Transcending Client/Server
By Ira Walters
Because of how browser-based applications work, they often feel different from custom software. A browser typically requests information from a server and displays it as a single transaction. Later, if the browser requests new information, it must reconnect to the server and perform another transaction. Then, the page refreshes to display the new data.
This limitation prompted Microsoft to develop remote scripting. This lets browsers execute methods on the Web server as easily as if those methods resided in a script on a Web page. More importantly, a browser can request new information from a server (for example, a database) and display that information without refreshing the page.
As a security measure, remote scripting restricts the data types you can transmit to integers and strings. Because structured data like arrays and objects create potential security risks (for example, a virus being sent to your server), you can't send them. And just as with forms, any input is automatically converted to a string. However, the server may send arrays and objects back to the client (although I've been able to do this only when I use JScript as the server-side programming language).
Implementing remote scripting is a three-step process:
- install three Microsoft files onto the Web Server;
- make methods available from the server side;
- enable the client side to access these server-side methods.
You'll need to get the following three files from Microsoft (
|
