Re: feedback requested on WAI CG Consensus Resolutions on Text alternatives in HTML 5 document

On Tue, 25 Aug 2009, Maciej Stachowiak wrote:
> On Aug 24, 2009, at 10:24 PM, Ian Hickson wrote:
> > 
> > I would be open to including references to documents that could help 
> > authors and implementors -- UAAG, ATAG, WCAG, UTR #36, CHARMOD, etc. 
> > Indeed, we already have a reference to CHARMOD and UNIVCHARDET.
> Can't speak for Steven, but that sounds like a really good set of 
> references to me. I think the references should be somewhat more 
> prominent than in HTML4 or SVG, maybe in the introduction instead of 
> relegated to an appendix.

I've added references to these in the intro.

On Tue, 25 Aug 2009, Steven Faulkner wrote:
> In the WAI-CG consenus document there are 2 recommendations that I am
> addressing in this email:
> "We recommend
>    - that HTML5 state that "For guidance on accessibility requirements for
>    text alternatives authors should consult WCAG
> 2.0<>
>    ."
>    - and that HTML should not provide any guidance that conflicts with WCAG.
>    "
> The first request is unambiguous to me, can you tell me what needs to be 
> made clearer?

The reasoning behind the request -- that is, the answer to the question "why?".

WCAG2 gives very little advice regarding alternative text, at least an 
order of magnitude less than HTML5 itself does. Why would we suggest that 
readers look at WCAG for advice on alternative text specifically?

Should we also reference WCAG at every other relevant point? It doesn't 
seem like that would be useful. We don't reference, say, CHARMOD 
everywhere that it might be useful, or the AWWW document, or any number of 
other informative resources or additional references that might help the 
reader -- so why WCAG, and why specifically in this section?

> The second request is ambiguous as the areas where the WAI-CG considers 
> that the content of the spec conflicts with WCAG has not been detailed. 
> The WAI groups are in the process of a full review of the HTML 5 
> specification and I understand that details of conflicts, if any, are 
> forthcoming.

As far as I can tell, WCAG2 refers only to HTML4 and XHTML1.x, so it's not 
possible for HTML5 to contradict it. However, if there is any advice in 
HTML5 that is demonstrably bad for accessibility, I would be very happy to 
change it.

> I would obviously be pleased if the accessibility references that point 
> authors and software develoeprs to the relevant specifications were 
> prominent within the spec.

I've added an intro section that references the WAI documents (amongst 

On Tue, 25 Aug 2009, Henri Sivonen wrote:
> On Aug 25, 2009, at 08:24, Ian Hickson wrote:
> > 
> > I don't really understand what problem we're trying to solve here. Why 
> > would we give authors using WYSIWYG tools a license to not care about 
> > making their pages accessible? That seems backwards.
> It's not about giving authors using WYSIWYG tools a license not to care 
> about making their pages accessible. It is about acknowledging that when 
> there's an abstraction layer that hides HTML syntax from the author, the 
> syntax error-based feedback loop to the author doesn't work and instead 
> the feedback is deflected by the tool developer.

I don't follow. If the author isn't checking the syntax of the document, 
and if the author is actively ignoring the tool's requests to make the 
document valid, why would the author then complain about the document 
being invalid? Or are you saying other people complain? I don't fully 
understand the situation you are describing.

> The problem being solved is removing the incentive not to conform to 
> ATAG 2 in order to perform the deflection. (The tool vendor can't 
> enforce WCAG 2 compliance of the tool users.)

(I don't think WCAG 2 need be invoked here -- as far as I can tell, HTML5 
itself has the relevant requirements too.)

> The right place to for software to complain at users of WYSIWYG tools 
> about lack of accessibility is the WYSIWYG tool complaining at the user. 
> A validator is the right place to complain at authors who don't use 
> WYSIWYG editor-like abstraction layers between them and HTML syntax.

Agreed. So why does it matter what the validator says to WYSIWYG tool 

Ian Hickson               U+1047E                )\._.,--....,'``.    fL       U+263A                /,   _.. \   _\  ;`._ ,.
Things that are impossible just take longer.   `._.-(,_..'--(,_..'`-.;.'

Received on Monday, 31 August 2009 08:54:25 UTC