- From: Al Gilman <asgilman@access.digex.net>
- Date: Fri, 18 Jul 1997 08:50:09 -0400 (EDT)
- To: jasonw@ariel.ucs.unimelb.EDU.AU (Jason White)
- Cc: w3c-wai-wg@w3.org
to follow up on what Jason White said: > > I think the use of server-side image maps should be strongly discouraged > in favour of client-side maps, which have been officially available since > the release of HTML 3.2. The above solution is appropriate to server-side > image maps and may perhaps be offered in guidelines as an alternative, but > I think the most important message which should be communicated to authors > is that for the benefit of accessibility, they need to make use of > client-side image maps with proper ALT attributes. In fact, if I remember > rightly, the ALT attribute is required in the elements that designate the > regions of a client-side image map. > Here is the situation the way I understand it, based entirely on unverified rumors and inferences: Authors and WebMasters would not presently use a user-side map without a server-side map because not enough browsers have implemented the user-side map feature. The auxiliary text navigation page that Geoff suggested is usable with a server-side map processing, or with user-side map processing. The need for this page is way less if you have a good USEMAP with well-written ALT entries. But its presence is no problem for the use of the USEMAP as well. The only difference is that if everyone were using USEMAP already, the recommended way of linking to an image description would not be by the default but by a dedicated region (perhaps including 0,0 ) with ALT="text description of the image". I haven't reviewed the structure of the USEMAP element to see how discursive one can get in wrapping the links in text. Sometimes it takes more than just a list with ALT text in the links to convey the graph created in the image, as for example: "The cabinet departments are arrayed around the White House in a circle to emphasize their equality of rank, as Arthur sat his Knights at a round table so none would fight over seats. In alphabetical order, they are <list of links>" The reason I didn't rely on USEMAP in my writeup is that I had internalized the assumption that it is so poorly supported by browsers in the field now that authors cannot use it except redundantly on top of server-side image map support. This, as I said earlier, is based entirely on unscientific methods of data collection and analysis and it would be helpful to know more about how thorough or spotty the implementation of USEMAP is among browsers in the field now. The whole issue of USEMAP vs. Yet Another HTML Page as the text-usable navigation bridge for a sensitive map merits further consideration. -- Al Gilman
Received on Friday, 18 July 1997 08:50:30 UTC