W3C home > Mailing lists > Public > w3c-wai-ig@w3.org > January to March 2005

Re: Copywriting for Screenreaders (was Alt text for URL's)

Date: Tue, 15 Feb 2005 09:15:11 -0500
Message-Id: <s211bdac.032@GWF2.EEOC.GOV>
To: <w3c-wai-ig@w3.org>

>>>> "david poehlman" <david.poehlman@handsontechnologeyes.com>
2/15/2005 8:20:03 AM >>>
>Ok,  I'll put it succinctly.  If site navigation is so bad that it
needs to 
>be skipped,

It's not a question of bad. It's a question of should the user have to
go through it each time he/she accesses a page. Clearly, nobody believes
the answer is yes. Everybody wants to find a solution for the user to
easily get to the section of the page - navigation, content, or
subsection of content - that he/she is looking for.

Most people here seem to favor skip links. You suggested earlier using
a table of contents approach. There's really little difference, except
that with the skip link approach it's one link, to one section of a
page, and in the table of contents approach it's many links, to many
sections of a page - what you seem to be advocating as good site
structure. But is that really good site structure? If you need to use a
screen reader, yes, a table of contents approach is great. It allows you
to naivgate through a page and find what you need, quickly, efficiently,
and easily.

The problem is, for a visual user, without motor impairments, a table
of contents may be a lousy approach. If a page is designed with multiple
elements, visually arranged so that the entire structure can be easily
seen and understood, with minimal, perhaps no, scrolling necessary,
what, to the average user, would be the point of a table of contents? It
would provide little to no benefit to navigation or understanding of
site structure. It would be one more visual element to be understood, so
that it can be used or ignored. There would be some confusion as to why
there's a list of links to something that's right on the same page,
visible on the same screen at the same time. It would use screen "real
estate" unnecessarily, and compromise the design of the page. 

Here's an example. Go to http://www.cnn.com/  There's a great deal of
content on that page. I've used it with a screen reader, and it's a
royal pain in the butt to use. I've tried it with Lynx, and it's
similarly bad. Clearly, a table of contents would be of significant
benefit, and would make the site much easier to use.

However, my vision and mobility are unimpaired. I can use Firefox or
Internet Explorer or another visual browser. I can see the visual layout
of the site. Although it's far from perfect, it's actually pretty easy
to use, relatively well laid-out, with a straight-forward, easily
understandable structure. Navigation within the page is easy. As a
sighted user, a table of contents, containing half a dozen or more
elements, would not help at all, would look out of place, would
introduce an element of confusion and clutter, and would generally be a
bad thing. That's why even the simple skip link is usually made
invisible, to get it out of the way of sighted users. That's a bad idea
for several reasons, but the practice started for a reason.

The simple, unfortunate reality is that a site's contents can and
should be made fully accessible to all users, but in many, perhaps most,
cases, the user experience cannot be optimized for sighted users,
non-sighted users, and mobility-impaired users all at the same time.
Skip links are a compromise - they make things somewhat better for users
of screen readers and text browsers, and for anyone using a keyboard
instead of a mouse. You're right, they don't do the job completely. But
revamping the page structure of most sites, or adding on a full table of
contents, often makes things worse for sighted, mouse-using users.
Whether it's right or wrong, very few site designers would even consider

Of course, every page is different. If you're talking about a single
long document with minimal external navigation, a table of contents may
be a perfect solution, but that's a minority of pages at best.
Received on Tuesday, 15 February 2005 14:16:07 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 13 October 2015 16:21:31 UTC