Re: CfC: to publish "HTML5 Image Description Extension" specification as a First Public Working Draft (FPWD)

Some of this is relevant to the TF, since it is technical discussion where  
we may want to express an opinion. Since I don't think we should try to  
merge the longdesc spec directly into some HTML spec yet, I don't think  
there is any action required, but below I explain why in more detail. Of  
course it is possible that people disagree with me, and these are only my  
personal opinions...

On Wed, 20 Feb 2013 01:04:49 +0100, Silvia Pfeiffer  
<silviapfeiffer1@gmail.com> wrote:

> The below are my personal opinions.
>
> I assume the implication of accepting this CFC is that "longdesc on img  
> elements" will be removed from the section that states that it is  
> non-conforming [1].

I'm not sure, since it is not currently proposed for inclusion in an HTML  
spec. But since the validator would allow longdesc, that seems to be an  
outcome that is consistent...

> I'm supportive of this CFC under the following conditions:
>
> 1. I'd like to make sure that "longdesc on iframe elements" remains  
> non-conforming.

As editor of this spec, I don't have any desire to extend it to frames (or  
anything else). But since it comes from a TF and I take the role of editor  
to mean that, rather than "author entitled to decide everything or stop  
playing", I can't make a guarantee. Withing the task force I ahve  
certainly argued not to extend this spec any further.

> 2. I'd also like to make sure we come up with a more generic means of  
> linking to external long descriptions for any replaced element, not just  
> images.

Me too. As far as I am concerned today (and for the 7-year life of Issue  
30), longdsc ain't it.

> In the long term I'd like to see such a feature and I'd like to see that  
> feature replace @longdesc on images, too.

I have no objection to that either. But one reason I support this is that  
the promised replacements haven't appeared yet.

> I'm therefore suggesting that for HTML5.1 @longdesc on img is added to  
> the list of obsolete but conforming features [2] with a statement that  
> it is there for legacy >reasons and will be replaced by a new aria-*  
> attribute.

I'll leave this question to when it is less hypothetical, but I am  
unlikely to simply reject the idea outright.

> I'm saying aria-* because if it is an accessibility-only attribute, it  
> should be in the aria namespace. Proponents that argue that it's not  
> just for accessibility should >use <a> instead.

Except that doesn't meet some of the use cases. But that's a detailed  
technically-based bikeshed that I think is best painted by whoever  
actually produces the proposed attribute (or other mechanism).

cheers, and thanks for the comments

Chaals


-- 
Charles McCathie Nevile - Consultant (web standards) CTO Office, Yandex
chaals@yandex-team.ru Find more at http://yandex.com