|
|
|
| |
        
|
 |
|
To access the contents, click the chapter and section titles.
HTML 4.0 Sourcebook
Use the ALT Attribute! If you use graphical banners to distinguish your pages, you must also use IMG element ALT attributes to assign alternative text strings to the images. Then, the important organizational information contained within the image is available to a visitor who is not able to view graphics. As an example, an Information Commons banner for a document describing e-mail services should have an ALT attribute value such as “INFORMATION COMMONS—Help With E-Mail.”
Multiple Document Versions Ideally, you should design all your pages, including the home page, so that they can be understood on any Web browser. However, sometimes this is not possible, or it is inconvenient—for example, you may be presenting information both to users who have learning disabilities and prefer graphics over text, and to users who are blind, and can only use text. To deal with this type of conflict, some Web sites offer multiple home pages.3 A good example of this is found at the University of Toronto’s Adaptive Technology Resource Centre site home page, shown in Figure 4.13. This page is graphically rich, but offers an alternative home page designed for users with screen readers and those who want to avoid graphics. Note also how the Silicon Graphics Web site (Figure 4.11A) offered this same alternative view.
You may also want to offer alternative navigational pages that structure the underlying content of a site from a different point of view. For example, a collection of material on Adaptive Technology (aids for persons with disabilities) can be organized either as a tree presenting important issues and technologies or as a problem-solving tree, where the various hypertext links represent options, in a decision-tree, for obtaining technology solutions for particular user disabilities. A schematic for this type of structure is given in Figure 4.14, which illustrates two organizational trees linked to the same underlying content. You must be careful with this type of design, since the documents at the bottom of the tree (the actual data in your collection) will display identically no matter if they were accessed from tree A or tree B. The navigational icons on each document must therefore point to both trees and to the appropriate places on each tree. This presents difficulties similar to those at the Yahoo! site, where several list categories appeared below multiple nodes in the tree. At Yahoo!, the referenced page lists all the possible parent nodes. Unfortunately, accessing these alternative parents can be quite disorienting for an inexperienced user, unless their nature as optional routes is well explained—note all the efforts the Yahoo! administrators have made to make this as easy to navigate as possible. You should visit Yahoo! and try this out for yourself. Of course, one way to improve the readability of the pages is to customize the page layout, dependent on the pages previously seen by the visitor. This is possible when the pages are generated by special software on the server, as discussed in the next section.
Gateway Filtering of Delivered Documents Although the Web is a dynamic medium, HTML documents by themselves are static and unintelligent—as just noted, a retrieved page contains no information about the history of the user’s interaction and cannot respond in a customized manner depending on the page(s) previously viewed. This means that, if you use multiple home pages, all documents that link back to the home pages must have multiple “Go to HOME” buttons to account for each possible origin—or these links must be omitted to avoid disorienting visitors. In principle, you can in part deal with this issue by using a gateway program (gateway programs are discussed in a Chapters 10 and 11), or some other server-side processing mechanism, to dynamically modify every document returned from the server. For example, suppose there are two alternative structures for a document collection, labeled pathA and pathB, each path starting from a different home page. You can implement this collection so that the home pages do not access the remaining documents directly, but instead access them through a gateway program. This program processes the documents so that they include the navigational icons and other information that reflects the path being explored. Suppose the first page to be explored from home page A is the document monsters.html. In this example, the page is accessed via the URL: <http://www.we.edu/cgi-bin/docfilter/dir1/dir2/monsters.html?path-a>
where the gateway filtering program is docfilter, the path to the desired document is /dir1/dir2/monsters.html, and the query string path-a indicates which path is being explored. The docfilter program retrieves the indicated document, and edits all URL strings in the document so that they also use docfilter to access documents from the local collection, and so that they also contain the query string path-a to indicate the path being explored. In addition, docfilter could modify the document banner, inserting the navigational icons and hypertext links appropriate to this path.
|
|
Products | Contact Us | About Us | Privacy | Ad Info | Home
Use of this site is subject to certain Terms & Conditions, Copyright © 1996-2000 EarthWeb Inc. All rights reserved. Reproduction whole or in part in any form or medium without express written permission of EarthWeb is prohibited. Read EarthWeb's privacy statement.
|
)
)
)
){kind=link}
){kind=link}
){kind=link}