What's wrong with image maps?
- They are invisible and unusable to text mode browsers, and to
graphic browsers with graphics disabled (usually to facilitate
browsing over slow lines, such as dial-ins).
- Unlike HTML lists or menus, they are not scalable; an image map 700 pixels
wide will not be visible on a screen or window 640 pixels wide without
- Their colors and text fonts are fixed, so they cannot be configured
by sight-impaired users.
- The graphic files are large, often in the range of 15-100k,
compared with a kilobyte or less for a text menu.
- The URL of the destination link does not appear to the user until
he or she selects the link. Nor can the browser tell exactly which
areas of the map are clickable.
- Most importantly, the results of an image-map transaction are
unknown to the browser, so that future accesses to the same map
location cannot be retrieved from the browser cache, nor from a
proxy cache, resulting in
much slower performance.
Some of these warnings do not apply to the client-side image map,
a new kind of image map that is being implemented on some browsers.
In a client-side map, the destination URL is visible to the browser,
hence cacheable. Client-side image maps are part of the new <OBJECT>
element; they are also implemented, somewhat differently, in the
<MAP> element, introduced by Spyglass Corporation
and partially supported by Netscape 2.0 and Microsoft Internet
Warren Steel (firstname.lastname@example.org)
16 October 2002
| Hints for Web Authors
| Warren Steel ]