Re: CSS Futures

Philip TAYLOR wrote:
> Can you explain this further, David ?  Are you arguing that,
> when a group of authors for a particular realm are unlikely

It's important to understand that the people you have to educate are not 
the authors but the authors' bosses, the salesmen and the purchasers.

> to car about accessibility, they should be pushed towards
> using a language that ignores the need for accessibility [1]
> rather than being pushed to realise that accessibility matters ?

I'm being pragmatic.  If people authored for accessibility first, then 
added what styling was compatible with the accessible structure, that 
would be the right approach, but I see no likelihood, in the foreseeable 
future, that authors will work in any other one than one that requires 
appearance and behaviour to match what the salesman sold or the 
purchaser requested.

In that context, it is better to have authors writing in a language that 
matches what they are being asked to do.  If they write in HTML/CSS, 
they will write really bad HTML/CSS and then bolt the accessibility onto 
that, making a result that is even more contorted and bloated.  With a 
tool matched to the presentational needs, the accessibility overlay can 
be done by someone else, who won't have to worry about making the 
accessibility work without changing the HTML more than minimally.

> Philip Taylor
> --------
> [1] I don't know if SVG /does/ "ignore the need for accessibility",
> but that is the drift I get from David's message, whence the
> reason for my question.

SVG has claims to accessibility, but I think they amount to:

- all text is text nodes, so will still be text (but probably in
   a totally random order!)
- there are various hooks for adding metadata;
- user agents are supposed to obey various user agent accessibility
   guidelines (but in practice that is a low implementation priority);
- it has a font mechanism that might avoid the need for image
   replacement (although with people putting WordArt through PDF, I'm
   not convinced that the mass of authors wouldn't end up bitmapping

Its bad feature is that it makes it very easy to construct pages in a 
completely random order.  The bad thing compared to PDF, is that the 
latest PDF has mechanisms to allow a logical re-arrangement and 
regrouping of the text into its semantic structure and logical reading 

David Woolley
Emails are not formal business letters, whatever businesses may want.
RFC1855 says there should be an address here, but, in a world of spam,
that is no longer good advice, as archive address hiding may not work.

Received on Tuesday, 3 July 2007 20:13:59 UTC