Re: Revert Request

Silvia and Chairs,

On 20th December the chairs said in their review of Jonas' CP, that: 
«this proposal suggests a material change, namely allowing 
aria-describedby to reference hidden elements». And that «this proposal 
should be updated to show that either some use cases cannot be fully 
met without this change, or that some cases cannot be met as 
effectively without this change, or some combination. Otherwise, there 
will be no reason to consider this proposal over the zero change 
proposal».

This review alone, could be enough to ask for a revert.

Note that I tend to agree with Silvia, that from ARIA's point of view, 
then @hidden - as noted in HTML5's ARIA section - is just a synonym for 
@aria-hidden='true'. My reading of ARIA is that it does permit that 
aria-describedby points to an element with @aria--hidden='true'. Like 
Silvia, I'm therefore not certain that the @aria-describedby/@hidden 
change contradicts with @longdesc any more than ARIA as such eventually 
does. A review from the ARIA community to clarify the issue would be 
welcome.

But that said, the @aria-describedby/@hidden change offers an advantage 
compared to the alternative: It is simpler and probably more secure [at 
least in the minds of many site owner] to use compared with the 
alternative, which would be 
      [aria-hidden=true]{display:none}
However, the consequence of Jonas' proposal would be that *no users*, 
except AT users with ARIA supporting software, would get to see such 
image descriptions. In contrast, it has never been important to the 
@longdesc supporters, that @longdesc content gets hidden from other 
users, rather, we want @longdesc content to be available to all. So 
this seems on its head, when we consider that those in opposition to 
@longdesc often argue that use of @longdesc causes content to be hidden 
from non-AT users.

Note as well that there is no permission to link to @hidden content, 
and as such @longdesc could not be used to link to @hidden content, 
whereas @aria-describedby could. Why? Is that fair or logical? Or 
perhaps it is @hidden that needs to change? After all, links are often 
'javascripted' to activate hidden sections. But HTML5 forbids such kind 
of linking.

Also if the zero change proposal, due to its current concurrence with 
Jonas' proposal in this detail, now is considered to reflect what Jonas 
proposed, then should it not offer offer the same use cases as Jonas 
was asked to provide?

So, while it *might* be that this issue is not as problematic Laura 
says, it definitely needs to be reviewed. One way - and the most 
logical way - to have that review, is if the editor reverts this change 
and Jonas updates his CP.

Therefore I concur with Laura's request.

Leif H Silli


Silvia Pfeiffer, Fri, 27 Jan 2012 00:25:46 +1100:
> Hi Laura,
> 
> I don't quite understand why that change is a problem. It expresses
> something that is already codified in ARIA and thus just brings it in
> line with another W3C spec. Also, it has no direct relationship to the
> longdesc issue, it just rectifies the use of aria-describedby with
> hidden text. longdesc has a very different purpose to
> aria-describedby.
> 
> Therefore, I don't see how this circumvented a WG process.
> 
> Regards,
> Silvia.
> 
> 
> On Thu, Jan 26, 2012 at 11:30 PM, Laura Carlson
> <laura.lee.carlson@gmail.com> wrote:
>> Hello Sam, Paul, and Maciej,
>> 
>> As you know the editor made changes to the hidden section [1]. This
>> biases an open issue [2] as it directly implements a material change
>> from a change proposal [3]. The Chairs specifically asked for
>> justification for this change in their change proposal review [4]. If
>> the proposal lacks justification, then the spec lacks justification.
>> 
>> I request that working group process not be bypassed and circumvented
>> and that the change be reverted until such time as the issue is fairly
>> heard and openly decided.
>> 
>> Thank you for your consideration.
>> 
>> Best Regards,
>> Laura
>> 
>> [1] http://html5.org/r/6895

>> [2] http://www.w3.org/html/wg/tracker/issues/30

>> [3] http://tinyurl.com/82rf7vq

>> [4] http://lists.w3.org/Archives/Public/public-html/2011Dec/0182.html

>> 
>> --
>> Laura L. Carlson
>> 
>

Received on Thursday, 26 January 2012 14:13:13 UTC