Thu, 19 2006, Alastair Campbell wrote:

Alastair Campbell wrote:
The only accessibility product I have ever recommended for
a client that
isn't from us:
http://www.udm4.com/
[snip]

Bob replied:
tables don't work in plain text worth beans, just continuous
lines of text with no spaces.

Sorry, I don't understand, what tables?

the tables that listed the various browsers and how they worked

under the drop down "browsers supported" I think

Bob

N response will ever be given to anyone using earthlink.net

"They that can give up essential liberty to obtain a little temporary
safety deserve Neither liberty nor safety", Benjamin Franklin
- - - - - - - - - - - - - - - - - - -
ASCII Ribbon Campaign accessBob
N HTML/PDF/RTF in e-mail accessys (AT) smartnospam (DOT) net
N MSWord docs in e-mail Access Systems, engineers
N attachments in e-mail, *LINUX powered* access is a civil right

THIS message and any attachments are CNFIDENTIAL and may be
privileged. They are intended NLY for the individual or entity named

Friday, Fri 20/10/2006 20:05, Chris wrote re the website www.gap.com <http://www.gap.com:

"The site uses a CSS layout which linnearizes nicely, has alt attributes for all of its images, and includes "skip navigation" links so keyboard users can move straight to the actual page content. It seems to be a good effort at accessibility, though on closer inspection there are some problems:

1) The alt attributes are often meaningless (e.g. "what to wear now").
2) The site uses an image map.
3) The skip navigation links refer to their targets by their physical location. "Side navigation" and "top navigation" are utterly meaningless if I can't see the whole page at once in a graphical browser."

I don't want to appear to be encouraging criticism of anyone, or organisation, who is striving to develop an accessible website. However, I would be unhappy if Anna were to consider that this website exhibited, as Chris wrote, "a good effort at accessibility".

In addition to the problems noted by Chris, I tried 3 basic tests on 1 web page (Women in the navigation bar), all of which failed viz:

*
It claims to be XHTML 1.0 Strict, yet it did not pass the W3C's XHTML validation test.
*
It uses CSS, yet it did not pass the W3C's CSS validation test.
*
It does not allow the resizing of text to accommodate any user who has a slight visual impairment - I use glasses to read and had left them in another room, so I wanted to increase the text size in Internet Explorer, but couldn't.

Whilst there are more tests which should be conducted on this website, I fail to see how, given these 6 problems, it could ever be considered to have made "a good effort at accessibility".

>From a Business point of view this latter problem could, and probably will, result in a loss of potential customers who when the web page's textual content fails to resize will simply surf to a competitor's website.

In my humble opinion there needs to be an agreed set of minimum requirements which should be met before a business-oriented website is considered to have made "a good effort at accessibility". The problem would seem to be the formulation of a universally agreed set of minimum requirements.

Dave

"Eadie, David" <D.Eadie (AT) gcal (DOT) ac.ukwrote:
1) The alt attributes are often meaningless (e.g. "what to wear now").

That's not meaningless, it's the heading from the text in the image.

2) The site uses an image map.

What's wrong with that? It's a client-side image-map and all the links
have alt text.

Jon,

If you read my e-mail properly then you will find that the points you are referring to come from Chris' e-mail. I merely used them as a basis for my final point.

I made 3 additional points: the validation failure for the XHTML and CSS, along with the inability to resize the text.

My 'final point' reworded, for Jon's benefit, is that whilst here in the UK it has been a legal requirement since 1999 to create an accessible website there appears to be little more than lip service paid to accessibility by a significant number of websites. Some of these websiters continue to receive praise as having made 'a good effort'. What exactly constitutes a good effort today?

Dave

From: w3c-wai-ig-request (AT) w3 (DOT) org on behalf of Jon Ribbens
Sent: Mon 23/10/2006 13:29
To: WAI Interest Group list
Subject: Re: examples of sites with good accessibility

"Eadie, David" <D.Eadie (AT) gcal (DOT) ac.ukwrote:
1) The alt attributes are often meaningless (e.g. "what to wear now").

That's not meaningless, it's the heading from the text in the image.

2) The site uses an image map.

What's wrong with that? It's a client-side image-map and all the links
have alt text.

Email has been scanned for viruses by Altman Technologies' email management service - www.altman.co.uk/emailsystems

10/23/06, Jon Ribbens <jon+w3c-wai-ig (AT) unequivocal (DOT) co.ukwrote:
--
"Eadie, David" <D.Eadie (AT) gcal (DOT) ac.ukwrote:
1) The alt attributes are often meaningless (e.g. "what to wear now").

That's not meaningless, it's the heading from the text in the image.

2) The site uses an image map.

What's wrong with that? It's a client-side image-map and all the links
have alt text.
--
That quote is from me rather than David. You're right. Chalk it up to me
starting to write the post at four in the morning, plus my general aversion
to seeing content and design sharing the same bed.

Chris

"Eadie, David" <D.Eadie (AT) gcal (DOT) ac.ukwrote:
If you read my e-mail properly then you will find that the points
you are referring to come from Chris' e-mail. I merely used them as
a basis for my final point.

Ah, my apologies. I was confused by your failure to use proper
quoting.

The problem is that the headings are styled as display: none. Many
screenreaders (Jaws among them) only read the screen-rendered content, and
so ignore any element with display: none.

solution might be to shrink the headings down to pixel size:

h1 {height: 0; width: 0; overflow: hidden;}

and then use absolute positioning to move them somewhere inconspicuous.

Chris

10/26/06, Geoff Stephens <GeoffsLists (AT) geoffandwen (DOT) comwrote:

And did you test with assistive technology? Some of the headings aren't
showing up for my screen reader. I do not know whether this is the
result of something defined in a style somewhere, a bug in the way the
screen reader handles the content, or something else.

The use of a list here is a good example of the trend to use lists as a
catch-all solution. This is made more unclear because, as I said, the
headings are not displayed as part of the content by my screen reader.

<div id="breadcrumbs">
<h1>You are in </h1>
<ul>
<li class="home"><a title="Transport for London" href="
http://www.tfl.gov.uk">Transport for London</a></li>
<li class="current"><span>Current section</span></li>
</ul>
</div>
of the definitions of list is:
The property possessed by a line or surface that departs from the vertical
--
the following heading is the only heading I can recognize in the rendered
version:
<h1>Please select from the options below before proceeding.</h1>

Maybe I'm out of the loop on the current practices with regard to Title
and Alt atributes, but is there any reason why you use both? Unless you
know how assistive technologies handle these attributes you run the risk of
possibly conveying an unintended meaning. This is further complicated by
the fact that assistive technology is apparently electing to handle Alt and
Title in whatever way they think best.

10/27/06, Lois Wakeman <lois (AT) lois (DOT) co.ukwrote:

"The problem is that the headings are styled as display: none
solution might be to shrink the headings down to pixel size"
A better one would be to rewrite the content so it uses headings properly
in
the context of the content - surely? And not just for users of AT.

The suggestion was an attempt at increasing accessibility while keeping
Andy's web app as close to its original form as possible.

Also, with the benefit of a graphical user agent, I can tell whet the
different sections of the page are by their position: navigation is at the
top or left side, the main content is in the middle, supplemental
information I can probably ignore is at the bottom. In that case explicit
headings for the sections aren't necessary. However, as soon as the context
of position is gone, headings become really really helpful for identifying
where I am on the page.

Chris