public version (XSLisapi.dll, version 2.1) also applies to custom versions. Output XML from Business Components: Many Commerce Server 2000 objects can use ActiveX Data Objects (ADOs) or XML formats to pass record sets (for example, product lists). In a possible case, the ASP script code requests information from the business object to the XML format. Write the front end script flexible and free to use the ASP

On the Microsoft platform, for most web-based applications, the code that drives the web front end is performed by using the ASP file written in VBScript. The main reason for using ASP for web programming is that it provides flexible development. In the field of rapidly developing e-commerce, it is important to quickly change the appearance of the site or the basic front end function.

Interpretation of scripting languages ​​Execution allow programmers and designers to quickly change the function of the site without having to go through the lengthy compilation process: You only need to save the source file, that is, the program has changed. However, it should be noted that in order to achieve optimal performance and scalability, the main business processing tasks are performed by components (provided by Commerce Server).

Business Reference Architecture B2C Solution Codes Add new changes in ASP programming, using the XSLisAPI filter to isolate the generation of content. That is, the pre-processed Active Server Pages (PASP) script does not generate HTML content directly, but generates an XML output, and then converts the XML output to the correct display format by the XSL style sheet. This conversion can be performed on a server or client (if the client has this capability), thereby alleviating the working pressure of the CPU.

Commerce Server component

Commerce Server 2000 provides a very rich tool for site developers, system administrators, and business managers that can be used to develop, deploy, and manage web business applications. Like any powerful large software system, Commerce Server is strong, and the wide range will definitely let you be at a loss unless you are familiar with the system. The reference architecture provides an appropriate definition for yourself: it is a reasonable design, indicating that a full small e-commerce application shows a lot of Commerce Server objects in use. Although there is no function of Commerce Server, the reference architecture B2C code demonstrates how to implement the following features:

Set the boundary, which content and operations can be accessed by anonymous users, which can be accessed by authentication. Do not rely on the method of introducing server session correlation, keeping the user's session status; and reserves the session information when accessed accesses and through the boundaries of validation. Create a product catalog, browse the directory level, and display the details of the product through the directory. If you do not require verification, you can allow customers to collect products you want to buy (using shopping carts). Store and retrieve user configuration information, so that users do not need to re-enter this when accessible. The Commerce Server 2000 object used in the reference architecture application is listed in the chapters of this guide.

Storage procedure in the data layer

This app does not write any custom stored procedures. All stored procedures in the Commerce Data Store on SQL Server are created by Commerce Server for providing data access capabilities for many Commerce objects.

Client script for input check

The user interface of this application uses the client JavaScript code to enter verification. The function of the client script is not very complicated, and there is no major association with the core function of the application to some extent, so this guide does not introduce these scripts.

to sum up

This chapter describes the background knowledge you need to make use of this guide. In addition, components constituting the reference architecture application (consolidateDretail.com application) are summarized. The "Solution Guide" details the code component, provides a code example, and provides links to other information.

Chapter 6: Solution Guide

Summary: This chapter provides a solution to the solution implementation component, including: Advanced graphical representation of the application, how the PASP file and how the XSLISAPI filter works, and a description of each primary component category in the consolidatedretail.com application.

Introduction

Reference Architecture Application contains many custom ASP files and code components. The sections of this chapter will detail the source code for these files. Figure 6-1 visually clearly shows the various components of the solution and its interrelationship. When reading the rest of this guide and viewing the source code, the map can be used as a basic reference.

Figure 6-1: Reference Architecture Application Component

As you can see from this figure, users use the browser (here INTERNET Explorer 5) to access the application. The ASP file in the website is implemented as a pre-processed ASP (PASP script), and the programmable object provided by Commerce Server 2000 (such as ProductCatalog, AuthManager, UserProfile, and ORDERGROUP) sends data to the Commerce Server site database in SQL Server 2000 and Retrieve data from this site database. Business processes such as order or see the shopping basket content are achieved through the Commerce Server pipeline. Pipe call Custom COM queuing components to process email messaging functions. The output of each web page is represented by XML and then intercepts by the XSLisapi 2.1 filter, and the XSLisApi 2.1 filter applies the corresponding XSL style table to the content and renders the content as HTML.

The next section of this chapter will explain how the PASP file and the XSLisAPI filter work, and a brief example is provided as a reference. This chapter will then introduce how each primary component category is implemented in a consolidatedretail.com application. Our discussion uses logical reasoning ways, first introduces Web services, then introduce the Commerce Server object, followed by discussing pipe components. PASP file and xslisapi filter

Active Server Page (ASP) is often used to generate representations of web solutions. In the consolidateDretail.com site, use their approach. At this site, most ASP files are implemented as pre-processed ASPs (ie, PASP files). The PASP file writes an XML to the response object, which is intercepted and processed by the XSLisapi filter. The script in the PASP file is used to call business logic in the Commerce Server 2000 object; the XSLisapi filter processes the representation.

The XSLISAPI filter performs syntax analysis from the XML output generated from the PASP file and checks the Processing instructions. If the processing explains the Server-Config property that references the XML configuration file, the XSLisapi filter reads the configuration file to determine which XSL style table should be applied to the XML output generated from the PASP file.

The XML configuration file contains information about which style table should be applied to a particular client type, as shown in the following code example:

IIS can identify the client program as long as you check the Hypertext Transfer Protocol (HTTP) request header that is sent with the page request. IIS uses the entries of the browscap.ini file in the Winnt / System32 / IneTsrv folder, you can identify many normal browsers. The XSLisapi application provides additional entries for devices for WAP phones in files called BrowsCap-add.ini, and users should copy and append them to the browscap.ini file.

Simple example

The following simple example shows how to use the XSLisAPI filter to present the XML output generated by the PASP file.

Note that this example is intended to help you understand the functionality of the XSLisapi filter. It does not meake any actual pages in the consolidatedretail.com site.

Assume that the user uses Internet Explorer 5 to request a page name myproducts.pasp. The system will explain the scripts in the file, then return to the following XML: by the following response object

Server-config = "productconfig.xml"

HREF = "csproducts.xsl"?>

Now, the XSLisAPI filter will perform syntax analysis of the XML and read the XML-Stylesheet processing instructions, which contains the following three properties:

TYPE: Refers to the type of style sheet to be applied (here XSL). Server-config: Refers to an XML configuration file that defines a style sheet to be used for a specific client. HREF: When the corresponding style sheet is not listed in the Server-Config file, the client will download and apply the default style sheet.

Then, the XSLisapi filter reads the specified Server-Config file (ProductConfig.xml), which contains the following entry:

Since IIS has identified the client browser based on the request header to Internet Explorer 5.0, the XSLisapi application now applies the Products-IE5.xsl style table to XML retrieved from MyProducts.PASP.

Assume that the Products-IE5.xsl style sheet contains the following XSL code:

Version = "2.0">