RE: Drop longdesc, get aria-describedat?

Judy Brewer, Thu, 08 Mar 2012 00:17:49 -0500:
> At 05:54 AM 3/8/2012 +0100, Leif Halvard Silli wrote:
>> John Foliot, Wed, 07 Mar 2012 22:22:44 -0500:

> The longdesc proposal has already been substantially updated in the 
> version of the TF-supported longdesc proposal that is awaiting 
> consideration under the HTML WG decision policy. The changes made 
> already reflect extensive adaptations according to discussion and 
> input from HTML WG participants. Generating yet more proposals, 
> rather than focusing attention on getting a hearing for the proposal 
> that is already on the table and has already received such extensive 
> work, is likely to cause further delay rather than help resolve the 
> issue.

My goal is not, in itself, that the CP proposes the status to become 
'obsolete but conforming', but rather that the transition argument is 
given more weight, because that could increase @longdesc's chances of 
becoming conforming. [I did not see it like that before - then I 
thought the best would to just emphasize the importance of @longdesc.]  
The transition argument could make the CP less controversial and thus 
increase its support.

Of course, 'obsolete but conforming' fits well with a transition 
status, but fully conforming can work too. I don't know for sure what 
the chairs would consider least controversial - but I know that they 
look for 'least controversial'. I now think the reason why 'obsolete 
but conforming' apparently was considered *most* controversial - it 
went out first - in the previous vote, was that it had a thin 
justification. One reason the argumentation was thin was that no one in 
the HTMLWG was aware of the plans about aria-describedAT - if they were 
such plans back then. The current CP does mention @aria-describedAT. As 
such  this - in all likelihood - temporary status of @longdesc, is thus 
not unreflected in the CP. If there comes to a new vote, now, then I 
will probably say that @longdesc should be conforming for the time 
being. [In practice, that would soon be equal to HTML5.]
Leif Halvard Silli

Received on Thursday, 8 March 2012 07:06:21 UTC