Re: First Stab at Set of Principles for 'Minimum Conformance'

"William Loughborough":
> At 09:04 AM 10/25/01 +0000, Jim Ley wrote:
> >there's no way to test that the illustrations apply to the text
>
> How do you test that the text (alt) applies to the image?

It's a property of the IMG, therefore you know that it applies to the
image (simply by definition).  There are no mechanisms in current HTML,
for saying "this is some alternative content/ an illustration for this
element" - so it's impossible for a tool to know what an image on a page
"applies to".

I don't think that was truly the question you intended to ask, but that's
the problem with the testability of this requirement.

As you say there is indeed a large problem of the appropriateness of the
alternative content, be textual or otherwise - How do tools recognise
this? - currently the approach with alt appears to be to check for common
mistakes with alt text (filename, filesize, empty etc.) and then if these
aren't there assume it is okay.  This is perhaps inadequate with alt
text, but I tend to think not, all but intentional misleading by the
author (which nothing will stop) seems to be covered by it.

The problem as I see it, is the lack of a way in HTML of associating
alternative content other than text to elements.

Jim.

Received on Thursday, 25 October 2001 10:08:48 UTC