RE: imagemaps

An alternative method would be to generate the same information as an
overlaid map, with alt / title information based on pieces of the the map. As
it happens, not every pixel actually needs to be a hot-spot, but the
processing that was used is based on a server side model rather than handing
a part of the information to the client. (That information may not be kept in
such detail, but I suspect in this case it is, since they can generate
instructions naming everything).

This would be very similar to the implementation required for dealing with
SVG and doing the processing on the server side - modulo an XSLT transform,
for those who can write XHTML.

The difference between the text based interface to these services (where you
can enter two addresses and get instructions about how to go from one to
another) and the map-based interfaces is that you are presented with a way of
choosing in the graphical version, whereas without it you pretty much have to
know exactly where you want to get from and to.

cheers

Charles McCN

On Thu, 15 Mar 2001, Bailey, Bruce wrote:

  Dear Charles,
  Thanks for thinking of this common counter example.
  Services like MapBlast and the like can be quite accessible.  The text
  directions are, for example, quite good.
  But what about the maps themselves -- where EVERY pixel is different hot
  spot?
  Can such a service achieve Single-A compliance or better?  (Without
  resorting to SVG as Charles suggests.)  Where does WCAG 1.0 make allowances
  for such things?  Is there not a way to eliminate all the extra verbiage
  about server side images maps, but still allow these kind of tools?
[snip]
  -- Bruce

Received on Thursday, 15 March 2001 09:16:32 UTC