Re: Re : Influence of valid code on screen readers

Roberto Castaldo wrote:

>Matt:
>We're not in a position to say what is and is not "allowed". The WCAG WG is
>not the Web police.
>
>Roberto C:
>I think we are in such position: the whole web community interested in web
>accessibility is strongly waiting for WCAG 2.0, and requests some guidelines
>that are very near the concept of "web accessibility police"; but we always
>say that Web should not be not accessible, so WCAG 2.0 will have a deep
>impact on the whole Web.
>  
>

Unless we put a requirement on them that they don't consider logical, or 
that they consider too much work. In which case, we will have a standard 
that appeals to Web standards people, and is ignored by everyone else.

>Roberto C:
>I think that we should avoid to give tag soup any kind of dignity, as it
>simply is the absence of rules in developing web pages. From Wikipedia: "Tag
>soup is HTML code, written without regard for the rules of HTML structure
>and semantic meaning"; that should be enough for any W3C recommendation to
>not allow this superficial way of approaching web development.
>  
>

I'm not defending tag soup. That practice needs to be fixed. I'll spend 
the next several years of my career pounding that fact into designers' 
heads. But we should only go so far as to instruct authors to use 
semantic markup and accessibility-related metadata appropriately at all 
times. Validity doesn't ensure that, since people can still be using 
<font> instead of <h1>, etc., and it requires doing a lot more work 
that's unrelated to accessibility.

Validity is a fragile thing. An unclosed <td> or an unterminated &nbsp 
buried in a CMS database somewhere doesn't mean an entire site is 
functionally inaccessible. But it does mean that at least one rendering 
of it is invalid. Getting people to spend their time making it valid 
instead of increasing its accessibility is not the solution. I would 
rather focus authors' energies on techniques that directly increase the 
accessibility of HTML markup than on a binary indicator that does not.

-
m

Received on Friday, 17 June 2005 20:28:19 UTC