Web applications: Separating the browser and the application

Navigation is an essential characteristic of the Web, much more so than on other platforms. Web authors can link any page to any other, and users can pursue information at will. The importance of navigation on the Web serves to demonstrate the separation between an application (such as a search engine or online store) and the browser (which simply shows the data on each page).

This split is perhaps clearest with the Back button, which is part of the browser, not the application. If the user clicks Back three times, the application does not "know" it. Even though it might seem that backing up over those pages is undoing the task they represent, the user is simply seeing the pages fly by in reverse order. The button re-displays the pages, but does not typically undo the state of the application. Similarly, clicking Forward is not a Redo command.

Thinking of the data as separate from the controls used to display it is new, because applications on other platforms combine data and controls in a seamless whole. Users of the Web, especially novices, do not recognize where the browser ends and the application begins. See an illustrated example.

Rather than having to rely on the browser’s Back and Forward buttons, users need links within the application to take them where they need to go. Servers can keep sufficient information about the previous page for an application to have its own Back button on each page.

Alternatively, Web applications could provide toolbars and menus (with Java, which introduces a different set of problems) so that users do not need the browser’s navigational tools.

Navigation in Web applications

1. Introduction
2. Basic navigation
3. Losing users and losing work
4. Separating the browser and the application
5. Lack of context
6. Delays caused by network connections
7. Conclusion

Download the full article (PDF, 52KB)