add line breaks through css?

[Spartanicus]

I suppose you are thinking of a speech browsers that interprets the HTML
source and ignores any CSS (or applies aural features in CSS, in the remote
future).

Opera 8.x supports a useful portion of aural/speech CSS from the 2.0
Spec/the 2.1proposals and properties from the CSS 3 proposals prefixed
with "-xv-".

But there's more. Think about seeing, on a graphic browser, links like

foo bar zap zip zap zup ump

Are there seven links, or less? Maybe "foo bar" is one link? If links are
underlined, breaks in underlining may give a hint. Too subtle?

Code a border: foo | bar | zap | zip | zap | zup | ump

 

[kchayka]

But it is common to use speech browsers that work "upon" graphic browsers,
reading aloud what appears on screen

This is true, but I'm not convinced hiding the marker is that big of a
problem. An experienced screen reader user sets the reading speed up
faster than you or I could ever follow. They may also navigate just
through links or other structural elements. In either case, they
probably wouldn't even notice whether the marker was on screen or not. A
less experienced user might notice something, depending on their
particular reader and how they have it configured.

But there's more. Think about seeing, on a graphic browser, links like

foo bar zap zip zap zup ump

Are there seven links, or less? Maybe "foo bar" is one link? If links are
underlined, breaks in underlining may give a hint. Too subtle?

I don't think you should rely on underlining, since that is a preference
setting in most browsers, though I've never known of anyone turning it
off by default. Regardless, if the designer opts for a horizontal list,
but doesn't provide sufficient separation on screen via spacing,
borders, colors or some other means, it's degraded usability for sure,
but not likely an accessibility issue.

BTW, 1em between links as the only separator is indeed too subtle for my
tastes, even if each link is a single word.

 

[kchayka]

It can, but does it?

Don't take my word for it:
<URL:http://www.access-matters.com/2005/03/27/111-alt-text-for-custom-list-markers/>

The fact that their scenerio uses image markers coded in HTML isn't
really relevant. Their test page still uses list markup and hides the
marker with list-style-type:none. Virtually all comments agree than
using list markup alone is enough.

This is a fantastic site for real-life web accessibility issues, BTW.
Actual screen reader users input a lot to the site. It's much more
informative than webaim.org, which seems rather stagnant.

foo bar zap zip zap zup ump

Are there seven links, or less?

If it were 7 links, JAWS would read this as:

“List with 7 items foo bar zap zip zap zup ump list end”

with a slight pause after each list item. There is link to a sample
recording of JAWS reading such a list on the above page. It ignores the
bullet marker whether it is shown on screen or not, but does use the
markup to identify it as a list. Window-Eyes also identifies lists from
the markup.

Not all screen readers behave this same way, but JAWS is the most-used
reader so I'd tend to keep it in mind first. I think it's up to the
individual to decide whether to cater to readers that perform less well
than JAWS. To me, it's the same issue as dealing with graphical browsers
that don't handle CSS well. It's likely a waste time trying to
accomodate a tiny percent (even less than 1%?) of your visitors that may
be using a deficient UA. YMMV, of course.

 

[Spartanicus]

http://www.access-matters.com/

[...]

This is a fantastic site for real-life web accessibility issues, BTW.

As a demonstration of accessibility issues, perhaps.

Actual screen reader users input a lot to the site. It's much more
informative than webaim.org, which seems rather stagnant.

Seems to have been coded by a mentally disabled person. The body text is
wrapped in an h2.

 

[Beauregard T. Shagnasty]

http://www.access-matters.com/


As a demonstration of accessibility issues, perhaps.


Seems to have been coded by a mentally disabled person. The body
text is wrapped in an h2.

Which page is that, Spartanicus? I looked at the access-matters.com
link provided, and the main page at webaim.org, and can't find any
mismatched pairs of h2. Maybe my Find is borked while viewing source?

 

[Spartanicus]

Which page is that, Spartanicus?

Whoops, one of my web filters is acting up on that page, apologies.