Re: Dave Raggett: Re: IMG ALT attribute in HTML 4.0 (fwd)

I support Daniel's proposal on all counts. More comments later.

- Mike


At 02:58 PM 9/15/97 +0200, Daniel Dardailler wrote:
>
>In looking at the matrix (OBJECT/IMG) X (MAP/no MAP) X (short/long
>desc) I came up with some issues and ideas.
>
>
>Using OBJECT
>
>Simple image with descriptions (both short and pointer to long)
>
><OBJECT data=foo.gif>
>  shortdesc and a pointer to
>  <a href=foo.html class=longdesc> long desc </a>
></OBJECT>
>
>Simple image with map and long description.
>
><OBJECT data=foo.gif shapes>
>this is long desc with
> <a>map coords</a> integrated 
> <a>map coords</a>
>but without long desc pointer
></OBJECT>
>
>To allow long desc pointer in the above example, the spec for OBJECT
>shapes needs to be changed to say it only looks for <A> that have a
>"coords" attributes (preferred) or A that do not have a class of
>longdesc.
>
>Introducing class=longdesc (maybe just in guidelines, but probably
>better in spec) for A seems like a useful addition in any case (since
>it allows the browser to identify it and present it to the user in
>some special way, in graphical environment or not).
>
>
>Using IMG and LONGDESC
>
><IMG src=foo.gif alt=shortdesc longdesc=foo.html>
>
>This is more or less equivalent to the first OBJECT example (no map),
>more strict in fact since it doens't use class=longdesc but a real
>attribute (but doesn't allow for integration short desc/long desc
>pointer).
>
>Now if we consider map:
>
><IMG src=foo.gif alt=shortdesc longdesc=foo.html usemap=map>
>
><MAP name=map>
> <area alt=shortdesc1>map coords</area>
> <area alt=shortdesc2>map coords</area>
></MAP>
> 
>We see that we run into a problem that was already mentioned:
>potential duplication in longdesc (talking about shortdesc1, 2, etc)
>and the MAP itself providing area and alt for the same shortdesc1. 2.
>
>One way out would be to extend MAP with the same OBJECT shapes feature:
>
><MAP name=map shapes>
>this is desc with
> <a>map coords</a> integrated 
> <a>map coords</a>
></MAP>
>
>
>I think the changes to the OBJECT-shapes-parsing-A rules and the
>class=longdesc convention are needed, that IMAGE longdesc is good to
>have and that MAP shapes extension is deluxe.
>
>
>
>

-------------------------
Michael G. Paciello                      Executive Director, 
E-Mail: mailto:paciello@yuri.org         Yuri Rubinsky Insight Foundation
Tel: +1 603 598 9544                     URL: http://www.yuri.org/
FAX: +1 603 598 2839                     Promoting Accessibility Awareness!

Please Make a Tax-Deductible Donation:   http://www.yuri.org/donate.html

Received on Monday, 15 September 1997 10:18:01 UTC