WebAIM - Web Accessibility In Mind

E-mail List Archives

RE: sales people


From: Jukka K. Korpela
Date: Oct 22, 2003 12:09AM

On Tue, 21 Oct 2003, Tim Harshbarger wrote:

> Michael,
> In your case, the word "image" may be correct because you provide additional
> text to describe the image.

Michael originally wrote:
"The full alternative text says that the image is a link to a site where
one can get a program that produces accessible client side image maps."
and later clarified that "full alternative text" means the alt text as a
whole. Without seeing the actual context and markup, it is difficult to
say exactly how adequate the alt text is, but it's probably basically OK.
The point however is not whether there is additional text to describe the
image. The point is whether the alt text acts as a suitable _replacement_
for the image, serving well in a situation where the image is not seen at
all, so that it would be rather pointless to _describe_ it.

The following would probably work as a link to a site where one can
get a program that produces accessible client side image maps:
<div><a href="...">FooBar</a>, a program for producing
accessible client side image maps.</div>
where FooBar is the name of the program. There are good reasons for
keeping link texts relatively short and name-like. And if you
then replace FooBar, in the link, by an icon of the program, for example,
then the replacement should apparently be <img alt="FooBar" src="...">,
shouldn't it? If you wish to _describe_ the icon, the title="..."
attribute could be used, but that would be somewhat pointless and
even disturbing here, when the image is a link.

> However, since evaluation tools are vastly
> limited in their capability to understand the intent of the value of an alt
> attribute, they can make mistakes when attempting to determine if the alt
> text is adequate or one of those situations where the words are simply a
> place filler.

Indeed. And the more heuristics an evaluation tool applies, the more often
it produces warnings that are completely nonsensical. Contrary to what one
might expect, a high level of heuristics (that is, lots of different
"practical" tests) is suitable for _experts_ only.

> If you are doing a presentation on evaluation tools, this may provide a very
> helpful example to you.

Actually I am currently writing an evaluation of evaluation tools, and I
find this (and the HiSoftware tool in general) an illustrating example.

> The fact is evaluation and repair tools are only
> part of the solution for making web sites more accessible.

I'm afraid that's not quite accurate. They are often part of the problem

> In this case,
> your knowledge of accessibility allowed you to interpret the feedback from
> the evaluation and repair tool correctly.

Hopefully this discussion helps in that direction. In any case, even
people who know accessibility issues rather well can be confused with the
reports of evaluation tools. Those reports are certainly, at times,
_very_ confusing and even misleading to people who are just starting to
try and understand accessibility. This implies that we should not
encourage people into using such tools too early. At worst, they will
spend their time "fixing" page features that needed no fixing, or
fixing them in the wrong direction, and making pages less accessible just
to please an evaluation tool.

Jukka "Yucca" Korpela, http://www.cs.tut.fi/~jkorpela/

To subscribe, unsubscribe, suspend, or view list archives,
visit http://www.webaim.org/discussion/