RE: Style sheet readability

Tim Springer wrote:

> I wanted to get your opinions on what it really means for a page to be
> readable when style sheets are disabled.

I think http://www.w3.org/TR/WCAG10-TECHS/#tech-order-style-sheets
says it rather clearly, though implicitly: the page must make perfect sense
without style sheets, and style sheets just add some suggested visual,
aural, or other presentation style to it. A style sheets could suggest some
particular way of making headings prominent, for example; you're still
supposed to use markup that indicates headings as headings so that _any_
browser, with or without style sheets, can present them as headings.

> The popular notion is that
> this constraint means that when a page is disabled elements 
> that are on the page should be visible to the user -
> so no black text on a black background.

Is it? If you apply the recommended approach, purely structural markup and
all presentational suggestions delegated to CSS, such problems don't arise
at all. You're not saying anything about any colors, except in CSS.

> As far as I can tell there are three broad cases when page content may
> not be "readable" with style sheets disabled. 

The most common case, which is getting more and more common I'm afraid, is
the use of <div>, <span>, or non-HTML XML, relying on style sheets for
"formatting", which is virtually always designed for visual on-screen
presentation only. When the document lacks headings, paragraphs, emphasis,
tables, and other structuring, it becomes plain text when presented without
style sheets (or, more exactly, without obeying all of the author's style
sheet). It could be even worse, of course; the text fragments might even be
out of order.

> The first is when content
> is generated on the client side via style sheets such as with the
> 'background-image', 'list-style', or 'content' properties.

Background images cause problems only when displayed, unless they're abused
in order to present actual content. The same applies to list style
properties. Generated content is a problem if used to generate real content.
The simple criterion is: disable style sheets and listen to the page.

> The second is when structural content is communicated with formatting
properties,
> such as 'border', 'border-width' and 'outline', instead of the
> appropriate structural markup.

Yes, the relevant thing is "instead of the appropriate structural markup".
It doesn't matter much what you say in CSS after deciding not to use
appropriate structural markup.

There are borderline cases, of course. There is, for example, no structural
way (in HTML) of saying "this paragraph is less important than the main flow
of content". Some people use "quasi-structutal markup"
<p><small>...</small></p>, with or without CSS, whereas others just use <p
class="...">...</p> with some CSS; it's hard to say which is better, or the
lesser of evils. But a failure to indicate a paragraph as less important is
hardly a _crucial_ matter in accessibility.

> The third is when positioning
> information is used in the style sheet that causes the page
> linearization to be different than the visual read order of the page.

The simple criterion is whether the page makes sense when read sequentially,
ignoring style sheets. If you use CSS for, say, putting a "navigation menu"
into the upper right corner, the real question is where it appears when CSS
is not in use. Hopefully at the bottom. It might matter where is appears
when CSS _is_ in use, since a screen reader might read the page by its
visual appearance, intermixing the menu items with actual content; but I'd
classify this as a user agent problem.

> Further it is my understanding that readability with style sheets is a
> constraint to ensure that older browsers could access page content.

No, it's _much_ more than that.

Since virtually all authors who use style sheets write them for certain
types of visual presentation only, it is sufficient for most purposes to
ask: does your page sound good when read aloud, by a speech synthesizer (or
a human being) that only looks at the markup, ignoring CSS?

-- 
Jukka Korpela, senior adviser
TIEKE Finnish Information Society Development Centre
http://www.tieke.fi
Phone: +358 9 4763 0397 Fax: +358 9 4763 0399

Received on Wednesday, 14 August 2002 02:35:40 UTC