W3C home > Mailing lists > Public > w3c-wai-ig@w3.org > April to June 2004

RE: Screen readers - usage stats?

From: Jewett, Jim J <jim.jewett@eds.com>
Date: Fri, 16 Apr 2004 10:13:26 -0400
Message-ID: <B8CDFB11BB44D411B8E600508BDF076C1E96D470@USAHM010.amer.corp.eds.com>
To: w3c-wai-ig@w3.org

>> ... having Konqueror identify itself as MSIE.

Steven Dale:
> I know not why Konqueror takes that path

The same reason that Opera does.

There are too many sites still using ancient 
browser sniffers that refuse to serve anything 
but MSIE and Netscape.  

If screenreader version were actually part of the
protocol, I'm sure there would be sites that said
"we've only tested with JAWS, so we won't serve
anything to other systems".  And so WE would start
offering users a chance to cloak...

Steven Dale:
> work more on separating content and presentation
> in the area of accessibility and embrace the CSS
> and XSL technologies.  For what happens when the
> next device comes to market?  Simple a new
> CSS stylesheet for that device.

Ian Anderson:

> What you are suggesting is a move towards balkanised 
> sites, code forking and multiple site versions for
> specific use cases, the way it was in 1998.

Not really; he is saying to make one site with the
information, and include enough structural hooks to 
get your presentation entirely from CSS.  Then do 
your tweaking strictly in CSS.

For instance, there was an example of "displaying" a
navigation link hundreds of pixels offscreen.  That 
way it didn't bother sighted viewers, but screenreaders
could still see it.

Mark up that link as a class that might be treated 
differently (I think there was already a pseudoclass
in the example), and move it offscreen in the CSS.  
Serve your default CSS plus a *small* browser/device/tech 
specific sheet which only includes the few changes that
are worth presenting differently.  You only have to make
the change once (not once/page).  You don't have to make 
it at all if the difference isn't important.

Received on Friday, 16 April 2004 10:14:53 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 13 October 2015 16:21:28 UTC