Re: Image Maps: A Presentational Structure

David Woolley wrote:

> Whilst this may be common usage and is presentational, image maps have
> a legitimate use which represents real content, and was probably the
> original reason for having them.  An image map can be a real map with
> regions for sales territories, that might otherwise be very difficult
> to describe in words.  It can be used to help people identify parts
> of an assembly, without knowing their technical names.  Etc.

   Ok, that's quite a good response, though I still think there is a 
very fine line between it's presentation and it's semantics.  Unless the 
image can be presented visually (screen, projection, etc.) all the 
semantics, if any, given by the shape and coords attributes cannot be 
perceived.  Non-visual or text only media types (aural, braille, tty, 
etc.) ignore the attributes and use the fallback mechanisms provided 
such as alt text and title attributes, or whatever else is provided.

   The examples provided for client side image maps in HTML 3.2 [1], 
HTML 4.01 [2] and XHTML 2 [3] (see example below) are all using an image 
map for a simple navigation toolbar.  These examples are presentational 
and don't fit your explanation above because neither the position, size 
nor shape of each area have anything to do with the purpose of the 
links.  It just seems like they're using an image for text, which is 
presentational.

Example from XHTML 2 draft.
<p src="navbar1.png" type="image/png" usemap="#map1">
     <nl id="map1">
         <label>Navigate the site:</label>
         <li href="guide.html" shape="rect" coords="0,0,118,28">
         Access Guide</li>
         <li href="shortcut.html" shape="rect" coords="118,0,184,28">
         Go</li>
         <li href="search.html" shape="circle" coords="184,200,60">
         Search</li>
         <li href="top10.html" shape="poly"
             coords="276,0,276,28,100,200,50,50,276,0">
         Top Ten</li>
     </nl>
</p>

> Actually, whilst general image maps were once used presentationally
> for text as images and icon hyperlinks, the tendency now is to
> create a mosaic of single images with a presentational abuse of tables.
   I rarely see real image maps in use.  “Image slicing”, as it's known 
in some authoring tools, to create the mosaic of images in a table, is 
easier and unfortunately it seems to be encouraged by making it such a 
prominent feature.  Image Maps are more accessible than presentational 
tables, though I still think they have some presentational aspects.

[1] http://www.w3.org/TR/REC-html32#map
[2] http://www.w3.org/TR/html401/struct/objects.html#edef-MAP
[3] http://www.w3.org/TR/xhtml2/mod-attribute-collections.html#col_Map

-- 
Lachlan Hunt

Received on Monday, 1 March 2004 04:06:04 UTC