When it comes to Web applications, the idea of state is sort of a no-no, because Web applications should be eminently scalable, and if you include state information, then that scalability goes away. Still, lots of Web applications use state information. And that means any .NET developer worth his salt better know what's what with state in ASP.NET. This tip, excerpted from InformIT, talks about the state concept and defines the kinds of state available.
Attempting to manage state in Web applications goes against the fundamental design principles of the Web. One of the primary goals of the Web and its underlying protocol, HTTP, is to provide a scalable medium for sharing information. Adding user state inherently reduces scalability because the pages shown to a particular user will be different from those shown to another user and thus cannot be reused or cached.
In spite of this fact, many applications deployed on the Web require user-specific state to function properly. Applications ranging from e-commerce shopping sites to local company intranet sites depend on the ability to track individual requests from distinct users and store state on behalf of each client, whether it's items in a shopping cart or which days were selected on a calendar as requested vacation days.
Although maintaining client-specific state is not officially part of the HTTP protocol, there is a proposal in place for adding state management to HTTP. RFC 2109 defines a proposed standard for state management for HTTP also known as cookies. Although it is only a proposed standard and not yet an official part of the HTTP specification, cookies are in widespread use today in almost all browsers, and many Web sites rely on cookies for their functionality.
As a consequence, Web programmers must be very conscious about state management. Unlike traditional applications, Web applications must be very explicit about any state that is maintained on behalf of a client, and there is no one standard way to maintain that state.
One of the most important decisions you face when designing a Web application is where to store your state. ASP.NET provides four types of state: application state, session state, cookie state, and view state.
ASP.NET, like its predecessor, ASP, provides a pair of objects for managing application-level state and session-level state. Application state is where information that is global to the application may be stored. For efficiency, this state is typically stored once and then read from many times.
Session state is maintained on a per-client basis. When a client first accesses any page in an application, an ASP.NET generated session ID is created. That session ID is then transmitted between the server and the client via HTTP either using client-side cookies or encoded in a mangled version of the URL. On subsequent accesses by the same client, state associated with that session ID may be viewed and modified.
Cookies provide the ability to store small amounts of data on a client's machine. Once a cookie is set, all subsequent pages accessed by the same client will transmit the cookie and its value.
Finally, view state is a yet another way of storing state on behalf of a client by saving and restoring values from a hidden field when a form is posted. Although this technique for retaining state has been used by Web developers in the past, ASP.NET provides a simplified mechanism for taking advantage of it. It is possible to place items into the ViewState property bag available in every Page-derived class. When that page issues a POST request to itself, the values placed in the ViewState property bag can then be retrieved, the key restriction being that view state works only when a page posts to itself.
To read the entire article from which this tip is excerpted, click over to InformIT. You have to register there, but the registration is free.
