Re: ISSUE-31 Change Proposal

On Thu, 28 Jan 2010, Laura Carlson wrote:
> 
> I have drafted a Change Proposal for HTML ISSUE-31.
> 
> Summary:
> The current guidance for conformance checkers for Section 4.8.2.1 the
> img element is unclear and does not implement WAI CG's advice on the
> validation of short text alternatives. This change proposal replaces
> the current guidance with clear guidance that lists all required short
> text alternative options that exist to be considered valid. It enables
> automatic validators to programmatically detect the presence or
> absence of text alternatives.
> 
> Full proposal is at:
> http://esw.w3.org/topic/HTML/ChangeProposals/ImgElement20090126
> 
> Ideas for improvement are most welcome.

Is there a bug that this corresponds to? Off-hand I couldn't see any 
reason I would reject the editorial aspects of this proposal, and it would 
be quickest for everyone if we could just deal with the uncontroversial 
aspects in a bug rather than through the big process.

Regarding the proposed normative changes, I disagree with several things:

  * Making the presence of ARIA attributes make otherwise invalid HTML 
    valid. ARIA is only exposed to AT tools, but we need to make sure it 
    is possible to have an accessible experience even without use of an 
    AT, which means pages have to be conforming without using ARIA. ARIA 
    adds to the experience, but shouldn't be required to understand the 
    page.

  * Adding an attribute whose exclusive purpose is shutting up a 
    validator. I do not believe this will result in an improvement in 
    accessibility. The one attribute from HTML4 that was like this (nohref 
    on the <area> element) was dropped in HTML5 because it turns out to 
    have not worked. I think it would be bad to make that mistake again.
    (Even worse would be to add _two_ such attributes on the same element, 
    as is proposed here.)

  * I do not think it is a good idea to make sighted people e-mailing 
    other sighted people privately have to either include alt="" text, or 
    have to write a spec that overrides this one. There's really no reason 
    to include alt text in that case.

  * There are images whose contents are not known; just not catering for 
    that use case is not acceptable, IMHO.

We should probably split these issues out into four (or more) separate 
change proposals though, rather than try to discuss them all at once, 
since they are somewhat orthogonal to each other.

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

Received on Thursday, 28 January 2010 22:10:01 UTC