Re: Grey Text

Well, user stylesheets are indeed readily available, but are difficult 
to implement successfully for many users - the complexities of visual 
layout are such that essentially users can choose to completely 
override everything an author does, and force a single layout, or can 
force a couple of properties like colour and font-size all the time 
hoping that sites will work (often they won't). In other words CSS 
hasn't yet reached its promise.

We did a bunch of work on this at Sidar, and our conclusion is that 
although we can solve the really hard cases the majority of people are 
in-between and aren't winning yet. The best bet is Opera, which 
provides a feature where you can easily swap between a collection of 
user stylesheets. This allows you to create various partial stylesheets 
to try and retain whatever of the author's style won't interfere. But 
writing stylesheets is still hard for many people, although there are 
more tools that are helpful now.

(Like lots of  stuff gets done for people who are totally blind and 
rely competely on a screen reader, but relatively little for the much 
larger number of people who have some residual sight and could take 
advantage of better accessibility in other ways, if people bothered).

It's getting better. It's heaps better with audio, which is relatively 
simpler, but screen-reader manufacturers seem to prefer using a 
proprietary styling mechanism that locks users into their product, in 
the main. (Kudos to folks like speakthis who have implemented audio 
styling, and the European Union website which is implementing it now.)

I suspect, based only on experience rather than a careful survey, that 
there are a lot of people who have problems reading grey, and that many 
of them don't know how to configure their systems. (It's the kind of 
problem faced by older people who are, as a group, generally less aware 
that they can easily tweak small things to solve the small difficulties 
they have)

So if the problem is solved today, my take is that means style 
changers. (A potential alternative is to ensure that there is generally 
high "brightness" contrast, so people can adjust their monitor 
contrast. But it isn't the way to make friends, I don't think).

Actually I don't think grey on white looks very nice, although I do 
like blue on white which has similar problems due to degeneration of 
the eyes.

The idea that WAI doesn't like things that look nice strikes me as a 
crock. The idea that there are certain popular design effects which are 
not, a priori, very helpful to accessibility is certainly true. But 
then there are always people who like things subtle and understated, 
and others who prefer them screamed out loud and in your face. It's 
inevitable that tastes differ. Fortunately, as Joe implies, the 
technical solutions WAI favours are those which least inhibit design 
choices, to the extent compatible with people being able to access 
content, and as new technology is adopted the situation generally 
improves.

An interesting case is the use of dynamic multimedia. It is very hard 
to change the colours of flash, so you are left with building 
stylechangers into each flash presentation. (Well, you could write a 
generic one and always include it). In SVG, which uses CSS for styling, 
you can rely on almost the same solutions as people already use for 
HTML pages. But currently flash has wider deployment - so the 
developers are the ones who have to carry the burden if things are 
going to be accessible.

cheers

Chaals

On Saturday, Dec 27, 2003, at 21:41 Europe/Rome, Isabelle wrote:
>> -----Original Message-----
>> Joe Clark
>> To provide real accommodation for the full range of low-vision people 
>> in
>> these and similar cases, stylesheet-switchers and user stylesheets are
>> required. WCAG WG would love to ban anything that looks nice, like
>> grey-on-white text, but that isn't the correct solution.
> Great suggestion.
>
> Thanks for the feedback - problem has been solved!
>
--
Charles McCathieNevile                          Fundación Sidar
charles@sidar.org                                http://www.sidar.org

Received on Saturday, 27 December 2003 19:13:00 UTC